2c:null 2d:["$","title",null,{"children":"From the NBA Draft Combine to NBA Career Performance | NBA Draft Combine GroupWork | Observable"}] 32:T407,## Why Is Reaching the NBA So Difficult? The NBA is one of the most exclusive major professional sports league in the world, with only 30 teams and roughly 450 active roster spots at any given time. Every year, tens of thousands of young athletes play organized basketball at the high school level, but only a tiny fraction ever advance to professional basketball, let alone the NBA. Even among Division I college players, statistically fewer than two percent are ever drafted, and many drafted players never sign a long-term contract or make a meaningful career out of the league. The competition is intensified by the increasingly global nature of the sport, as talented prospects now emerge not only from American high schools and colleges but also from international leagues in Europe, Asia, and Australia. On top of raw talent, players must also survive injuries, adapt to increasingly specialized positional demands, and stand out in a scouting environment where teams have access to more data and analysis than ever before. 33:T670,### NBA Draft Combine Data The [NBA Draft Combine dataset on Kaggle](https://www.kaggle.com/datasets/marcusfern/nba-draft-combine) contains data from more than 1,700 players who participated from 2000 onwards. The original data were collected from [NBA Stats](https://www.nba.com/stats/). Not every invited prospect completes all assessments. Highly rated players may decline to participate or skip individual measurements because their expected draft position is already secure. #### Included Attributes | Attribute | Description | Unit | |---|---|---| | **YEAR** | Year of the NBA Draft Combine | Year | | **PLAYER** | Player name | — | | **POS** | Playing position | — | | **HGT** | Player height | Inches | | **WGT** | Player weight | Pounds | | **BMI** | Calculated Body Mass Index | kg/m² | | **BF** | Body-fat percentage | Percent | | **WNGSPN** | Wingspan | Inches | | **STNDRCH** | Standing reach | Inches | | **HANDL** | Hand length | Inches | | **HANDW** | Hand width | Inches | | **STNDVERT** | Standing vertical leap | Inches | | **LPVERT** | Maximum vertical leap | Inches | | **LANE** | Time required to complete the lane-agility test | Seconds | | **SHUTTLE** | Time required to complete the shuttle run | Seconds | | **SPRINT** | Time required to complete the three-quarter-court sprint | Seconds | | **BENCH** | Maximum number of bench-press repetitions | Repetitions | | **BAR** | Ratio of wingspan to height | Ratio | | **PAN** | Hand length multiplied by hand width | Square inches | | **PBHGT** | Standing reach plus standing vertical leap | Inches | | **PDHGT** | Standing reach plus maximum vertical leap | Inches | 34:T8a9,### NBA Draft Results and Career Statistics Historical [NBA Draft results and career statistics from Basketball Reference](https://www.basketball-reference.com/draft/) are used to determine whether a Combine participant was drafted, how early he was selected and how he subsequently performed in the NBA. Basketball Reference provides a separate page for every draft year. These pages contain information about the draft selection as well as the player’s accumulated NBA career statistics. #### Included Attributes | Attribute | Description | |---|---| | **YEAR** | Year of the NBA Draft | | **PICK** | Overall draft position | | **TEAM** | Team that selected the player | | **PLAYER** | Player name | | **COLLEGE** | College attended before entering the NBA | | **YEARS** | Number of NBA seasons played | | **G** | Career games played | | **MP** | Career minutes played | | **PTS** | Career points scored | | **TRB** | Career total rebounds | | **AST** | Career assists | | **FG%** | Career field-goal percentage | | **3P%** | Career three-point percentage | | **FT%** | Career free-throw percentage | | **WS** | Career Win Shares | | **WS/48** | Win Shares per 48 minutes | | **BPM** | Career Box Plus/Minus | | **VORP** | Career Value Over Replacement Player | The overall pick number is used to measure how early a player was selected. A lower value represents an earlier selection: Pick 1 is the first player selected, while Pick 60 is one of the final selections. Career performance should not be assessed using points alone. Older players have had more time to accumulate games, minutes and points. Therefore, efficiency and impact metrics such as **WS/48**, **BPM** and **VORP** should be considered alongside career totals. - **WS/48 (Win Shares per 48 Minutes):** Estimates how much a player contributes to team wins per 48 minutes played. - **BPM (Box Plus/Minus):** Estimates how many points per 100 possessions a player contributes compared with an average player. A positive value indicates above-average performance. - **VORP (Value Over Replacement Player):** Estimates a player's total contribution compared with a replacement-level player. A higher value indicates greater overall impact. 35:T567,### Matching Procedure Combine records and career outcomes were joined on player name and draft year, leaving many participants unmatched due to spelling variants, suffix differences, and name changes between the two sources. A recovery script addressed part of this gap by matching the remaining records against Basketball-Reference draft tables within a bounded year window. An initial surname-based pass produced 83 automatic matches, 70 ambiguous candidates, and 516 records that stayed unmatched, most plausibly undrafted players. Manual verification found 51 of the 83 automatic matches to be incorrect (e.g. a “James” match linking to LeBron James), since surname matching alone did not check first-name plausibility; the script was patched accordingly and covered with regression tests. This yielded 33 net-new verified matches (28 confirmed automatic plus 5 resolved manually), with three ambiguous candidates and four outcome-value lookups still open at the time of writing. ### Consolidation The verified recoveries were merged into the base dataset via an update-in-place operation rather than concatenation, filling missing outcome values into existing rows without duplicating entries or altering the original file. The resulting dataset (2_combine_with_outcomes_with_defense.csv) contains approximately 964 players with complete combine and career-outcome data.36:Tbcd,// ------------------------------------------------------------ // 1. Define the NBA Draft Combine metrics // direction = 1: a higher value is considered better // direction = -1: a lower value is considered better // ------------------------------------------------------------ const combineMetrics = [ {key: "HGT", label: "Height", direction: 1}, {key: "WGT", label: "Weight", direction: 1}, {key: "BMI", label: "BMI", direction: 1}, {key: "WNGSPN", label: "Wingspan", direction: 1}, {key: "STNDRCH", label: "Standing Reach", direction: 1}, {key: "STNDVERT", label: "Standing Vertical", direction: 1}, {key: "LPVERT", label: "Max Vertical", direction: 1}, {key: "LANE", label: "Lane Agility", direction: -1}, {key: "SHUTTLE", label: "Shuttle Run", direction: -1}, {key: "SPRINT", label: "Sprint", direction: -1}, {key: "BENCH", label: "Bench Press", direction: 1} ]; // ------------------------------------------------------------ // 2. Convert values to numbers // Empty or invalid values are returned as null // ------------------------------------------------------------ const toNumber = value => { if ( value === null || value === undefined || value === "" ) { return null; } const number = Number( String(value) .replace("%", "") .replace(",", ".") .trim() ); return Number.isFinite(number) ? number : null; }; // ------------------------------------------------------------ // 3. Standardize player positions // Example: "PG-SG" is assigned to the primary position "PG" // ------------------------------------------------------------ const getPosition = value => { if (!value) return "Unknown"; return String(value) .toUpperCase() .trim() .split("-")[0]; }; // ------------------------------------------------------------ // 4. Determine whether a player was drafted // A player is considered drafted when a valid draft pick exists // or when MATCH_STATUS equals "matched_drafted" // ------------------------------------------------------------ const isDrafted = player => { const draftStatus = String( player["drafted / undrafted"] ?? "" ) .trim() .toLowerCase(); const draftPick = toNumber(player.PICK); return ( draftStatus === "drafted" || Number.isFinite(draftPick) ); }; // ------------------------------------------------------------ // 5. Prepare and enrich the original dataset // ------------------------------------------------------------ const players = cleanData.map(player => ({ ...player, YEAR_NUM: toNumber(player.YEAR), POSITION_GROUP: getPosition(player.POS), DRAFTED: isDrafted(player), PICK_NUM: toNumber(player.PICK) })); // ------------------------------------------------------------ // 6. Extract all available Combine years // ------------------------------------------------------------ const availableYears = Array.from( new Set( players .map(player => player.YEAR_NUM) .filter(Number.isFinite) ) ).sort((a, b) => a - b); 37:Tb75,// ------------------------------------------------------------ // Calculate comparable heatmap data for: // 1. All Combine participants // 2. Drafted players only // ------------------------------------------------------------ const heatmapComparison = (() => { const allPositions = ["PG", "SG", "SF", "PF", "C"]; const positionsToShow = selectedPosition === "All" ? allPositions : [selectedPosition]; // Filter players by the selected Combine year const yearPlayers = players.filter(player => effectiveYear === "All" || player.YEAR_NUM === Number(effectiveYear) ); // Store the standardized results const allPlayerRows = []; const draftedPlayerRows = []; for (const metric of combineMetrics) { // Use all players in the selected year as the common baseline const baselineValues = yearPlayers .map(player => toNumber(player[metric.key])) .filter(Number.isFinite); const baselineMean = d3.mean(baselineValues); const baselineDeviation = d3.deviation(baselineValues); // Skip metrics without enough valid observations if ( baselineValues.length < 2 || !Number.isFinite(baselineMean) || !Number.isFinite(baselineDeviation) || baselineDeviation === 0 ) { continue; } for (const position of positionsToShow) { const positionPlayers = yearPlayers.filter( player => player.POSITION_GROUP === position ); const draftedPositionPlayers = positionPlayers.filter( player => player.DRAFTED ); // Calculate values for all players at this position const allValues = positionPlayers .map(player => toNumber(player[metric.key])) .filter(Number.isFinite); // Calculate values for drafted players at this position const draftedValues = draftedPositionPlayers .map(player => toNumber(player[metric.key])) .filter(Number.isFinite); if (allValues.length > 0) { const groupMean = d3.mean(allValues); const score = ((groupMean - baselineMean) / baselineDeviation) * metric.direction; allPlayerRows.push({ group: "All participants", position, metric: metric.label, score, groupMean, baselineMean, playerCount: allValues.length }); } if (draftedValues.length > 0) { const groupMean = d3.mean(draftedValues); const score = ((groupMean - baselineMean) / baselineDeviation) * metric.direction; draftedPlayerRows.push({ group: "Drafted players", position, metric: metric.label, score, groupMean, baselineMean, playerCount: draftedValues.length }); } } } return { all: allPlayerRows, drafted: draftedPlayerRows }; })();38:T5c3,// Create a heatmap const createHeatmap = (data, title) => { return Plot.plot({ width: 560, height: 440, marginTop: 50, marginLeft: 125, marginRight: 20, marginBottom: 45, title, x: { label: "Position", domain: ["PG", "SG", "SF", "PF", "C"] }, y: { label: null, domain: combineMetrics.map(d => d.label) }, color: { type: "diverging", scheme: "RdBu", pivot: 0, domain: [-1.5, 1.5], legend: true, label: "Relative to all players", ticks: [-1, 0, 1], tickFormat: value => { if (value === -1) return "Below average"; if (value === 0) return "Average"; if (value === 1) return "Above average"; return ""; } }, marks: [ Plot.cell(data, { x: "position", y: "metric", fill: "score", inset: 1, title: d => `${d.position} – ${d.metric} Score: ${d.score.toFixed(2)} Mean: ${d.groupMean.toFixed(2)} Players: ${d.playerCount}` }), // Display the actual group mean inside each cell Plot.text(data, { x: "position", y: "metric", text: d => { const decimals = [ "Lane Agility", "Shuttle Run", "Sprint" ].includes(d.metric) ? 2 : 1; return d.groupMean.toFixed(decimals); }, // The text color is based on the standardized score fill: d => Math.abs(d.score) >= 0.7 ? "white" : "black", fontSize: 10 }), Plot.frame() ] }); };39:T4eb,// ------------------------------------------------------------ // Define NBA career metrics for drafted players // ------------------------------------------------------------ const nbaPositionMetrics = [ { key: "GAMES", label: "Career Games", decimals: 0 }, { key: "MP_PER_G", label: "Minutes per Game", decimals: 1 }, { key: "PTS_PER_G", label: "Points per Game", decimals: 1 }, { key: "TRB_PER_G", label: "Rebounds per Game", decimals: 1 }, { key: "AST_PER_G", label: "Assists per Game", decimals: 1 }, { key: "FG_PCT", label: "Field Goal %", format: "percentage" }, { key: "FG3_PCT", label: "Three-Point %", format: "percentage" }, { key: "FT_PCT", label: "Free Throw %", format: "percentage" }, { key: "WS", label: "Win Shares", decimals: 1 }, { key: "WS/48", label: "Win Shares per 48", decimals: 3 }, { key: "BPM", label: "Box Plus/Minus", decimals: 1 }, { key: "VORP", label: "VORP", decimals: 1 }, { key: "DWS", label: "Defensive Win Shares", decimals: 1 }, { key: "DBPM", label: "Defensive Box Plus/Minus", decimals: 1 } ];3a:T7b1,// ------------------------------------------------------------ // Calculate average NBA performance by playing position // Only drafted players are included // ------------------------------------------------------------ const nbaPositionHeatmapData = (() => { const allPositions = ["PG", "SG", "SF", "PF", "C"]; const positionsToShow = selectedPosition === "All" ? allPositions : [selectedPosition]; // Select drafted players for the selected year or all years const draftedPlayers = players.filter(player => { const yearMatches = effectiveYear === "All" || player.YEAR_NUM === Number(effectiveYear); return player.DRAFTED && yearMatches; }); const rows = []; for (const metric of nbaPositionMetrics) { // Use all drafted players as the baseline const baselineValues = draftedPlayers .map(player => toNumber(player[metric.key])) .filter(Number.isFinite); const baselineMean = d3.mean(baselineValues); const baselineDeviation = d3.deviation(baselineValues); if ( baselineValues.length < 2 || !Number.isFinite(baselineMean) || !Number.isFinite(baselineDeviation) || baselineDeviation === 0 ) { continue; } for (const position of positionsToShow) { const positionPlayers = draftedPlayers.filter( player => player.POSITION_GROUP === position ); const positionValues = positionPlayers .map(player => toNumber(player[metric.key])) .filter(Number.isFinite); if (positionValues.length === 0) { continue; } const groupMean = d3.mean(positionValues); const score = (groupMean - baselineMean) / baselineDeviation; rows.push({ position, metric: metric.label, metricKey: metric.key, score, groupMean, baselineMean, playerCount: positionValues.length }); } } return rows; })();3b:T720,// ------------------------------------------------------------ // Create the NBA performance heatmap // ------------------------------------------------------------ const nbaPositionPlot = Plot.plot({ width: 850, height: 500, marginTop: 60, marginLeft: 145, marginRight: 30, marginBottom: 50, title: `NBA career performance by position – ${ effectiveYear === "All" ? "All years" : effectiveYear } (drafted players only, n = ${playerCounts.drafted})`, x: { label: "Position", domain: selectedPosition === "All" ? ["PG", "SG", "SF", "PF", "C"] : [selectedPosition] }, y: { label: null, domain: nbaPositionMetrics.map( metric => metric.label ) }, color: { type: "diverging", scheme: "RdBu", pivot: 0, domain: [-1.5, 1.5], legend: true, label: "Relative to all drafted players", ticks: [-1, 0, 1], tickFormat: value => { if (value === -1) return "Below average"; if (value === 0) return "Average"; if (value === 1) return "Above average"; return ""; } }, marks: [ // Draw colored cells based on relative performance Plot.cell(nbaPositionHeatmapData, { x: "position", y: "metric", fill: "score", inset: 1, title: d => `${d.position} – ${d.metric} Average: ${formatNbaPositionValue(d)} Overall drafted-player average: ${d.baselineMean.toFixed(2)} Players with valid data: ${d.playerCount}` }), // Display the actual average value inside each cell Plot.text(nbaPositionHeatmapData, { x: "position", y: "metric", text: d => formatNbaPositionValue(d), fill: d => Math.abs(d.score) >= 0.7 ? "white" : "black", fontSize: 10 }), Plot.frame() ] });3c:Td04,### Player Biography James Harden was born on August 26, 1989, in Los Angeles, California. He played college basketball at Arizona State University before declaring for the 2009 NBA Draft, where he was selected 3rd overall by the Oklahoma City Thunder. Over a 17-year career, he has gone on to play for the Houston Rockets, Brooklyn Nets, Philadelphia 76ers, and Los Angeles Clippers. As of 2026, he plays for the Cleveland Cavaliers, having been traded there from the Clippers in February 2026 in exchange for Darius Garland. ### Career Development Harden began his career as Sixth Man of the Year in Oklahoma City (2012) before being traded to Houston, where he developed into one of the most dominant offensive players of his generation, winning three consecutive scoring titles and the NBA MVP award in 2018. His Brooklyn stint showcased an all-around, playmaking-heavy version of his game, averaging close to a 23/11/8 stat line, while his time in Philadelphia saw him lead the league in assists. In Los Angeles, he shifted into more of a facilitator role, still averaging around 22–25 points per game on efficient shooting well into his mid-30s. Across his career, he has earned 11 All-Star selections and multiple All-NBA team honors. ### Personality, Playing Style and Strengths Beyond the numbers, James Harden's game and public image have been shaped by a distinctive blend of offensive craft, confidence, and a career-long willingness to reshape his circumstances on his own terms. #### Personality - Confident, outspoken, and self-assured in interviews and public statements - Has repeatedly pushed for trades when dissatisfied with his situation (Houston, Brooklyn, Philadelphia) - Known for responding to criticism or trash talk through his on-court performance rather than verbal exchanges - Polarizing public figure — provokes strong reactions from both supporters and critics #### Playing Style - Signature step-back three-pointer, a shot he helped popularize league-wide - Historically isolation-heavy scorer, especially during his Houston prime - Exceptional ability to draw fouls and get to the free-throw line - Has shifted from primary scorer to facilitator/playmaker as his athleticism has declined with age #### Strengths - Elite scoring efficiency across multiple eras of his career - Advanced footwork and change-of-pace ball handling - High-level court vision and passing as a secondary/primary playmaker - Free-throw generation among the best in NBA history - Longevity — remains a productive offensive player 17 seasons into his career ### Public Perception Harden is widely respected among peers and analysts as one of the most offensively gifted players in NBA history — a former teammate recently noted that despite never winning a championship, he ranks among the greatest scorers the league has seen, while receiving disproportionate criticism compared to other all-time greats. Much of that criticism centers on his playoff track record, a narrative the media has repeatedly revisited throughout his career. His 2023 comment "I am the system" after joining the Clippers also became a widely mocked moment on social media following the team's early-season struggles, reflecting how his public image is often shaped by a handful of viral moments as much as by his statistical dominance.3d:Tdf8,### Player Biography Devin Armani Booker was born on October 30, 1996, in Grand Rapids, Michigan, and played one season at the University of Kentucky (2014–15), where he was named SEC Sixth Man of the Year. The Phoenix Suns drafted him 13th overall in the 2015 NBA Draft, and he has played his entire career with the franchise since. Now 29 years old and in his 11th NBA season, he remains Phoenix’s franchise cornerstone after signing a contract extension in 2025 that keeps him with the Suns through at least the 2029–30 season, even as the roster around him was rebuilt following the Kevin Durant trade and Bradley Beal buyout. ### Career Development Booker has been named an NBA All-Star five times (2020–2022, 2024, 2026), made the All-NBA First Team in 2022 and All-NBA Third Team in 2024, and won the NBA Three-Point Contest in 2018. His breakout moment came early: at age 20 he scored a career-high 70 points against the Boston Celtics on March 24, 2017. He led Phoenix to its first NBA Finals appearance in 28 years in 2021, and has since won Olympic gold with Team USA in Tokyo 2020 and a second gold medal in Paris 2024. An NBA championship remains the one accolade missing from his résumé. Over 650 career games he has averaged 24.4 points, 4.0 rebounds, and 5.1 assists per game. ### Personality, Playing Style and Strengths Devin Booker built his game around one of the smoothest scoring repertoires in the league, anchored by a quick-release jumper and an increasingly featured playmaking role. Off the court he carries himself with unusual maturity for his age, while on it he channels a sharper, more combative alter-ego that stands in contrast to his easygoing public image. ### Personality - Widely regarded as mature beyond his years; teammate Chris Paul once called him "the oldest 24-year-old I ever met in my life" - Runs the charitable fund Devin Booker's Starting Five - More vocal and demanding on court than his easygoing public image suggests, publicly pushing for better in-game communication in 2024-25, which reportedly prompted coach Mike Budenholzer to ask him to tone it down - Keeps an on-court alter-ego, "Book," distinct from his mild-mannered public persona; former teammate Marquese Chriss calls him a relentless trash-talker - Booker describes it as mostly cheering himself on in the third person, a habit visible since his rookie season, when he led all first-year players in technical fouls ### Playing Style - Functions as a high-usage perimeter scorer built around a quick-release jumper and pull-up game - Historically more effective creating his own shot through isolation and pick-and-roll than scoring off movement ### Strengths - Elite shot-making from midrange and three-point range - Reliable free-throw shooting - Growing playmaking responsibility as his role has shifted toward more point-guard duties in recent seasons ### Public Perception Perception of Booker is split. The Athletic still ranked him among the league’s top shooting guards entering 2025–26 despite a down prior season, yet a 2025 NBA general managers’ survey left him out of the top five at both guard positions, and he stood just 17th in early 2026 All-Star fan voting despite outscoring several players ranked above him. He still made the team regardless. His 2024 Olympic run pushed back on that undervaluation: head coach Steve Kerr called him the team’s “unsung MVP” for his reduced, defense-first role next to LeBron James and Stephen Curry. The missing NBA title remains the most frequently cited gap in his legacy.3e:Td91,### Player Biography Stephen Curry was born on March 14, 1988, in Akron, Ohio, and grew up in Charlotte, North Carolina. He played college basketball at Davidson College, where his exceptional shooting and NCAA Tournament performances attracted national attention. Curry was selected 7th overall by the Golden State Warriors in the 2009 NBA Draft. As of 2026, he is 38 years old and has spent his entire 17-year NBA career with the Warriors. ### Career Development Curry initially struggled with recurring ankle injuries, but developed into the central figure of the Warriors' dynasty. He led Golden State to four NBA championships in 2015, 2017, 2018 and 2022, while winning two regular-season MVP awards, including the first unanimous MVP award in NBA history in 2016. In 2022, he received his first NBA Finals MVP award after leading the Warriors past the Boston Celtics. Curry also became the NBA's all-time leader in three-point field goals and fundamentally changed how teams approach shooting, spacing and offensive strategy. ### Personality, Playing Style and Strengths Stephen Curry stands out through his combination of an approachable personality, extraordinary shooting ability and constant movement without the ball. Unlike many NBA superstars, he can dominate a game without controlling possession for long periods. #### Personality - Combines a calm and team-oriented personality with visible confidence during decisive moments - Leads primarily through performance and consistency rather than confrontation - Teammates often emphasize his willingness to share the spotlight and adapt his game for the success of the team - Shows a notable contrast between his approachable public image and highly competitive on-court presence #### Playing Style - Moves constantly without the ball to create shooting opportunities and disrupt defensive structures - Attempts and converts shots from far beyond the traditional three-point line - Releases the ball exceptionally quickly, both after receiving a pass and after dribbling - Uses advanced ball handling to create shots from almost any position - Forces defenders to guard him unusually far from the basket, creating additional space for his teammates #### Strengths - Widely regarded as the greatest three-point shooter in NBA history - Exceptional combination of shooting range, accuracy and release speed - Extraordinary conditioning that allows him to move continuously while maintaining shooting efficiency - Creates exceptional defensive “gravity”, attracting multiple defenders even when he does not have the ball - Difficult to defend because he remains dangerous both with and without possession - Has overcome physical disadvantages in size and strength through skill, coordination and spatial awareness ### Public Perception Curry is widely regarded as one of the most influential players in basketball history because his success helped establish three-point shooting as a central part of modern NBA offense. Fans and analysts frequently praise his shooting ability, team-oriented approach and role in transforming the Warriors into a dynasty. Some critics have questioned his defense, physical size and ability to dominate without strong teammates, but his 2022 championship and Finals MVP performance weakened many of these arguments. His approachable personality, spectacular shooting and visible enjoyment of the game have made him one of the NBA's most popular and recognizable players.3f:Tfd4,### Player Biography LeBron James was born on December 30, 1984, in Akron, Ohio. He entered the NBA directly out of St. Vincent–St. Mary High School, selected 1st overall by the Cleveland Cavaliers in the 2003 NBA Draft. Over a 24-year career, he has played for the Cleveland Cavaliers (2003–2010, 2014–2018), Miami Heat (2010–2014), Los Angeles Lakers (2018–2026), and, as of 2026, the Philadelphia 76ers. His departure from Los Angeles came after he informed the Lakers he would play for a different team in the 2026-27 season, entering his 24th NBA season at age 41. ### Career Development James has won four NBA championships (2012, 2013, 2016, 2020), four NBA Finals MVP awards, and four NBA Most Valuable Player awards, alongside 22 All-Star selections. His career-average stat line stands at 26.8 points, 7.5 rebounds and 7.4 assists per game, and in his final Lakers season (2025-26) he still averaged 20.9 points, 6.1 rebounds and 7.2 assists across 60 games at age 41. That postseason, he improved his play in the playoffs to 23.2 points, 6.7 rebounds and 7.3 assists across 10 postseason appearances, leading the Lakers past Houston before falling to Oklahoma City in the second round. He remains widely regarded, per Britannica, as one of the greatest all-around players of all time, having won titles across three different franchises. ### Personality, Playing Style and Strengths LeBron James built his game around an unmatched combination of size, speed and basketball IQ for his position, evolving from a downhill physical scorer in his youth into a pass-first floor general in the twilight of his career. Off the court he operates less like a traditional athlete and more like a media and business executive, carefully building a brand and an empire around the same stage-driven instincts that shape how he plays. ### Personality - Extrinsically motivated and stage-driven — his 2010 free-agency announcement drew over 13 million television viewers, reflecting a career built on high-visibility moments - Competes best against named rivals and public doubters rather than abstract personal benchmarks - Combines on-court dominance with an unusually deliberate, business-minded off-court persona - Deeply tied to his hometown Akron, both symbolically and through his philanthropic and educational ventures ### Playing Style - Rare combination of size, strength and court vision for his position, able to play and defend multiple positions - Elite passer and secondary/primary playmaker, having led the league in assists in 2020 despite his frame - Physical, downhill scoring style in his prime, evolved into a more efficient, facilitation-heavy game in his 40s - Exceptional durability and conditioning — remained a productive 20-point, 7-assist player at age 41, playing 71 games in one recent season, his highest total since 2017-18 ### Strengths - Longevity: still an elite, multi-category producer more than two decades into his career - Basketball IQ and adaptability — has reshaped his game repeatedly to fit new teammates, coaches and physical limits - Leadership and off-court influence extending well beyond basketball - A business empire including a production company, part-ownership stakes in Fenway Sports Group, Liverpool FC and other ventures, and status as one of the first billionaire active athletes ### Public Perception James is a central figure in the modern GOAT (Greatest of All Time) debate, with supporters citing his championships across three franchises, longevity and off-court impact, including the I PROMISE School he founded in Akron for at-risk children. Critics and commentators have also scrutinized his business-first approach to activism and celebrity, with some arguing his empire-building sits in tension with the more grassroots tradition of athlete activism. His departure from the Lakers ahead of the 2026-27 season, in search of a fifth championship with Philadelphia, has renewed public debate about how much longer he can remain a difference-maker at the highest level.40:T422,const players1 = [ {player: "James