Player development
Every young player season from 2017-18 to 2025-26, 2,258 of them: who grew beyond the path players like him normally take, counted only where the growth helped a team win, and whether any team does it again and again.
Players develop. No team’s development has repeated.
−0.02how well a team’s results from 2017-18 to 2020-21 predict its results from 2021-22 to 2025-26 (1 would be perfectly, 0 not at all)
17%of random shuffles of players between teams spread the teams as widely as the real ones did
3ways of measuring a player’s season, the same answer from each
With the site’s chained ratings the two halves line up a little more (+0.18, and 6% of shuffles spread as widely), which is not proof. A 2015 study at Harvard Sports Analysis found the same thing with a different measure: team effects on development did not hold up as statistically real. The table below is what happened, team by team. It is not a ranking of who is good at it.
What counts as development
- Growth beyond his normal path. Every player has an expected next season from his age, his years in the league, his draft slot, his level and how many possessions that level rests on. Young players improve on their own until about 25 to 27, and a bad year tends to bounce back. Only growth past that counts.
- Only where it helps win. Value counts above replacement level (−1.3 per 100 possessions here, the level of players any team can sign). A player going from awful to merely bad adds nothing, however many minutes he gets. Teams that played young players the most had only a little more of it.
- Per possession. A bigger role alone is not growth.
- Everyone counts. A young player with real minutes who didn’t play in the NBA again (through 2025-26) counts at replacement level (202 seasons), so waiving the misses doesn’t help a team here.
- Credited to the team he played for in the season the growth showed, split by possessions after a trade. In wins, at 30 points a win.
- Past the path, short of it, or on it. A season beat the path when he finished 0.5 per 100 possessions or more past it, and fell short when he finished that much below it. A season counts for a team when a quarter or more of his possessions that season came there.
What happened, team by team
wins from growth beyond the normal path, 2017-18 to 2025-26
1 Thunder+33.81st to 5th+5.9+27.91stShai Gilgeous-Alexander +13.9
Loading the Thunder’ report
2 Spurs+24.61st to 9th+6.7+17.92ndDejounte Murray +6.3
Loading the Spurs’ report
3 Raptors+15.82nd to 14th+10.7+5.13rdPascal Siakam +5.2
Loading the Raptors’ report
4 Grizzlies+13.93rd to 15th−3.2+17.113thJaren Jackson Jr. +5.2
Loading the Grizzlies’ report
5 Mavericks+13.82nd to 20th+9.1+4.716thLuka Dončić +9.6
Loading the Mavericks’ report
6 Celtics+12.44th to 15th+2.0+10.511thJayson Tatum +5.5
Loading the Celtics’ report
7 Heat+12.43rd to 17th+10.5+1.915thTyler Herro +4.6
Loading the Heat’ report
8 Pacers+11.63rd to 21st+5.2+6.412thTyrese Haliburton +6.6
Loading the Pacers’ report
9 Nuggets+10.03rd to 24th+9.4+0.628thNikola Jokić +7.7
Loading the Nuggets’ report
10 Bucks+9.33rd to 23rd+10.5−1.220thGiannis Antetokounmpo +9.3
Loading the Bucks’ report
11 Pistons+8.73rd to 26th+5.4+3.34thAndre Drummond +6.3
Loading the Pistons’ report
12 Cavaliers+8.35th to 22nd−2.6+10.918thDarius Garland +6.9
Loading the Cavaliers’ report
13 Hornets+8.15th to 24th+1.4+6.87thDevonte' Graham +4.2
Loading the Hornets’ report
14 Suns+7.95th to 22nd+1.5+6.45thCollin Gillespie +3.6
Loading the Suns’ report
15 76ers+2.96th to 29th+1.2+1.725thTyrese Maxey +8.2
Loading the 76ers’ report
16 Knicks+2.710th to 26th+1.0+1.78thImmanuel Quickley +2.1
Loading the Knicks’ report
17 Nets+2.511th to 25th+5.6−3.16thJarrett Allen +2.7
Loading the Nets’ report
18 Jazz+1.911th to 26th−0.5+2.423rdLauri Markkanen +2.9
Loading the Jazz’ report
19 Clippers+1.814th to 24th−1.0+2.821stIsaiah Hartenstein +1.3
Loading the Clippers’ report
20 Magic+1.69th to 28th−4.5+6.19thFranz Wagner +5.0
Loading the Magic’ report
21 Kings+1.511th to 27th+3.5−2.027thDe'Aaron Fox +3.3
Why it ranks low. Its young seasons beat expected 11 times in 66 (17%), 26th in the league, where the average is 22%.
Loading the Kings’ report
22 Pelicans+1.112th to 26th+4.4−3.319thAnthony Davis +2.7
Why it ranks low. No one measure stands out. It is a little short across the board, as the table below shows.