Harden", team: "Cleveland Cavaliers", age: 37, seasons: 17, championships: 0, awards: "MVP (2018), 3x Scoring Champion, 6th Man of the Year (2012), 2nd all-time in career 3-pointers made", allStar: 11, salary: 32300000, endorsements: 16000000}, {player: "Devin Booker", team: "Phoenix Suns", age: 29, seasons: 11, championships: 0, awards: "All-NBA First Team (2022), All-NBA Third Team (2024), 70-point game (2017), 2x Olympic gold (2020, 2024)", allStar: 5, salary: 57100000, endorsements: 10000000}, {player: "Stephen Curry", team: "Golden State Warriors", age: 38, seasons: 17, championships: 4, awards: "2x MVP (2015, 2016), Finals MVP (2022), NBA all-time leader in career 3-pointers made, 2x Scoring Champion", allStar: 12, salary: 62600000, endorsements: 50000000}, {player: "LeBron James", team: "Philadelphia 76ers", age: 41, seasons: 23, championships: 4, awards: "4x Finals MVP, 4x MVP, NBA all-time leading scorer, record 21x All-NBA selections", allStar: 21, salary: 3900000, endorsements: 80000000} ]41:Tc1e,// ------------------------------------------------------------ // Define the four selected players // The dataset uses the format "Last name, First name" // ------------------------------------------------------------ const comparisonPlayers = [ { playerName: "Harden, James" }, { playerName: "Booker, Devin" }, { playerName: "Curry, Stephen" }, { playerName: "James, LeBron" } ]; // ------------------------------------------------------------ // Define the NBA Draft Combine metrics // direction = 1: higher values receive a higher relative score // direction = -1: lower values receive a higher relative score // ------------------------------------------------------------ const playerCombineMetrics = [ { key: "HGT", label: "Height", direction: 1, decimals: 1 }, { key: "WGT", label: "Weight", direction: 1, decimals: 1 }, { key: "BMI", label: "BMI", direction: 1, decimals: 1 }, { key: "WNGSPN", label: "Wingspan", direction: 1, decimals: 1 }, { key: "STNDRCH", label: "Standing Reach", direction: 1, decimals: 1 }, { key: "STNDVERT", label: "Standing Vertical", direction: 1, decimals: 1 }, { key: "LPVERT", label: "Max Vertical", direction: 1, decimals: 1 }, { key: "LANE", label: "Lane Agility", direction: -1, decimals: 2 }, { key: "SHUTTLE", label: "Shuttle Run", direction: -1, decimals: 2 }, { key: "SPRINT", label: "Sprint", direction: -1, decimals: 2 }, { key: "BENCH", label: "Bench Press", direction: 1, decimals: 0 } ]; // ------------------------------------------------------------ // Define the NBA career metrics // ------------------------------------------------------------ const playerNbaMetrics = [ { key: "MP_PER_G", label: "Minutes per Game", direction: 1, decimals: 1 }, { key: "PTS_PER_G", label: "Points per Game", direction: 1, decimals: 1 }, { key: "TRB_PER_G", label: "Rebounds per Game", direction: 1, decimals: 1 }, { key: "AST_PER_G", label: "Assists per Game", direction: 1, decimals: 1 }, { key: "FG_PCT", label: "Field Goal %", direction: 1, format: "percentage" }, { key: "FG3_PCT", label: "Three-Point %", direction: 1, format: "percentage" }, { key: "FT_PCT", label: "Free Throw %", direction: 1, format: "percentage" }, { key: "WS", label: "Win Shares", direction: 1, decimals: 1 }, { key: "WS/48", label: "Win Shares per 48", direction: 1, decimals: 3 }, { key: "BPM", label: "Box Plus/Minus", direction: 1, decimals: 1 }, { key: "VORP", label: "VORP", direction: 1, decimals: 1 }, { key: "DWS", label: "Defensive Win Shares", direction: 1, decimals: 1 }, { key: "DBPM", label: "Defensive Box Plus/Minus", direction: 1, decimals: 1 } ];42:Tc01,// ------------------------------------------------------------ // Format the actual values displayed inside the heatmap cells // ------------------------------------------------------------ const formatPlayerMetricValue = (value, metric) => { if (!Number.isFinite(value)) return "N/A"; if (metric.format === "percentage") { const percentageValue = Math.abs(value) <= 1 ? value * 100 : value; return `${percentageValue.toFixed(1)}%`; } return value.toFixed(metric.decimals ?? 1); }; // ------------------------------------------------------------ // Calculate the player profiles // All selected players are compared with shooting guards // ------------------------------------------------------------ const calculatePlayerHeatmap = metrics => { const rows = []; // Use all shooting guards as the common comparison group const shootingGuards = cleanData.filter( player => getPosition(player.POS) === "SG" ); for (const selectedPlayer of selectedPlayerRecords) { for (const metric of metrics) { const playerRecord = selectedPlayer.record; // Create an empty cell when the player was not found if (!playerRecord) { rows.push({ player: selectedPlayer.displayName, playerName: selectedPlayer.playerName, metric: metric.label, value: null, formattedValue: "N/A", score: null, baselineMean: null, position: "Unknown", baselinePosition: "SG", baselineCount: 0 }); continue; } const actualPosition = playerRecord.POS; const playerValue = toNumber( playerRecord[metric.key] ); // Calculate the SG comparison values const baselineValues = shootingGuards .map(player => toNumber(player[metric.key])) .filter(Number.isFinite); const baselineMean = d3.mean(baselineValues); const baselineDeviation = d3.deviation(baselineValues); const validScore = Number.isFinite(playerValue) && Number.isFinite(baselineMean) && Number.isFinite(baselineDeviation) && baselineDeviation > 0; // Calculate the player's deviation from the SG average const score = validScore ? ( (playerValue - baselineMean) / baselineDeviation ) * (metric.direction ?? 1) : null; rows.push({ player: selectedPlayer.displayName, playerName: selectedPlayer.playerName, metric: metric.label, value: playerValue, formattedValue: formatPlayerMetricValue( playerValue, metric ), score, baselineMean, position: actualPosition, baselinePosition: "SG", baselineCount: baselineValues.length }); } } return rows; }; // Calculate the two datasets const selectedCombineHeatmapData = calculatePlayerHeatmap(playerCombineMetrics); const selectedNbaHeatmapData = calculatePlayerHeatmap(playerNbaMetrics);43:T96b,// ------------------------------------------------------------ // Create a player comparison heatmap // // Cell value: actual player statistic // Cell color: performance relative to the player's position // Grey cell: measurement is not available // ------------------------------------------------------------ const createPlayerHeatmap = (data, metrics, title) => { const playerDomain = selectedPlayerRecords.map( player => player.displayName ); const validData = data.filter( row => Number.isFinite(row.score) ); return Plot.plot({ width: 720, height: 500, marginTop: 55, marginLeft: 150, marginRight: 25, marginBottom: 110, title, x: { label: null, domain: playerDomain, tickRotate: -25 }, y: { label: null, domain: metrics.map(metric => metric.label) }, color: { type: "diverging", scheme: "RdBu", pivot: 0, domain: [-2, 2], legend: true, label: "Relative to SG average", ticks: [-1, 0, 1], tickFormat: value => { if (value === -1) return "Below"; if (value === 0) return "Average"; if (value === 1) return "Above"; return ""; } }, marks: [ // Draw a grey background for every possible cell Plot.cell(data, { x: "player", y: "metric", fill: "#eeeeee", inset: 1, title: d => `${d.playerName} Position: ${d.position} ${d.metric}: ${d.formattedValue} Comparison group: ${d.baselineCount} players` }), // Draw colored cells where valid values are available Plot.cell(validData, { x: "player", y: "metric", fill: "score", inset: 1, title: d => `${d.playerName} Actual position: ${d.position} Comparison position: SG ${d.metric}: ${d.formattedValue} Relative score: ${d.score.toFixed(2)} SG average: ${ Number.isFinite(d.baselineMean) ? d.baselineMean.toFixed(2) : "N/A" } Comparison group: ${d.baselineCount} shooting guards` }), // Display the actual values inside the cells Plot.text(data, { x: "player", y: "metric", text: "formattedValue", fill: d => Number.isFinite(d.score) && Math.abs(d.score) >= 1 ? "white" : "black", fontSize: 10 }), Plot.frame() ] }); };44:T646,The player profiles demonstrate that successful NBA careers can emerge from substantially different physical and athletic characteristics. James Harden recorded comparatively high weight, BMI and bench-press results, while Devin Booker performed particularly well in the lane-agility test. Stephen Curry entered the league with a lighter physical profile but later developed into the strongest three-point and free-throw shooter among the four selected players. LeBron James had exceptional size and weight for his position, although several athletic Combine measurements are unavailable in the dataset. The NBA career results do not follow a single consistent Combine pattern. LeBron James records the highest values across several career-impact measures, including points, rebounds, Win Shares, BPM and VORP. Harden combines strong scoring with high assist production and advanced impact values. Curry’s NBA profile is particularly distinguished by his three-point percentage, free-throw percentage and BPM. Booker also records strong scoring numbers, despite lower cumulative and advanced career values than the other three players. Overall, the comparison suggests that no single physical or athletic Combine profile is sufficient to explain later NBA success. Players with different Combine strengths can achieve high professional performance through abilities such as shooting, playmaking, decision-making and technical development. The four-player comparison therefore provides an exploratory overview and motivates the subsequent statistical analysis of shooting guards across the full dataset.45:Tfb7,const regressionHelpers = (() => { const predictors = [ "HGT", "WGT", "WNGSPN", "STNDRCH" ]; const solveLinearSystem = (matrix, vector) => { const size = vector.length; const augmented = matrix.map((row, index) => [ ...row, vector[index] ]); for (let column = 0; column < size; column++) { let pivotRow = column; for (let row = column + 1; row < size; row++) { if ( Math.abs(augmented[row][column]) > Math.abs(augmented[pivotRow][column]) ) { pivotRow = row; } } [augmented[column], augmented[pivotRow]] = [ augmented[pivotRow], augmented[column] ]; const pivot = augmented[column][column]; if (Math.abs(pivot) < 1e-12) { throw new Error("Regression matrix is singular."); } for (let item = column; item <= size; item++) { augmented[column][item] /= pivot; } for (let row = 0; row < size; row++) { if (row === column) continue; const factor = augmented[row][column]; for (let item = column; item <= size; item++) { augmented[row][item] -= factor * augmented[column][item]; } } } return augmented.map(row => row[size]); }; const fitModel = (rows, target) => { const modelRows = rows .map(row => ({ row, targetValue: target.value(row) })) .filter(item => predictors.every(key => Number.isFinite(Number(item.row[key])) ) && Number.isFinite(item.targetValue) ); const means = Object.fromEntries( predictors.map(key => [ key, d3.mean(modelRows, item => Number(item.row[key])) ]) ); const deviations = Object.fromEntries( predictors.map(key => [ key, d3.deviation( modelRows, item => Number(item.row[key]) ) || 1 ]) ); const designMatrix = modelRows.map(item => [ 1, ...predictors.map(key => (Number(item.row[key]) - means[key]) / deviations[key] ) ]); const outcomes = modelRows.map( item => item.targetValue ); const parameterCount = predictors.length + 1; const xtx = Array.from( {length: parameterCount}, (_, rowIndex) => Array.from( {length: parameterCount}, (_, columnIndex) => d3.sum( designMatrix, values => values[rowIndex] * values[columnIndex] ) ) ); const xty = Array.from( {length: parameterCount}, (_, columnIndex) => d3.sum( designMatrix, (values, rowIndex) => values[columnIndex] * outcomes[rowIndex] ) ); const coefficients = solveLinearSystem(xtx, xty); const predictions = modelRows.map((item, index) => { const predicted = d3.sum( designMatrix[index], (value, coefficientIndex) => value * coefficients[coefficientIndex] ); const playerName = String(item.row.PLAYER) .split(", ") .reverse() .join(" "); return { playerName, actual: item.targetValue, predicted, featured: [ "James Harden", "Devin Booker", "Stephen Curry", "LeBron James" ].includes(playerName) }; }); const outcomeMean = d3.mean(outcomes); const totalVariation = d3.sum( outcomes, value => (value - outcomeMean) ** 2 ); const residualVariation = d3.sum( predictions, point => (point.actual - point.predicted) ** 2 ); return { ...target, predictors, predictions, sampleSize: predictions.length, rSquared: totalVariation > 0 ? 1 - residualVariation / totalVariation : null }; }; return { predictors, fitModel }; })();46:T582,## Can Physical Combine Measurements Predict NBA Career Performance? Three multiple linear regression models were estimated to examine whether physical measurements from the NBA Draft Combine are associated with later career performance. The models predict **WS/48**, **BPM**, and **VORP per season** for drafted shooting guards and players whose listed position includes shooting guard. Only four Combine variables were used as predictors: - **Height** - **Weight** - **Wingspan** - **Standing reach** These variables were selected because they were available for all four featured players. LeBron James did not complete several athletic Combine drills, meaning that values such as vertical jump, lane agility, shuttle run, sprint time and bench press were unavailable. Restricting the models to the four shared physical measurements allows James Harden, Devin Booker, Stephen Curry and LeBron James to be included without artificially replacing missing values. The horizontal axis shows the performance predicted from the four Combine measurements, while the vertical axis shows the player's actual career performance. Grey points represent the remaining players in the regression sample, and the four featured players are highlighted. The dashed diagonal line represents a perfect prediction: players above the line performed better than predicted, while players below it performed worse than predicted.47:T9a6,const regressionPlots = careerRegressionModels.map(model => { const allValues = model.predictions.flatMap(point => [ point.actual, point.predicted ]); const [minimum, maximum] = d3.extent(allValues); const padding = (maximum - minimum) * 0.08 || 1; const domain = [ minimum - padding, maximum + padding ]; const featuredPlayers = model.predictions.filter( point => point.featured ); return Plot.plot({ width: 470, height: 430, marginTop: 55, marginRight: 35, marginBottom: 65, marginLeft: 70, title: `${model.label}: Predicted vs. Actual`, subtitle: `n = ${ model.sampleSize } & R² = ${ model.rSquared.toFixed(3) }`, x: { label: `Predicted ${model.label} from Combine profile`, domain, grid: true }, y: { label: `Actual ${model.label}`, domain, grid: true }, color: { domain: [ "James Harden", "Devin Booker", "Stephen Curry", "LeBron James" ], range: [ "#7f1d1d", "#f97316", "#2563eb", "#7e22ce" ], legend: true }, marks: [ Plot.line( [ {predicted: domain[0], actual: domain[0]}, {predicted: domain[1], actual: domain[1]} ], { x: "predicted", y: "actual", stroke: "#777", strokeDasharray: "6,5" } ), Plot.dot(model.predictions, { x: "predicted", y: "actual", r: 3.5, fill: "#b8b8b8", opacity: 0.55, tip: true, title: point => `${point.playerName} Predicted: ${point.predicted.toFixed(3)} Actual: ${point.actual.toFixed(3)}` }), Plot.dot(featuredPlayers, { x: "predicted", y: "actual", fill: "playerName", stroke: "white", strokeWidth: 1.5, r: 7, tip: true }), Plot.text(featuredPlayers, { x: "predicted", y: "actual", text: "playerName", dx: 9, dy: -9, textAnchor: "start", fontSize: 10, fontWeight: 600 }) ] }); }); display(html`

${regressionPlots.map(plot => html`

${plot}

`)}

`);48:T83d,# Shooting Statistics Across Careers ## Approach Season-by-season shooting statistics (FG%, 3P%, FT%, eFG%, TS%) and value metrics (WS/48, BPM, VORP) were retrieved from Basketball-Reference for all four players, covering 68 player-seasons in total (Harden: 17, Curry: 17, James: 23, Booker: 11). Career-year alignment (season 1 = rookie year), rather than calendar season, is used throughout, so that a player's fifth season can be compared directly against another player's fifth season regardless of when either was drafted. Every value is percentile-ranked against the broader Shooting Guard cohort used elsewhere in this project (n = 338, drafted players with SG in their position label), so a percentile in this section always answers the same question: how does this number compare to everyone drafted at the same position since 2000, not just to three other superstars. Single-season values in the radar's season view are ranked against this same career-aggregate reference distribution, since season-level data for the full cohort was not collected; this is an approximation worth keeping in mind for early-career seasons, where small sample sizes are more likely to distort the picture. ## A shared ceiling, reached by different roads Plot true shooting percentage against career year for all four players and the first thing that stands out is how little separates them. Despite four visibly different combine profiles, all four settle into a comparatively narrow band of 0.53 to 0.68 over the sample, close enough that the lines cross each other repeatedly rather than staying neatly stacked. Booker's path there is the clearest development story in the dataset: a TS% of 0.535 as a 19-year-old rookie, climbing gradually to a career-best 0.618 in year five. Curry's line tells a different kind of story: a sharp single-season dip to 0.557 in 2019-20 that looks, out of context, like a real decline. It is not. He played five games that season before a hand injury ended it early, a reminder that a single data point built on a handful of games can look like a signal when it is really noise.49:Tb4a,sgTrendChart = { const width = 640, height = 420; const margin = {top: 24, right: 24, bottom: 36, left: 48}; const metricLabel = sgTrendMetric === "TS_pct" ? "TS%" : "eFG%"; const data = sgTrendData.filter(d => sgRadarSelectedPlayers.includes(d.Player)); const maxYear = d3.max(data, d => d.career_yr); const x = d3.scaleLinear().domain([1, maxYear]).range([margin.left, width - margin.right]); const y = d3.scaleLinear().domain(d3.extent(data, d => d[sgTrendMetric])).nice() .range([height - margin.bottom, margin.top]); const svg = d3.create("svg") .attr("viewBox", [0, 0, width, height]) .attr("width", width) .attr("height", height); svg.append("g") .attr("transform", `translate(0,${height - margin.bottom})`) .call(d3.axisBottom(x).ticks(8).tickFormat(d3.format("d"))) .call(g => g.append("text") .attr("x", width - margin.right).attr("y", 32) .attr("fill", "currentColor").attr("text-anchor", "end") .text("Season")); svg.append("g") .attr("transform", `translate(${margin.left},0)`) .call(d3.axisLeft(y).tickFormat(d3.format(".0%"))) .call(g => g.select(".domain").remove()); const line = d3.line().x(d => x(d.career_yr)).y(d => y(d[sgTrendMetric])); const byPlayer = d3.groups(data, d => d.Player); const paths = svg.append("g") .attr("fill", "none").attr("stroke-width", 2.5) .selectAll("path") .data(byPlayer) .join("path") .attr("d", ([, rows]) => line(rows.sort((a, b) => a.career_yr - b.career_yr))) .attr("stroke", ([player]) => sgRadarColor(player)) .attr("stroke-opacity", 0.9); svg.append("g") .selectAll("circle") .data(data) .join("circle") .attr("cx", d => x(d.career_yr)) .attr("cy", d => y(d[sgTrendMetric])) .attr("r", d => Math.max(2.5, Math.sqrt(d.GP) * 0.6)) // GP-skaliert: Currys 5-Spiele-Saison faellt auf .attr("fill", d => sgRadarColor(d.Player)) .on("mouseover", (event, d) => { paths.attr("stroke-opacity", p => p[0] === d.Player ? 1 : 0.08); sgRadarTooltip.style("opacity", 1) .html(`${d.Player}
${d.Season} (Jahr ${d.career_yr})
${metricLabel}: ${d3.format(".1%")(d[sgTrendMetric])}
GP: ${d.GP}`); }) .on("mousemove", (event) => { sgRadarTooltip.style("left", (event.pageX + 12) + "px").style("top", (event.pageY - 12) + "px"); }) .on("mouseout", () => { paths.attr("stroke-opacity", 0.9); sgRadarTooltip.style("opacity", 0); }); const legend = svg.append("g").attr("font-size", 11); byPlayer.forEach(([player], i) => { const ly = margin.top + i * 16; legend.append("circle").attr("cx", width - margin.right - 90).attr("cy", ly).attr("r", 4).attr("fill", sgRadarColor(player)); legend.append("text").attr("x", width - margin.right - 80).attr("y", ly + 4).text(player); }); return svg.node(); }4a:T59d,## Where the four actually stop looking alike Career value metrics tell a deceptively flat story: WS/48, BPM and VORP place all four players in a narrow band at the very top of the SG cohort, 89.6 to 99.8 percent across the board. That is not really a sign the four are similar players, it is a selection effect. None of them is a replacement-level player padding the sample, so naturally all four sit near the ceiling on metrics built to separate stars from everyone else. Shooting numbers tell the real story. LeBron James ranks at just the 33rd percentile on free-throw shooting, worse than two-thirds of everyone ever drafted at his position, right next to a 99.8 percentile WS/48: one of the most valuable players in the cohort's history, and a below-average free-throw shooter, at once. Curry's three-point percentage runs the opposite play: at the 96th percentile, it's the single sharpest outlier in the entire six-metric comparison. Two numbers need a caveat before being taken at face value, though. Booker's WS/48 of 0.110 barely clears the NBA-wide average of roughly 0.100, yet still lands in the 90th percentile of the draft cohort, "average NBA player" and "average draftee" are very different bars. And LeBron's VORP of 159.4 dwarfs Booker's 20.2 largely because he has played twice as many seasons, not because the talent gap is twice as large; rate stats like WS/48 control for that, cumulative ones like VORP do not.4b:Teb3,sgRadarChart = { const width = 520, height = 520, margin = 80; const radius = Math.min(width, height) / 2 - margin; const angleSlice = (2 * Math.PI) / sgRadarMetrics.length; const rScale = d3.scaleLinear().domain([0, 100]).range([0, radius]); const angle = i => angleSlice * i - Math.PI / 2; const toXY = (value, i) => [rScale(value) * Math.cos(angle(i)), rScale(value) * Math.sin(angle(i))]; const lineGen = d3.line().curve(d3.curveLinearClosed); const svg = d3.create("svg") .attr("viewBox", [0, 0, width, height]) .attr("width", width) .attr("height", height); const g = svg.append("g").attr("transform", `translate(${width/2},${height/2})`); // Perzentil-Gitter const gridLevels = [20, 40, 60, 80, 100]; g.selectAll(".sgRadarGrid") .data(gridLevels) .join("circle") .attr("class", "sgRadarGrid") .attr("r", d => rScale(d)) .attr("fill", "none") .attr("stroke", "#ccc") .attr("stroke-dasharray", "2,2"); g.selectAll(".sgRadarGridLabel") .data(gridLevels) .join("text") .attr("x", 4) .attr("y", d => -rScale(d)) .attr("font-size", 9) .attr("fill", "#999") .text(d => `${d}.`); // Achsen const axisG = g.selectAll(".sgRadarAxis") .data(sgRadarMetrics) .join("g"); axisG.append("line") .attr("x1", 0).attr("y1", 0) .attr("x2", (d, i) => toXY(100, i)[0]) .attr("y2", (d, i) => toXY(100, i)[1]) .attr("stroke", "#999"); axisG.append("text") .attr("x", (d, i) => toXY(114, i)[0]) .attr("y", (d, i) => toXY(114, i)[1]) .attr("text-anchor", "middle") .attr("font-size", 12) .attr("font-weight", 600) .text(d => d.label); // Spieler-Polygone (nutzt jetzt die aktive Datenauswahl: Karriere oder Saison) const polygonData = sgRadarActiveData.map(row => ({ player: row.player, points: sgRadarMetrics.map(m => ({ metric: m.label, raw: row[`${m.key}_raw`], pctile: row[`${m.key}_pctile`] })) })); g.selectAll(".sgRadarArea") .data(polygonData) .join("path") .attr("class", "sgRadarArea") .attr("d", d => lineGen(d.points.map((p, i) => toXY(p.pctile, i)))) .attr("fill", d => sgRadarColor(d.player)) .attr("fill-opacity", 0.15) .attr("stroke", d => sgRadarColor(d.player)) .attr("stroke-width", 2); // Punkte + Tooltip (flach, ein join statt verschachtelter Loops) const flatDots = polygonData.flatMap(d => d.points.map((p, i) => ({ player: d.player, ...p, xy: toXY(p.pctile, i) })) ); g.selectAll(".sgRadarDot") .data(flatDots) .join("circle") .attr("class", "sgRadarDot") .attr("cx", d => d.xy[0]) .attr("cy", d => d.xy[1]) .attr("r", 4) .attr("fill", d => sgRadarColor(d.player)) .on("mouseover", (event, d) => { sgRadarTooltip.style("opacity", 1) .html(`${d.player}
${d.metric}: ${d.raw} (${d.pctile}. Perzentil)`); }) .on("mousemove", (event) => { sgRadarTooltip.style("left", (event.pageX + 12) + "px") .style("top", (event.pageY - 12) + "px"); }) .on("mouseout", () => sgRadarTooltip.style("opacity", 0)); // Legende const legend = svg.append("g") .attr("transform", `translate(20, ${height - 20 - sgRadarActiveData.length * 16})`); legend.selectAll(".sgRadarLegendItem") .data(sgRadarActiveData.map(d => d.player)) .join("g") .attr("class", "sgRadarLegendItem") .attr("transform", (d, i) => `translate(0, ${i * 16})`) .call(gEl => { gEl.append("rect") .attr("width", 10).attr("height", 10) .attr("fill", d => sgRadarColor(d)); gEl.append("text") .attr("x", 14).attr("y", 9) .attr("font-size", 11) .text(d => d); }); return svg.node(); }4c:Te52,await FileAttachment("all_play_types_23_24.csv").csv({typed: true}).then(data => { // 1. Die Heatmap generieren (Breite auf 550 reduziert für mehr Platz) const heatmap = Plot.plot({ title: "The Tactical Fingerprint (Heatmap)", subtitle: "Frequency of specific play types per player (%)", width: 550, height: 500, marginLeft: 120, padding: 0.05, x: { domain: ["James Harden", "Devin Booker", "Stephen Curry", "LeBron James"] }, color: { type: "linear", scheme: "Blues", legend: true, domain: [0, 35] }, marks: [ Plot.cell(data, { x: "PLAYER", y: "PLAY_TYPE", fill: "FREQUENCY_PCT", inset: 0.5 }), Plot.text(data, { x: "PLAYER", y: "PLAY_TYPE", text: d => d.FREQUENCY_PCT > 0 ? (d.FREQUENCY_PCT + "%") : "", fill: d => d.FREQUENCY_PCT > 15 ? "white" : "black" }) ] }); // 2. Das HTML-Layout (STRICKT NEBENEINANDER ERZWUNGEN) return html`

${heatmap}

📖 Tactical Glossary

Cut: The Backdoor Slash. The player darts toward the basket without the ball, catching a sudden pass for a quick layup or dunk.
Handoff: The Dribble Handoff (DHO). A player hands the ball directly to a teammate running tightly past them.
Isolation (Iso): 1-on-1. The player holds the ball and attacks their direct defender while teammates clear the space.
Misc (Miscellaneous): The Outliers. Desperation heaves at the end of the shot clock, or broken plays.
OffRebound: Second-Chance Points. The player scores on a putback right after grabbing an offensive rebound.
OffScreen: Marathon Mode. The player runs off a set screen without the ball, catches the pass, and usually shoots immediately.
Postup: Back to the Basket. The player establishes position down low, using their physical strength to beat the defender.
PRBallHandler (Pick & Roll): The Playmaker Move. The ball-handler uses a screen from a teammate to create space.
PRRollMan: The Finisher. The player who sets the screen in a pick-and-roll and then dives hard to the rim.
Spotup: Catch & Shoot. The player waits for the pass at the three-point line and either shoots directly or drives.
Transition: Fastbreak. Rapid offensive attacks before the opposing defense has time to set up.

`; })4d:T28a6,# Discussion ## The Multidimensional Role of a Shooting Guard The positional comparison shows that shooting guards occupy an intermediate physical position between point guards and forwards. However, the role cannot be defined by a single physical or athletic profile. Shooting guards must combine several abilities, including shooting, ball handling, off-ball movement, playmaking, agility and defensive versatility. The importance of each ability also depends on the player's tactical role and the structure of the team. The four selected players illustrate this diversity. James Harden creates advantages through isolation scoring, physical strength and playmaking. Devin Booker combines agility, footwork and technically advanced shot creation. Stephen Curry influences the game through shooting range, rapid release and continuous off-ball movement, while LeBron James uses exceptional size, strength and court vision to control the offense. Their different tactical fingerprints demonstrate that successful perimeter players can dominate through fundamentally different combinations of skills. This multidimensionality helps explain why shooting guards do not show one clearly dominant Combine profile. A player can compensate for average results in one area through exceptional abilities in another. Booker, for example, combined elite agility with less impressive jumping and strength measurements, while Curry overcame a comparatively light physical profile through shooting, movement and spatial awareness. Evaluating shooting guards therefore requires considering how different abilities interact rather than searching for one ideal physical profile. ## Why Combine Results Do Not Tell the Whole Story The Combine remains valuable because it provides standardized and comparable measurements of size and athletic ability. The positional heatmaps reveal clear differences between guards, forwards and centers, particularly in height, weight, wingspan, reach, speed and agility. These measurements can help teams evaluate whether a prospect possesses the physical characteristics required for a particular role. However, the connection between these measurements and later career performance is weak. The regression models based on height, weight, wingspan and standing reach explain only **1.6% of the variation in WS/48**, **0.3% in BPM**, and **2.6% in VORP per season**. All four highlighted players achieved better career outcomes than predicted by their physical profiles. This suggests that the characteristics separating exceptional NBA players from other prospects are largely located outside these four measurements. The regression results should nevertheless be interpreted carefully. Only four predictors were included because they were the only Combine measurements available for all four featured players. LeBron James did not complete several athletic tests, so vertical jump, sprint, lane agility, shuttle run and bench press could not be included without excluding him or artificially replacing missing values. The results therefore demonstrate the limited predictive value of the **shared physical measurements**, rather than proving that every component of the NBA Draft Combine is irrelevant. Combine measurements also represent a prospect at one specific moment before entering the NBA. They cannot account for how players develop their technique, adapt to new roles or improve within a professional environment. Booker's gradual development into an efficient scorer and playmaker demonstrates that career performance is dynamic, while Combine measurements are largely static. Team quality, coaching, playing time, injuries and tactical responsibilities can further influence a player's development and career statistics. ## What the NBA Draft Combine Does Not Measure The project began with a straightforward question: can two days of testing in Chicago predict who will become a productive NBA player years later? Evidence from previous validity studies suggests that the answer is largely no. Teramoto et al. (2018) found that only anthropometric measurements such as height, wingspan and standing reach showed moderate correlations with future performance (*r* = 0.313–0.545), while agility and jumping tests showed almost none. García-Rubio et al. (2020), analysing 723 drafted players from 1999 to 2018, found that Combine results predicted draft position only for hand width, height among frontcourt players and vertical jump. No other tested attribute reliably predicted draft position or subsequent performance. This is primarily a limitation of measurement scope rather than simply a data-quality problem. A standardized physical assessment cannot capture every factor that determines the development of an NBA career: - **Shooting under pressure:** The ability to maintain accuracy against active defenders and during decisive moments. NBA players have been shown to perform worse on free throws during the final seconds of close games, a situation that standardized shooting drills cannot reproduce (Cao, Price & Stone, 2011). - **Basketball intelligence:** The ability to read defensive structures, anticipate plays and make appropriate decisions. Cognitive testing among 356 NBA prospects found that intelligence scores predicted shooting efficiency and Player Efficiency Rating even after controlling for draft position (Hogan et al., 2023). - **Off-ball movement:** The use of positioning, timing and movement without possession to create space for both the player and his teammates. - **Endurance and conditioning:** The ability to maintain movement, concentration and technical execution throughout a game or season. During live competition, players can sustain heart rates above 85% of their maximum for extended periods, a demand that short Combine drills do not reproduce (Stojanović et al., 2018). - **Mental strength and composure:** The ability to respond to pressure, mistakes, criticism, injuries and changing responsibilities. - **Communication and leadership:** The ability to organize teammates, communicate during play and contribute to collective decision-making. Team cohesion has been identified as an important contributor to performance across 46 studies (Carron et al., 2002). - **Work ethic and adaptability:** The capacity to develop new skills and adjust to different coaches, teammates, roles and tactical systems over time. - **Creativity and technical skill:** The ability to create scoring opportunities through footwork, ball handling, timing, changes of pace and unconventional decision-making. - **Durability:** The capacity to remain healthy and productive throughout a professional career. Combine athletic results did not differ meaningfully between players who later required lower-limb surgery and those who did not (Patel et al., 2025). Comparable NFL research found that documented injury history, rather than athletic testing, predicted career length (Brophy et al., 2008). - **Team and opportunity context:** Access to playing time, appropriate coaching, roster stability and a tactical role that supports the player's development. Some of these qualities may be observed through interviews, scrimmages, private workouts and game film, but they are difficult to represent through standardized numerical measurements. They are also absent from the quantitative dataset used in this project. Curry's defensive gravity, Harden's change-of-pace playmaking, Booker's technical development and James's decision-making illustrate abilities that cannot be adequately captured by physical measurements alone. The same limitation is described explicitly in the more extensively studied NFL Combine literature. Kuzmits and Adams (2008) characterize the Combine as testing isolated, uncontested and single-effort athleticism, whereas professional performance requires repeated execution under fatigue, opposition pressure and real competitive consequences. This represents a mismatch between the test and the actual demands of the job rather than a simple measurement error. The Combine measures size and explosiveness effectively, but many qualities with potential long-term relevance remain outside its quantitative scope. ## Limitations The analysis is affected by missing Combine values because not every prospect completes every assessment. Highly rated prospects may selectively skip individual drills or the entire event, meaning that the missing observations are unlikely to be completely random. The regression sample also includes only drafted players with sufficient career data, which may introduce selection bias. Position labels provide another limitation because modern NBA roles are flexible. Curry is listed as a point guard–shooting guard and James as a small forward–shooting guard, while both have performed responsibilities that extend beyond traditional positional definitions. The selected players are illustrative cases and should not be considered representative of all NBA prospects. Future analyses could include shooting-drill results, college performance, age at the time of the draft, draft position, injuries and measures of opportunity such as career minutes. Predictive models should also be evaluated using players who were not involved in estimating the models to determine how well they perform on previously unseen cases. # Conclusion The NBA Draft Combine provides useful standardized information about the physical and athletic characteristics of prospects. However, the analysis shows that successful shooting guards and related perimeter players can emerge from substantially different Combine profiles. The regression models and individual player comparisons indicate that physical measurements alone cannot reliably predict long-term NBA performance. Harden, Booker, Curry and James all exceeded the career outcomes expected from their shared physical measurements, but achieved this success through different playing styles and developmental paths. The Combine should therefore be treated as one component of a broader evaluation process. Physical measurements should be considered alongside game statistics, shooting ability, tactical analysis, game film and qualitative assessments of decision-making, mental strength, communication, durability and adaptability. NBA success is multidimensional and cannot be reduced to a small set of physical attributes. 4e:[{"id":"0d75f3767800a4191a7e24f2a94db14c1b617568297af9a7a770cda7a796a158da77bcd837f7d11e00f2f08cd13b113f3c624c9f4c8fe70d909011c5f12aa638","href":"https://static.observableusercontent.com/files/0d75f3767800a4191a7e24f2a94db14c1b617568297af9a7a770cda7a796a158da77bcd837f7d11e00f2f08cd13b113f3c624c9f4c8fe70d909011c5f12aa638","name":"Draft Combine - Kaggle.csv","type":"text/csv","lastModified":1788250605192.161,"size":204626},{"id":"5b7198846520f4ece190d32a3745636398bfead636ca2ef5c45bba9d63b10ebfda9cc9be4aa91b69d0317588eb0c1f5de678c6a49bc117c895367da4781a4eee","href":"https://static.observableusercontent.com/files/5b7198846520f4ece190d32a3745636398bfead636ca2ef5c45bba9d63b10ebfda9cc9be4aa91b69d0317588eb0c1f5de678c6a49bc117c895367da4781a4eee","name":"combine_with_outcomes.csv","type":"text/csv","lastModified":1788338293895,"size":371511},{"id":"77f29ae63af0f9c1b69954a6ce6a1896884b081e6a854920065539227378c2c1dfd3ffeb39a5e8574ab7e0f61fe4964157a5fc3f0244cf203682ab2d72ddba45","href":"https://static.observableusercontent.com/files/77f29ae63af0f9c1b69954a6ce6a1896884b081e6a854920065539227378c2c1dfd3ffeb39a5e8574ab7e0f61fe4964157a5fc3f0244cf203682ab2d72ddba45","name":"merge_combine_with_outcomes_3.py","type":"text/x-python","lastModified":1788338176801,"size":18956},{"id":"3726e111899a2f08a0b2d15f1ba80c8fa192132d1285659277a93d5b56bde21c9e8f738d2c01af3cf71f2a3ede1b1cdc84d20467d9ee893eaea8cf8598199628","href":"https://static.observableusercontent.com/files/3726e111899a2f08a0b2d15f1ba80c8fa192132d1285659277a93d5b56bde21c9e8f738d2c01af3cf71f2a3ede1b1cdc84d20467d9ee893eaea8cf8598199628","name":"1_combine_with_outcomes_with_defense.csv","type":"text/csv","lastModified":1788353373354,"size":316912},{"id":"07731d62f942250afb96ef02da03e84cbacce4df450cceda327e8215c3b8ff60929c66e84d43f7d43c3713016a2a6eac93d0c227ca923ee1943f39c2825d68e1","href":"https://static.observableusercontent.com/files/07731d62f942250afb96ef02da03e84cbacce4df450cceda327e8215c3b8ff60929c66e84d43f7d43c3713016a2a6eac93d0c227ca923ee1943f39c2825d68e1","name":"2_combine_with_outcomes_with_defense.csv","type":"text/csv","lastModified":1788370600389,"size":316032},{"id":"18219b8ad59caa51c3f69d8e1071b0c9eaf5f7e7f9c5e77cdd117a40e59e478e6d35bc7a08e27a001e80ec39bdf84ea12225b0711275f647a3582bd4a77c17a2","href":"https://static.observableusercontent.com/files/18219b8ad59caa51c3f69d8e1071b0c9eaf5f7e7f9c5e77cdd117a40e59e478e6d35bc7a08e27a001e80ec39bdf84ea12225b0711275f647a3582bd4a77c17a2","name":"FGA_Lebron_23_24.png","type":"image/png","lastModified":1788376237309,"size":111647},{"id":"55fcb2ea668202f46c41b235510a6f380b41c2990a5adaa38e4cc33488cc9f1b158c30d1876e857ded904382328e6e24169ddca6d5c4f3ae226f480a553bf345","href":"https://static.observableusercontent.com/files/55fcb2ea668202f46c41b235510a6f380b41c2990a5adaa38e4cc33488cc9f1b158c30d1876e857ded904382328e6e24169ddca6d5c4f3ae226f480a553bf345","name":"FGA_Curry_23_24.png","type":"image/png","lastModified":1788376215546,"size":166043},{"id":"769913ce8a847fb802eca7a71dc34a82c491afd4749051174193226c0fbc3ba0be43a82ad9c3eef16a0b09aa78fa59ae39c779a796b1881188caeeb7ac978a34","href":"https://static.observableusercontent.com/files/769913ce8a847fb802eca7a71dc34a82c491afd4749051174193226c0fbc3ba0be43a82ad9c3eef16a0b09aa78fa59ae39c779a796b1881188caeeb7ac978a34","name":"FGA_Harden_23_24.png","type":"image/png","lastModified":1788376199807,"size":153729},{"id":"1ebe6fc4273d6da10dade440d67f5b5bc940c57627d9beb20b768adfa1c57d11e70d00e4b8ab3ac72eb28b381be1f8378f98302f06021c70895018e8c37ce023","href":"https://static.observableusercontent.com/files/1ebe6fc4273d6da10dade440d67f5b5bc940c57627d9beb20b768adfa1c57d11e70d00e4b8ab3ac72eb28b381be1f8378f98302f06021c70895018e8c37ce023","name":"FGA_Booker_23_24.png","type":"image/png","lastModified":1788376102290,"size":142720},{"id":"0f0bf29ff83ef668bb67678642a061088cb3e8ba0a4555116391b491e0aa8cb73e8938094352eee639cc018e09cce849f0088e868834f33ea8ad68229469592a","href":"https://static.observableusercontent.com/files/0f0bf29ff83ef668bb67678642a061088cb3e8ba0a4555116391b491e0aa8cb73e8938094352eee639cc018e09cce849f0088e868834f33ea8ad68229469592a","name":"booker_jumpshot.mp4","type":"video/mp4","lastModified":1788379566321,"size":5071885},{"id":"a362b5b2a8e66d7f36aa7889acda44f4aaf394eefef363b26306b581625b04f322eeb73cd14329f3704a92607b84ab4de2cf5e2917dd29b3f338ae59cf02a315","href":"https://static.observableusercontent.com/files/a362b5b2a8e66d7f36aa7889acda44f4aaf394eefef363b26306b581625b04f322eeb73cd14329f3704a92607b84ab4de2cf5e2917dd29b3f338ae59cf02a315","name":"curry_jumpshot.mp4","type":"video/mp4","lastModified":1788379349454,"size":8932478},{"id":"2eb4be85e216a3f0a4033c8d3ac2472149b655af1d1690cb415e174b4efb8d41a376152e9f2c776504b530976cafadc8d3f9e0f1ef5c9a12216e9c482a0335f4","href":"https://static.observableusercontent.com/files/2eb4be85e216a3f0a4033c8d3ac2472149b655af1d1690cb415e174b4efb8d41a376152e9f2c776504b530976cafadc8d3f9e0f1ef5c9a12216e9c482a0335f4","name":"harden_stepback.mp4","type":"video/mp4","lastModified":1788378569857,"size":3090238},{"id":"6c1abe77180d41903ef7aef90a18490c3a0b808b23c388bfcc4d455836bd5f44bd217ff2f39b98f091e7047a45020519e65151a728d8d651b7ef683f4a8e8e07","href":"https://static.observableusercontent.com/files/6c1abe77180d41903ef7aef90a18490c3a0b808b23c388bfcc4d455836bd5f44bd217ff2f39b98f091e7047a45020519e65151a728d8d651b7ef683f4a8e8e07","name":"lebron_layup.mp4","type":"video/mp4","lastModified":1788379897011,"size":6082226},{"id":"1f3d9e53f51995bfe035dde3be27e40013637c8f91c311161c3fbc916593bcec331996c2b9cf22341691dc31f330c1a16badba8c2312f5882c170fa326c0192a","href":"https://static.observableusercontent.com/files/1f3d9e53f51995bfe035dde3be27e40013637c8f91c311161c3fbc916593bcec331996c2b9cf22341691dc31f330c1a16badba8c2312f5882c170fa326c0192a","name":"play_types_23_24.csv","type":"text/csv","lastModified":1788416900967,"size":471},{"id":"e1404dafc0ccbf4e167846076c4623b52db18ba5c1e904f5b8861cf7174beab2aff9ceb17dc73fb3aa26891df175676f9e836ebe8bbb6d439b28e1092ffdd86e","href":"https://static.observableusercontent.com/files/e1404dafc0ccbf4e167846076c4623b52db18ba5c1e904f5b8861cf7174beab2aff9ceb17dc73fb3aa26891df175676f9e836ebe8bbb6d439b28e1092ffdd86e","name":"all_play_types_23_24.csv","type":"text/csv","lastModified":1788435327876,"size":1190},{"id":"bff4f2511c2b828f8df9fbc884f25c0fbb3491a53f93c88a383bce69aaf83d11766433cc59e6c693ae3d6b552dd57bf1cd2a95d619065f6ccbac994b52d5a860","href":"https://static.observableusercontent.com/files/bff4f2511c2b828f8df9fbc884f25c0fbb3491a53f93c88a383bce69aaf83d11766433cc59e6c693ae3d6b552dd57bf1cd2a95d619065f6ccbac994b52d5a860","name":"lebron-james.png","type":"image/png","lastModified":1788436455929,"size":214782},{"id":"dcc3258adb185c09289ad61961002fce505c8c9d7aaf57ac4b5ea88f0f4fbeb43c92f56420251c6d117ef4dffdc6609534fc953cdbf4374cb9f85295adf16ae7","href":"https://static.observableusercontent.com/files/dcc3258adb185c09289ad61961002fce505c8c9d7aaf57ac4b5ea88f0f4fbeb43c92f56420251c6d117ef4dffdc6609534fc953cdbf4374cb9f85295adf16ae7","name":"stephen-curry.png","type":"image/png","lastModified":1788436455496,"size":187656},{"id":"b73ede8f2e39b115117b6a98dbb4998a4b998a6361081082537129d7c55ee0a6566dbf5ecfd64b794522b11320a40e8d595f310d41f9116bef6d7b21f59a4a67","href":"https://static.observableusercontent.com/files/b73ede8f2e39b115117b6a98dbb4998a4b998a6361081082537129d7c55ee0a6566dbf5ecfd64b794522b11320a40e8d595f310d41f9116bef6d7b21f59a4a67","name":"devin-booker.png","type":"image/png","lastModified":1788436454562,"size":194740},{"id":"37c43cd033d0a7c944f00bec1a96345ffb8237dba3af77e66eaf28db1059a2b08af0d28c0649f2912418b56d368922da192ec845d224f8d0d75405b37f03bcb2","href":"https://static.observableusercontent.com/files/37c43cd033d0a7c944f00bec1a96345ffb8237dba3af77e66eaf28db1059a2b08af0d28c0649f2912418b56d368922da192ec845d224f8d0d75405b37f03bcb2","name":"james-harden.png","type":"image/png","lastModified":1788436414214,"size":227727},{"id":"4c90dbffac3bd6c0ec58015bdcded15d17d405d39aefb6ef8e95df57fcce981fb5814c92700ceef7bd18c6d3e82860c9775b663af97c2f942bfaaf0741034d35","href":"https://static.observableusercontent.com/files/4c90dbffac3bd6c0ec58015bdcded15d17d405d39aefb6ef8e95df57fcce981fb5814c92700ceef7bd18c6d3e82860c9775b663af97c2f942bfaaf0741034d35","name":"player_tracking_23_24.csv","type":"text/csv","lastModified":1788445539630,"size":455},{"id":"159f664d59a9c9ea0ef93ad8693120afc14f88d8f000588d490e2b8da2d86b91901295d1c62ee360f1e66ed4e9fa7c37b85d70fdf26efbdb36374c84056312fd","href":"https://static.observableusercontent.com/files/159f664d59a9c9ea0ef93ad8693120afc14f88d8f000588d490e2b8da2d86b91901295d1c62ee360f1e66ed4e9fa7c37b85d70fdf26efbdb36374c84056312fd","name":"sg_radar_data.csv","type":"text/csv","lastModified":1788455720568,"size":426},{"id":"4580ed56cfa2a1f4c3d3e67d236e4e79fffea38ede6502f2907a85c1d071623c24ab3cbaa11b7db8ccf2489cb38bfe751a89d519473f281c286f4e8c566d82bc","href":"https://static.observableusercontent.com/files/4580ed56cfa2a1f4c3d3e67d236e4e79fffea38ede6502f2907a85c1d071623c24ab3cbaa11b7db8ccf2489cb38bfe751a89d519473f281c286f4e8c566d82bc","name":"shooting_stats_4players.csv","type":"text/csv","lastModified":1788461698494,"size":6919},{"id":"c7e9d29364cbae96a75d5296614d2744316363630f107dd22fec86368b46054a750f8ea32b548015f2eedaa4ae2179e8f5acf6847372c1823d9a25e0b28077eb","href":"https://static.observableusercontent.com/files/c7e9d29364cbae96a75d5296614d2744316363630f107dd22fec86368b46054a750f8ea32b548015f2eedaa4ae2179e8f5acf6847372c1823d9a25e0b28077eb","name":"sg_radar_data_by_season.csv","type":"text/csv","lastModified":1788462125290,"size":5985},{"id":"8d406c4bffc20514940155bbb1d6118525a0953f9da1c3277b29f80171d1291de7f849254d9ee698174545565c6ff25b8661b88f64317706d118517e2808fe9b","href":"https://static.observableusercontent.com/files/8d406c4bffc20514940155bbb1d6118525a0953f9da1c3277b29f80171d1291de7f849254d9ee698174545565c6ff25b8661b88f64317706d118517e2808fe9b","name":"NBA_Draft.jpg","type":"image/jpeg","lastModified":1788522234355,"size":299920}] 30:{"id":"edcbf8535d0584e9","slug":"from-the-nba-draft-combine-to-draft-night","title":"From the NBA Draft Combine to NBA Career Performance","thumbnail":null,"access_level":"unlisted","create_time":"$D2026-09-01T08:07:10.343Z","update_time":"$D2026-09-04T13:04:44.915Z","view_time":null,"first_public_time":"$D2026-09-02T08:57:14.117Z","fork":{"id":"ee9e4fb45c009de1","slug":"search-table","title":"Search + Table","owner":{"id":"f35c755083683fe5","login":"observablehq","name":"Observable"}},"stars":4,"starred":false,"owner":{"type":"team","id":"c1751318e65a435a","login":"nba-draft-combine-groupwork","name":"NBA Draft Combine GroupWork","avatar_url":"https://avatars-next.observableusercontent.com/5058f1e63fd906af9a13b3fe87a33b4bd4710a5ba437df6d151a408d773cc50b","tier":"pro_2024"},"version":1483,"publish_version":null,"role":null,"authored":false,"authors":[{"type":"individual","id":"244d1fd997a9455a","login":"michael-nguyen","name":"Michael Nguyen","avatar_url":null,"tier":null},{"type":"individual","id":"0aa59571979f67b2","login":"gregor-wyss","name":"Gregor.wyss","avatar_url":"https://avatars-next.observableusercontent.com/def0cb94b8edbb0441620858c5b10c1551c95daece445a436a41637193172910","tier":null},{"type":"individual","id":"e2f1ff37d7e67475","login":"joel-rieser","name":"Joel rieser","avatar_url":"https://avatars-next.observableusercontent.com/c95af7b7cf856538dd13e4f2b9a8cc4d87931e0d8d6be6a2f80d76cec65a93aa","tier":null},{"type":"individual","id":"1cde8b727d4eeeb5","login":"faebu","name":"faebu","avatar_url":"https://avatars-next.observableusercontent.com/51a3427b54dfb4a089987b06892b2555463bbf6b6b4f0cbc2be99a66255b1a4f","tier":null}],"members":[{"type":"individual","id":"244d1fd997a9455a","login":"michael-nguyen","name":"Michael Nguyen","avatar_url":null,"tier":null,"role":"editor"},{"type":"team","id":"c1751318e65a435a","login":"nba-draft-combine-groupwork","name":"NBA Draft Combine GroupWork","avatar_url":"https://avatars-next.observableusercontent.com/5058f1e63fd906af9a13b3fe87a33b4bd4710a5ba437df6d151a408d773cc50b","tier":"pro_2024","role":"editor"}],"comments":[{"id":"44f4de712ed3e8eb","reply_to_id":null,"content":"1. Welche körperlichen und athletischen Merkmale des NBA Draft Combine hängen mit der Wahrscheinlichkeit zusammen, dass ein Spieler gedraftet wird, und unterscheiden sich diese Zusammenhänge je nach Spielposition?\r\n2. Inwieweit hängen die Ergebnisse des NBA Draft Combine damit zusammen, wie früh ein Spieler im NBA Draft ausgewählt wird?\r\n3. Wiesen später erfolgreiche NBA-Spieler bereits überdurchschnittliche Ergebnisse beim NBA Draft Combine auf, und welche erfolgreichen Spieler wurden durch die gemessenen Merkmale möglicherweise unterschätzt?","create_time":"$D2026-09-02T07:15:23.222Z","update_time":null,"resolved":false,"user":{"type":"individual","id":"244d1fd997a9455a","login":"michael-nguyen","name":"Michael Nguyen","avatar_url":null,"tier":null}},{"id":"ce1b761a1ebd9c3b","reply_to_id":null,"content":"- Einleitung -> Fabian\r\n- Data prep. und cleaning --> Joel\r\n- Grundwerte (Michael)\r\n- Weshalb SG (Michael)\r\n- weshalb diese 4 spieler (Michael)\r\n- Indiv. Spieleranalyse (Michael)\r\n-- Pro Spieler eine kleine Biografie (3. sätze) + spezielle infos + öffentliche Medien + wo liege seine stärken (was andere über ihn reden zu seinem spielstil) --> Alle\r\n-- Erfolge (MVP, Titel usw.) -> Fabian\r\n-- was sie verdienen --> Fabian (zusammen mit Erfolge)\r\n-- Wurfkategorien Vergleich -> Gregor\r\n-- Schuss Vergleich --> Gregor\r\n-- wurfstatistik über karriere -> Joel\r\n-- Optional defense \r\n- Zusammenfassung Tabelle wo welche Spieler startk bzw. schwach sind -> Morgen\r\n- aufzeigen, dass im sg viele verschiedene werte wichtig sind und dass combine werte nicht alles bedeuten ->Fabian\r\n- welche daten fehlen bei den combine bzw. ausdauer, mentale stärke, provokation, kommunikation. -> Joel","create_time":"$D2026-09-03T09:08:55.168Z","update_time":null,"resolved":false,"user":{"type":"individual","id":"244d1fd997a9455a","login":"michael-nguyen","name":"Michael Nguyen","avatar_url":null,"tier":null}}],"custom_thumbnail":false,"trashed":false,"subscribed":null,"forks":0,"collection":null,"type":"notebook","body":{"title":"From the NBA Draft Combine to NBA Career Performance","cells":[{"id":17,"value":"# From the NBA Draft Combine to NBA Career Performance\n\nThe path from a promising young basketball player to an NBA draft pick runs through a rigorous, multi-stage evaluation process, at the center of which stands the NBA Draft Combine — two days of standardized physical and athletic testing that shape how teams and media assess prospects. This report traces that process from the difficulty of reaching the NBA, through the mechanics of the Combine and the draft, to a data-driven test of how well Combine measurements predict NBA career performance. The analysis finds that they explain only a small share of it.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":25,"value":"$32","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":69,"value":"## From Prospect to Draft Pick\nBefore a player ever hears their name called on draft night, they typically pass through several distinct stages that shape how NBA teams evaluate them. Most prospects first build a track record in high school, college basketball, or an international professional league, where their statistics, physical development, and game film are closely tracked by scouts. As draft eligibility approaches, players declare for the draft and enter a pre-draft evaluation period that includes private workouts, interviews, and the NBA Draft Combine. This period is where raw potential gets converted into quantifiable data, allowing teams to compare players across different leagues and competition levels on a more standardized basis. It is worth noting, however, that participating in the Combine is not a requirement for being selected on draft night — several top prospects have skipped it entirely and still gone on to become franchise cornerstones.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":46,"value":"### The NBA Draft Process\nThe NBA Draft consists of two rounds of 30 picks each, for a total of 60 selections, typically held in late June. The order of the first 14 picks is determined by the NBA Draft Lottery, a weighted random drawing among non-playoff teams designed to discourage intentional losing, while the remaining picks follow reverse order of regular-season record. Draft rights can be traded between teams independently of the players themselves, meaning picks are often exchanged years in advance as part of broader trade packages. Once selected, a player signs a rookie-scale contract with the drafting team, which holds exclusive negotiating rights to that player for a set period, even if the player was drafted while still playing overseas.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":48,"value":"### The Purpose of the NBA Draft Combine\nThe NBA Draft Combine exists to give all 30 teams standardized, comparable information on draft prospects beyond what can be observed from game film alone. Because prospects come from vastly different leagues, competition levels, and playing systems, raw statistics are often difficult to compare directly, making objective physical and medical data especially valuable. The Combine allows teams to verify a player's true height, wingspan, and athletic testing results, which can differ meaningfully from the numbers listed by colleges or media outlets. It also gives front offices direct access to medical staff and interview time with prospects, helping them assess durability, character, and basketball intelligence alongside physical tools. Ultimately, the Combine functions as a risk-reduction tool, helping teams refine their draft boards before committing significant guaranteed money to a rookie contract.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":333,"value":"FileAttachment(\"NBA_Draft.jpg\").image({\n width: 650,\n alt: \"NBA Draft Combine\"\n})","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":60,"value":"### What the Combine Measures\nThe Combine is organized around several distinct categories of testing that together build a comprehensive athletic profile of each prospect. Anthropometric measurements capture body dimensions such as height, wingspan, standing reach, weight, and body fat percentage. Strength and agility drills assess explosiveness and lateral quickness through tests like the bench press, lane agility drill, shuttle run, and vertical jump. Shooting drills and controlled scrimmages give scouts a chance to observe shooting mechanics and basketball skills in a live, if limited, competitive setting. Finally, medical evaluations and formal interviews round out the picture, allowing teams to assess injury risk and gather qualitative insight into a player's mindset and decision-making that no physical test can capture.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":55,"value":"# Research Question\n\nHow well do NBA Draft Combine measurements predict the career performance of shooting guards?","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":58,"value":"# Data and Study Scope\nThe analysis combines standardized NBA Draft Combine measurements with historical draft results and NBA career statistics. Connecting these datasets makes it possible to follow a player’s journey from the Combine to draft night and, finally, to his professional NBA career.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":62,"value":"$33","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":78,"value":"$34","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":83,"value":"## Data Preparation\nThe analysis in this notebook rests on a merged dataset combining NBA Draft Combine testing results with players’ subsequent career outcomes. This section documents the data sources and the matching procedure used to link the two.\n\n\n### Data Sources\n\nCombine attributes (height, wingspan, vertical leap, sprint time, lane agility, shuttle run, bench press, and related measures) were sourced from the NBA Draft Combine dataset covering 1,749 player-year records from the 2000 through 2026 draft classes.\nCareer outcomes, including draft position (PICK), the performance metrics WS, WS/48, BPM, and VORP, and basic per-game statistics such as points, assists, rebounds, and shooting percentages, were scraped from NBA Draft results and career statistics from Basketball-Reference. WS, WS/48, BPM, and VORP are Basketball-Reference proprietary calculations and are not available through the official NBA stats API, which exposes only raw box-score totals.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":155,"value":"$35","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":5,"value":"rawData = FileAttachment(\"2_combine_with_outcomes_with_defense.csv\").csv({ typed: true })","pinned":false,"hidden":true,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":34,"value":"cleanData = rawData.map(row =>\n Object.fromEntries(\n Object.entries(row).map(([column, value]) => {\n if (\n value === null ||\n value === undefined ||\n value === \"\" ||\n value === \"NA\"\n ) {\n return [column, null];\n }\n\n if (typeof value === \"string\") {\n return [column, value.trim()];\n }\n\n return [column, value];\n })\n )\n)","pinned":false,"hidden":true,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":6,"value":"viewof search = Inputs.search(cleanData)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":7,"value":"Inputs.table(search, {\n format: {\n YEAR: d3.format(\"d\"),\n PICK: d3.format(\"d\")\n }\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":92,"value":"## From Combine Profiles to NBA Career Outcomes","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":347,"value":"Pearson correlation coefficients were computed between all 18 NBA Draft Combine attributes and three career-outcome metrics — WS/48, BPM, and cumulative VORP — for the subset of drafted players with complete outcome data (n ≈ 938). WS/48 was treated as the primary reference metric, since it is not confounded by career length in the way cumulative statistics are.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":340,"value":"const combineData = await FileAttachment(\"combine_with_outcomes.csv\").csv({typed: true});","pinned":false,"hidden":false,"mode":"js","author":"$undefined","summary":"$undefined"},{"id":341,"value":"const draftedPlayers = combineData.filter(d => d.MATCH_STATUS === \"matched_drafted\");","pinned":false,"hidden":false,"mode":"js","author":"$undefined","summary":"$undefined"},{"id":342,"value":"function pearsonCorrelation(rows, xKey, yKey) {\n // Parses BAR (stored as \"105.9%\") into a plain number; other fields are already numeric via {typed: true}\n const parseValue = (v) => typeof v === \"string\" ? parseFloat(v.replace(\"%\", \"\")) : v;\n\n const pairs = rows\n .map(d => [parseValue(d[xKey]), parseValue(d[yKey])])\n .filter(([x, y]) => x != null && y != null && !isNaN(x) && !isNaN(y));\n\n const n = pairs.length;\n if (n < 10) return {r: null, n}; // too few observations to trust\n\n const meanX = d3.mean(pairs, p => p[0]);\n const meanY = d3.mean(pairs, p => p[1]);\n\n let num = 0, denX = 0, denY = 0;\n for (const [x, y] of pairs) {\n const dx = x - meanX, dy = y - meanY;\n num += dx * dy;\n denX += dx * dx;\n denY += dy * dy;\n }\n\n const den = Math.sqrt(denX * denY);\n return {r: den === 0 ? null : num / den, n};\n}","pinned":false,"hidden":false,"mode":"js","author":"$undefined","summary":"$undefined"},{"id":343,"value":"const combineAttributes = [\n {key: \"HGT\", label: \"Height\"},\n {key: \"WGT\", label: \"Weight\"},\n {key: \"BMI\", label: \"BMI\"},\n {key: \"BF\", label: \"Body Fat %\"},\n {key: \"WNGSPN\", label: \"Wingspan\"},\n {key: \"STNDRCH\", label: \"Standing Reach\"},\n {key: \"HANDL\", label: \"Hand Length\"},\n {key: \"HANDW\", label: \"Hand Width\"},\n {key: \"STNDVERT\", label: \"Standing Vertical\"},\n {key: \"LPVERT\", label: \"Max Vertical\"},\n {key: \"LANE\", label: \"Lane Agility (lower = faster)\"},\n {key: \"SHUTTLE\", label: \"Shuttle Run (lower = faster)\"},\n {key: \"SPRINT\", label: \"Sprint (lower = faster)\"},\n {key: \"BENCH\", label: \"Bench Press\"},\n {key: \"BAR\", label: \"Wingspan/Height Ratio\"},\n {key: \"PAN\", label: \"Hand Length × Width\"},\n {key: \"PBHGT\", label: \"Standing Reach + Standing Vert.\"},\n {key: \"PDHGT\", label: \"Standing Reach + Max Vert.\"}\n];","pinned":false,"hidden":false,"mode":"js","author":"$undefined","summary":"$undefined"},{"id":344,"value":"const correlationResults = combineAttributes.map(attr => {\n const ws48 = pearsonCorrelation(draftedPlayers, attr.key, \"WS_PER_48\");\n const bpm = pearsonCorrelation(draftedPlayers, attr.key, \"BPM\");\n const vorp = pearsonCorrelation(draftedPlayers, attr.key, \"VORP\");\n return {\n attribute: attr.label,\n key: attr.key,\n r_ws48: ws48.r,\n n_ws48: ws48.n,\n r_bpm: bpm.r,\n r_vorp: vorp.r\n };\n}).filter(d => d.r_ws48 !== null)\n .sort((a, b) => Math.abs(b.r_ws48) - Math.abs(a.r_ws48));","pinned":false,"hidden":false,"mode":"js","author":"$undefined","summary":"$undefined"},{"id":345,"value":"Plot.plot({\n title: \"Combine attributes vs. WS/48 (career quality, length-adjusted)\",\n subtitle: `Pearson correlation, drafted players only (n ≈ ${correlationResults[0]?.n_ws48 ?? \"?\"})`,\n marginLeft: 200,\n width: 700,\n x: {label: \"Correlation coefficient (r)\", domain: [-0.3, 0.3]},\n y: {label: null},\n color: {domain: [-1, 1], scheme: \"RdBu\"},\n marks: [\n Plot.ruleX([0]),\n Plot.barX(correlationResults, {\n x: \"r_ws48\",\n y: \"attribute\",\n fill: \"r_ws48\",\n sort: {y: \"x\"},\n tip: true\n })\n ]\n})","pinned":false,"hidden":false,"mode":"js","author":"$undefined","summary":"$undefined"},{"id":346,"value":"Inputs.table(correlationResults.map(d => ({\n Attribute: d.attribute,\n \"r (WS/48)\": d.r_ws48?.toFixed(3),\n \"r (BPM)\": d.r_bpm?.toFixed(3),\n \"r (VORP, cumulative)\": d.r_vorp?.toFixed(3),\n \"n\": d.n_ws48\n})))","pinned":false,"hidden":false,"mode":"js","author":"$undefined","summary":"$undefined"},{"id":350,"value":"Three main takeaways emerge. First, even the strongest correlation — height, at r = 0.216 — explains less than 5% of the variance in WS/48, confirming that no individual Combine attribute meaningfully predicts career quality on its own. Second, a clear pattern separates the attributes: size-related measurements (height, wingspan, standing reach, weight, hand dimensions) cluster around r = 0.16–0.22, while athletic and agility measurements (vertical leap, sprint, shuttle run, lane agility, bench press) show almost no relationship, several near zero. Third, correlations weaken further for BPM and nearly vanish for cumulative VORP, consistent with the career-length confound discussed earlier in the report.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":96,"value":"### A positional comparison of physical measurements, athletic test results and subsequent NBA performance\nFollowing data preparation, the analysis first examines how NBA Draft Combine profiles differ across playing positions. The comparison includes all Combine participants and the subset of players who were subsequently drafted. Their later NBA career statistics are then compared to determine whether positional differences in physical and athletic profiles are also reflected in professional performance.\n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":166,"value":"$36","pinned":false,"hidden":true,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":167,"value":"const selectedPosition = view(\n Inputs.select(\n [\"All\", \"PG\", \"SG\", \"SF\", \"PF\", \"C\"],\n {\n label: \"Position\",\n value: \"All\"\n }\n )\n);","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":168,"value":"// Include all available Combine years\nconst useAllYears = view(\n Inputs.toggle({\n label: \"Use all years\",\n value: true\n })\n);\n\n// Select one NBA Combine year\nconst selectedYear = view(\n Inputs.range(\n [\n Math.min(...availableYears),\n Math.max(...availableYears)\n ],\n {\n label: \"Combine year\",\n value: Math.max(...availableYears),\n step: 1\n }\n )\n);","pinned":false,"hidden":true,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":185,"value":"// Use either all years or the selected year\nconst effectiveYear = useAllYears\n ? \"All\"\n : selectedYear;","pinned":false,"hidden":true,"mode":"js","author":"$undefined","summary":"$undefined"},{"id":170,"value":"const filteredPlayers = players.filter(player => {\n const positionMatches =\n selectedPosition === \"All\" ||\n player.POSITION_GROUP === selectedPosition;\n\n const yearMatches =\n effectiveYear === \"All\" ||\n player.YEAR_NUM === Number(effectiveYear);\n\n return (\n positionMatches &&\n yearMatches\n );\n});","pinned":false,"hidden":true,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":173,"value":"const summary = (() => {\n const draftedPlayers = filteredPlayers.filter(\n player => player.DRAFTED\n );\n\n const draftPicks = draftedPlayers\n .map(player => player.PICK_NUM)\n .filter(Number.isFinite);\n\n return {\n totalPlayers: filteredPlayers.length,\n draftedPlayers: draftedPlayers.length,\n\n draftRate: filteredPlayers.length\n ? draftedPlayers.length / filteredPlayers.length\n : 0,\n\n averagePick: draftPicks.length\n ? d3.mean(draftPicks)\n : null\n };\n})();","pinned":false,"hidden":true,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":174,"value":"$37","pinned":false,"hidden":true,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":175,"value":"$38","pinned":false,"hidden":false,"mode":"js","author":"$undefined","summary":"$undefined"},{"id":186,"value":"const yearLabel =\n effectiveYear === \"All\"\n ? \"All years\"\n : effectiveYear;","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":183,"value":"// Count players included in the selected year and position\nconst playerCounts = (() => {\n const relevantPlayers = players.filter(player => {\n const yearMatches =\n effectiveYear === \"All\" ||\n player.YEAR_NUM === Number(effectiveYear);\n\n const positionMatches =\n selectedPosition === \"All\" ||\n player.POSITION_GROUP === selectedPosition;\n\n return yearMatches && positionMatches;\n });\n\n return {\n all: relevantPlayers.length,\n\n drafted: relevantPlayers.filter(\n player => player.DRAFTED\n ).length\n };\n})();","pinned":false,"hidden":true,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":180,"value":"const allPlayersPlot = createHeatmap(\n heatmapComparison.all,\n `All Combine participants – ${yearLabel} (n = ${playerCounts.all})`\n);\n\nconst draftedPlayersPlot = createHeatmap(\n heatmapComparison.drafted,\n `Drafted players only – ${yearLabel} (n = ${playerCounts.drafted})`\n);\n\nhtml`\n

${allPlayersPlot}

${draftedPlayersPlot}

\n`","pinned":false,"hidden":true,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":182,"value":"html`\n

\n ${allPlayersPlot}\n

\n\n

\n ${draftedPlayersPlot}\n

\n`","pinned":false,"hidden":true,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":196,"value":"$39","pinned":false,"hidden":true,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":197,"value":"$3a","pinned":false,"hidden":true,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":198,"value":"// ------------------------------------------------------------\n// Format the values displayed inside the NBA heatmap\n// ------------------------------------------------------------\n\nconst formatNbaPositionValue = row => {\n const metric = nbaPositionMetrics.find(\n item => item.key === row.metricKey\n );\n\n if (!metric || !Number.isFinite(row.groupMean)) {\n return \"N/A\";\n }\n\n if (metric.format === \"percentage\") {\n const percentage =\n Math.abs(row.groupMean) <= 1\n ? row.groupMean * 100\n : row.groupMean;\n\n return `${percentage.toFixed(1)}%`;\n }\n\n return row.groupMean.toFixed(\n metric.decimals ?? 1\n );\n};","pinned":false,"hidden":true,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":199,"value":"$3b","pinned":false,"hidden":true,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":213,"value":"### Comparison of all Combine participants and drafted players\n\nThe two heatmaps present the average Combine measurements for point guards, shooting guards, small forwards, power forwards and centers. The left heatmap includes all Combine participants, while the right heatmap includes only players who were drafted. The values inside the cells represent the actual group averages. The colors indicate whether a value is below or above the overall average.\n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":200,"value":"html`\n

${allPlayersPlot}

${draftedPlayersPlot}

\n`","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":210,"value":"The results show a clear positional gradient in physical characteristics. Average height, weight, wingspan and standing reach increase progressively from point guards to centers. Guards generally perform better in speed and agility tests, while forwards and centers tend to record higher values in strength-related measurements.\n\nThese patterns remain largely unchanged when only drafted players are considered. However, drafted guards show slightly better results in several athletic tests, including sprint and vertical-jump measurements. This suggests that athletic performance may play a role in the selection of perimeter players, whereas size and reach are more defining characteristics for forwards and centers.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":214,"value":"### Average career outcomes of drafted Combine participants\nTo connect Combine measurements with subsequent professional performance, the third heatmap compares the NBA career statistics of drafted players across the five positions. The analysis includes playing time, scoring, rebounding, assists, shooting efficiency and efficiency and advanced performance indicators. Cell values represent the actual positional averages, while the colors show performance relative to all drafted players in the dataset.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":209,"value":"html`\n

\n ${nbaPositionPlot}\n

\n`","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":215,"value":"The NBA results reflect the different roles associated with each position. Point guards record the highest average number of assists per game and relatively high scoring values. Centers lead in rebounds, field-goal percentage, Win Shares per 48 minutes and Defensive Box Plus/Minus. These results are consistent with the primary responsibilities of guards as playmakers and centers as interior scorers, rebounders and defenders.\n\nShooting guards occupy an intermediate physical profile between point guards and forwards. However, they record the lowest average \"Box Plus/Minus\", \"Win Shares per 48 minutes\" and \"VORP\" among the five positions in this dataset. This indicates that physical and athletic Combine measurements alone may not sufficiently explain the later NBA performance of shooting guards","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":217,"value":"# Why Focus on Shooting Guards?\nThe positional overview shows that centers have the greatest physical variation and are strongly characterized by size-related measurements. Shooting guards, by contrast, occupy a less physically extreme position and combine requirements for size, speed, agility, shooting and decision-making. Their relatively low advanced NBA performance averages also raise the question of whether the Combine captures the abilities that are most relevant to their professional success.\n\nShooting guards are therefore selected as a focused case study to investigate which Combine measurements are associated with draft selection and later NBA performance, and which important performance dimensions may not be captured by the Combine.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":113,"value":"## Illustrative Comparison of Four NBA Players\nAs an initial player-level overview, four prominent NBA players were selected: James Harden, Devin Booker, Stephen Curry and LeBron James. Harden and Booker represent shooting guards, while Curry and James provide additional reference profiles from related perimeter and forward roles. The purpose of this comparison is not to draw statistical conclusions from four cases, but to illustrate how different Combine profiles can be associated with successful NBA careers.\n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":268,"value":"## James Harden","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":269,"value":"display(html`\n

\n \n

\n Age 37
\n Cleveland Cavaliers\n

\n`);","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":271,"value":"$3c","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":272,"value":"## Devin Booker","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":273,"value":"display(html`\n

\n \n

\n Age 29
\n Phoenix Suns\n

\n`);","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":274,"value":"$3d","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":275,"value":"## Stephen Curry","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":276,"value":"display(html`\n

\n \n

\n Age 38
\n Golden State Warriors\n

\n`);","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":277,"value":"$3e","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":278,"value":"## LeBron James","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":279,"value":"display(html`\n

\n \n

\n Age 41
\n Philadelphia 76ers\n

\n`);","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":280,"value":"$3f","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":289,"value":"## Career Achievements and Earnings","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":282,"value":"$40","pinned":false,"hidden":true,"mode":"js","author":"$undefined","summary":"$undefined"},{"id":283,"value":"html`\n \n \n ${players1.map(p => html`\n \n \n \n \n \n \n \n \n \n `)}\n \n

Player	Team	Age	Seasons Played	Titles	Major Individual Awards	All-Star Selections	Salary	Endorsements
${p.player}	${p.team}	${p.age}	${p.seasons}	${p.championships}	${p.awards}	${p.allStar}	${(p.salary/1e6).toFixed(1)}M $	${(p.endorsements/1e6).toFixed(1)}M $

`","pinned":false,"hidden":false,"mode":"js","author":"$undefined","summary":"$undefined"},{"id":290,"value":"The comparison illustrates that career success cannot be evaluated through championships alone. Stephen Curry and LeBron James have each won four NBA titles and received multiple MVP awards, while James Harden has earned an MVP award and established himself as one of the greatest scorers of his generation despite never winning a championship. Devin Booker has fewer individual awards but has developed into a consistent All-Star and one of the league's leading scorers.\n\nThe earnings figures also reflect more than on-court performance. Salary depends on factors such as contract timing, age, experience and team circumstances, while endorsement income is strongly influenced by popularity, personal branding and global recognition. Curry and James therefore generate substantially higher endorsement earnings than Harden and Booker. These values are estimates and should be interpreted as an indication of commercial impact rather than exact or directly comparable income.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":187,"value":"$41","pinned":false,"hidden":true,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":188,"value":"// ------------------------------------------------------------\n// Normalize player names to make the matching more robust\n// ------------------------------------------------------------\n\nconst normalizePlayerName = value => {\n return String(value ?? \"\")\n .toLowerCase()\n .normalize(\"NFD\")\n .replace(/[\\u0300-\\u036f]/g, \"\")\n .replace(/[.'’]/g, \"\")\n .replace(/\\s+/g, \" \")\n .trim();\n};\n\n\n// ------------------------------------------------------------\n// Find the four players in the dataset\n// ------------------------------------------------------------\n\nconst selectedPlayerRecords = comparisonPlayers.map(selection => {\n const record = cleanData.find(player =>\n normalizePlayerName(player.PLAYER) ===\n normalizePlayerName(selection.playerName)\n );\n\n return {\n ...selection,\n record: record ?? null,\n\n displayName: selection.playerName\n .split(\", \")\n .reverse()\n .join(\" \")\n };\n});","pinned":false,"hidden":true,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":190,"value":"$42","pinned":false,"hidden":true,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":191,"value":"$43","pinned":false,"hidden":true,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":193,"value":"const combinePlayerPlot = createPlayerHeatmap(\n selectedCombineHeatmapData,\n playerCombineMetrics,\n \"NBA Draft Combine Profile\"\n);\n\nconst nbaPlayerPlot = createPlayerHeatmap(\n selectedNbaHeatmapData,\n playerNbaMetrics,\n \"NBA Career Performance\"\n);","pinned":false,"hidden":true,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":218,"value":"## Draft Combine and NBA Career Data\nThe first heatmap compares their physical and athletic Combine measurements relative to the average of their respective positions. The second heatmap presents their subsequent NBA career performance using scoring, playmaking, shooting-efficiency and advanced impact metrics. Missing Combine measurements are displayed as N/A and are not interpreted as zero.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":216,"value":"html`\n

${combinePlayerPlot}

${nbaPlayerPlot}

\n`","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":219,"value":"$44","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":291,"value":"$45","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":292,"value":"const careerRegressionModels = (() => {\n const shootingGuards = cleanData.filter(row => {\n const positions = String(row.POS ?? \"\")\n .split(\"-\")\n .map(position => position.trim());\n\n return (\n row[\"drafted / undrafted\"] === \"drafted\" &&\n positions.includes(\"SG\")\n );\n });\n\n const targets = [\n {\n key: \"WS48\",\n label: \"WS/48\",\n value: row => Number(row[\"WS/48\"])\n },\n {\n key: \"BPM\",\n label: \"BPM\",\n value: row => Number(row.BPM)\n },\n {\n key: \"VORP_PER_SEASON\",\n label: \"VORP per Season\",\n value: row => {\n const vorp = Number(row.VORP);\n const seasons = Number(row.YEARS_PLAYED);\n\n return Number.isFinite(vorp) &&\n Number.isFinite(seasons) &&\n seasons > 0\n ? vorp / seasons\n : NaN;\n }\n }\n ];\n\n return targets.map(target =>\n regressionHelpers.fitModel(\n shootingGuards,\n target\n )\n );\n})();","pinned":false,"hidden":true,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":294,"value":"$46","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":293,"value":"$47","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":295,"value":"The low R² values show that height, weight, wingspan and standing reach explain only **1.6% of the variation in WS/48**, **0.3% in BPM**, and **2.6% in VORP per season**. Physical Combine measurements alone therefore have very limited value for predicting later NBA performance. All four highlighted players performed better than predicted, suggesting that important qualities such as shooting, basketball intelligence, decision-making and mental strength are not captured by these measurements.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":235,"value":"const bookerCareerSeasons = [\"15-16\",\"16-17\",\"17-18\",\"18-19\",\"19-20\",\"20-21\",\"21-22\",\"22-23\",\"23-24\",\"24-25\",\"25-26\"];","pinned":false,"hidden":true,"mode":"js","author":"$undefined","summary":"$undefined"},{"id":236,"value":"const bookerCareerMetrics = ({\n \"WS/48\": [0.043,0.035,0.063,0.076,0.143,0.104,0.156,0.157,0.181,0.111,0.145],\n \"BPM\": [-2.8,-1.9,0.9,1.0,2.2,0.3,4.1,4.2,3.0,0.4,2.2],\n \"PER\": [11.9,14.6,18.1,20.2,20.6,19.2,21.3,22.0,21.9,19.3,20.4]\n});","pinned":false,"hidden":true,"mode":"js","author":"$undefined","summary":"$undefined"},{"id":304,"value":"$48","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":337,"value":"viewof sgTrendMetric = Inputs.radio(\n new Map([[\"TS%\", \"TS_pct\"], [\"eFG%\", \"eFG_pct\"]]),\n {value: \"TS_pct\", label: \"Metrik\"}\n)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":338,"value":"$49","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":339,"value":"$4a","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":305,"value":"sgRadarRaw = FileAttachment(\"sg_radar_data.csv\").csv({typed: true})","pinned":false,"hidden":true,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":307,"value":"sgRadarMetrics = [\n {key: \"WS48\", label: \"WS/48\"},\n {key: \"BPM\", label: \"BPM\"},\n {key: \"VORP\", label: \"VORP\"},\n {key: \"FG\", label: \"FG%\"},\n {key: \"FG3\", label: \"3P%\"},\n {key: \"FT\", label: \"FT%\"}\n]\n","pinned":false,"hidden":true,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":308,"value":"sgRadarColor = d3.scaleOrdinal()\n .domain(sgRadarRaw.map(d => d.player))\n .range([\"#E4572E\", \"#29335C\", \"#F3A712\", \"#669BBC\"])","pinned":false,"hidden":true,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":309,"value":"viewof sgRadarSelectedPlayers = Inputs.checkbox(\n sgRadarRaw.map(d => d.player),\n {value: sgRadarRaw.map(d => d.player), label: \"Spieler anzeigen\"}\n)","pinned":false,"hidden":true,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":310,"value":"sgRadarTooltip = {\n const div = d3.create(\"div\")\n .style(\"position\", \"fixed\")\n .style(\"pointer-events\", \"none\")\n .style(\"background\", \"white\")\n .style(\"border\", \"1px solid #ccc\")\n .style(\"border-radius\", \"4px\")\n .style(\"padding\", \"4px 8px\")\n .style(\"font-size\", \"12px\")\n .style(\"box-shadow\", \"0 1px 4px rgba(0,0,0,0.2)\")\n .style(\"opacity\", 0)\n .style(\"z-index\", 10);\n document.body.appendChild(div.node());\n invalidation.then(() => div.node().remove());\n return div;\n}\n","pinned":false,"hidden":true,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":317,"value":"sgRadarMaxYear = d3.max(sgRadarSeasonRaw, d => d.career_yr)","pinned":false,"hidden":true,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":318,"value":"viewof sgRadarMode = Inputs.radio([\"Career\", \"Season\"], {\n value: \"Career\",\n label: \"View\",\n format: (d) => d === \"Career\" ? \"Career (full career average)\" : \"Season (single year)\"\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":319,"value":"viewof sgRadarYear = Inputs.range([1, sgRadarMaxYear], {value: 1, step: 1, label: \"Carrear Year\"})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":311,"value":"sgRadarActiveData = {\n const base = sgRadarMode === \"Season\"\n ? sgRadarSeasonRaw.filter(d => d.career_yr === sgRadarYear)\n : sgRadarRaw;\n return base.filter(d => sgRadarSelectedPlayers.includes(d.player));\n}","pinned":false,"hidden":true,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":312,"value":"$4b","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":320,"value":"sgRadarChart","pinned":false,"hidden":true,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":322,"value":"sgRadarSeasonRaw = FileAttachment(\"sg_radar_data_by_season.csv\").csv({typed: true})","pinned":false,"hidden":true,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":329,"value":"sgTrendRaw = FileAttachment(\"shooting_stats_4players.csv\").csv({typed: true})","pinned":false,"hidden":true,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":330,"value":"sgTrendData = {\n const grouped = d3.groups(sgTrendRaw, d => d.Player);\n return grouped.flatMap(([player, rows]) => {\n const sorted = rows.slice().sort((a, b) => d3.ascending(a.Season, b.Season));\n return sorted.map((d, i) => ({...d, career_yr: i + 1}));\n });\n}","pinned":false,"hidden":true,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":334,"value":"## The limits of value metrics at the top\n\nOn the value metrics, these four players are nearly indistinguishable. Their shooting profiles are not: the differences there are real and easy to measure, and value metrics miss them completely. That gap is the finding. Once a player has already reached an elite level, metrics like WS/48 or VORP stop being useful for telling him apart from his peers, the real differences show up in how he scores, not in how much value he produces overall. Testing that idea properly needs the full dataset, which spans the whole range of career outcomes rather than four proven stars.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":287,"value":"## The Tactical Fingerprint: Four Superstars, Four Different Ways to Dominate\nTo truly understand these stylistic differences, we have to look beyond aggregate value metrics and examine the geometry and tactics of the court. While these four highly successful superstars operate in roughly the same position and share similar offensive responsibilities, they possess radically unique basketball DNAs. They are all defined by their historic scoring prowess, yet their varying shot selections prove that they dismantle defenses in completely different ways.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":302,"value":"$4c","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":297,"value":"### James Harden: The Isolation Architect\nJames Harden operates in a world of his own, heavily relying on Isolation plays. His teammates routinely clear out the floor, setting the stage for his signature 1-on-1 masterpieces. He methodically breaks down his defender to drive straight to the basket, but what makes him a true nightmare to guard is his lethal three-point shooting and that iconic step-back jumper. Fascinatingly, a look at his shot heatmap clearly highlights the right side of the floor as his preferred hunting ground.\n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":247,"value":"html`\n

\n\n

Shot Location Heatmap

\n\n

Signature Move: Step Back

\n\n

\n`","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":351,"value":"","pinned":true,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":298,"value":"### Devin Booker: The Mid-Range Maestro\nOf this group, Devin Booker stands out as the ultimate offensive all-rounder. Rather than leaning into one extreme play type, his shot selection is incredibly balanced. He is a genuine three-level scorer who can hit from anywhere, though he has a distinct affinity for operating inside the three-point arc. His bread and butter—and the move he is arguably most feared for—is a silky pull-up jumper after one or two quick dribbles to create separation.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":245,"value":"html`\n

\n\n \n

Shot Location Heatmap

\n \n

\n\n \n

Signature Move: Pullup Jumpshot

\n \n \n

\n\n

\n`","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":299,"value":"### Stephen Curry: The Off-Ball Marathon Runner\nThe picture looks drastically different for Stephen Curry, especially when analyzing \"Off-Screen\" actions. No other player in this quartet even comes close to his 15.7% frequency in this category. Being on the smaller side, Curry utilizes his teammates' screens as weapons to create slivers of space for an open shot. Defending him is universally described as a horror show because guarding Curry means running a relentless marathon. At an exhausting average of nearly 120 meters per minute, defenders are forced to chase him endlessly through a maze of bodies. His heatmap confirms what we already know: a massive, glowing concentration of shots from beyond the arc. He isn't just the greatest shooter of all time by accident.\n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":248,"value":"html`\n

\n\n

Shot Location Heatmap

\n\n

Signature Move: Pullup Jumpshot

\n\n

\n`","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":300,"value":"### LeBron James: The Freight Train\nLeBron James leverages his massive, powerful frame to dictate an entirely different style of play, turning his sheer physicality into an overwhelming advantage. Once he gets downhill, he attacks the rim like a freight train. Defenders are largely helpless to stop him because he is simply too strong, a fact that becomes especially evident during rapid transition plays. His heatmap tells the exact same story, featuring a scorching hot zone directly under the basket. Yet, intriguingly, the data also uncovers a distinct \"cold spot\" on the right side of the court where he almost never attempts a shot.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":249,"value":"html`\n

\n\n

Shot Location Heatmap

\n\n

Signature Move: Driving Layup

\n\n

\n`","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined"},{"id":296,"value":"$4d","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"}],"files":"$W4e","stdlib":"2","titling":"auto"}} 31:"$30:body"