Loading the Pelicans’ report
23 Rockets+1.112th to 28th−1.4+2.522ndAmen Thompson +3.7
Why it ranks low. Its 21 rookie seasons started 0.10 per 100 below the normal first season for their slot, 23rd. Seven young players didn’t play in the NBA again through 2025-26 after a season here, 10th most in the league (the average is 5.9).
Loading the Rockets’ report
24 Hawks+0.610th to 29th+2.8−2.126thTrae Young +4.7
Why it ranks low. Its young seasons beat expected 13 times in 79 (16%), 27th in the league, where the average is 22%. Nine young players didn’t play in the NBA again through 2025-26 after a season here, 4th most in the league (the average is 5.9).
Loading the Hawks’ report
25 Lakers+0.213th to 28th−0.7+1.010thAustin Reaves +2.1
Why it ranks low. Its young players got few possessions, 130,892, 6th fewest in the league.
Loading the Lakers’ report
26 Timberwolves+0.29th to 30th−8.2+8.417thKarl-Anthony Towns +4.5
Why it ranks low. Its young players got the 6th most possessions in the league, 177,268, and 16 of their 74 seasons beat expected (22%, the league average is 22%).
Loading the Timberwolves’ report
27 Trail Blazers−0.812th to 28th+3.0−3.814thJusuf Nurkić +4.4
Why it ranks low. Its young seasons beat expected 12 times in 88 (14%), 29th in the league, where the average is 22%. 11 young players didn’t play in the NBA again through 2025-26 after a season here, 3rd most in the league (the average is 5.9).
Loading the Trail Blazers’ report
28 Wizards−4.119th to 29th−2.5−1.730thOtto Porter Jr. +2.2
Why it ranks low. Its young seasons beat expected 9 times in 80 (11%), 30th in the league, where the average is 22%. Its 25 rookie seasons started 0.13 per 100 below the normal first season for their slot, 24th.
Loading the Wizards’ report
29 Warriors−5.321st to 30th−1.7−3.529thKevon Looney +1.7
Why it ranks low. Its young players got few possessions, 112,154, 3rd fewest in the league. Its young seasons beat expected 10 times in 59 (17%), 25th in the league, where the average is 22%.
Loading the Warriors’ report
30 Bulls−8.123rd to 30th−1.3−6.724thZach LaVine +2.1
Why it ranks low. Its young players got the 2nd most possessions in the league, 193,892, and 18 of their 97 seasons beat expected (19%, the league average is 22%). Nine young players didn’t play in the NBA again through 2025-26 after a season here, 5th most in the league (the average is 5.9).
Loading the Bulls’ report
Open a team for its report: every young player behind its numbers, with his draft pick, how he got there and his seasons. Likely: where the team’s place falls 8 times in 10 when its players are drawn again at random. Raw growth: its place if every young player’s improvement counted, normal progression and all, the way most lists do it. The two agree only partly.
The biggest developers
wins from growth beyond the normal path, young seasons only- 1. Shai Gilgeous-Alexanderpicks 11 to 20, OKC +13.9+13.9wins2018-19−2.1per 1002023-24+7.6per 100Biggest jump+3.42022-23
- 2. Luka Dončićtop 3 pick, DAL +9.6, LAL +0.2+9.9wins2018-19+0.9per 1002024-25+6.4per 100Biggest jump+2.22019-20
- 3. Giannis Antetokounmpopicks 11 to 20, MIL +9.3+9.3wins2016-17+3.5per 1002019-20+8.0per 100Biggest jump+2.62018-19
- 4. Tyrese Maxeypicks 21 to 30, PHI +8.2+8.2wins2020-21−2.3per 1002025-26+3.3per 100Biggest jump+1.82023-24
- 5. Nikola Jokićsecond round, DEN +7.7+7.7wins2016-17+5.7per 1002020-21+5.9per 100Biggest jump+2.12020-21
- 6. Darius Garlandpicks 4 to 10, CLE +6.9+6.9wins2019-20−3.3per 1002024-25+2.6per 100Biggest jump+3.72021-22
- 7. Tyrese Haliburtonpicks 11 to 20, IND +6.6, SAC +0.2+6.8wins2020-21−0.9per 1002024-25+3.7per 100Biggest jump+2.72022-23
- 8. Andre Drummondpicks 4 to 10, DET +6.3+6.3wins2016-17−1.1per 1002018-19+3.9per 100Biggest jump+1.82018-19
- 9. Dejounte Murraypicks 21 to 30, SAS +6.3+6.3wins2016-17−2.5per 1002021-22+3.1per 100Biggest jump+1.92020-21
- 10. Victor Wembanyamatop 3 pick, SAS +5.9+5.9wins2023-24+2.5per 1002025-26+8.6per 100Biggest jump+3.82025-26
- 11. Jayson Tatumtop 3 pick, BOS +5.5+5.5wins2017-18+2.2per 1002023-24+4.1per 100Biggest jump+2.32021-22
- 12. Jaren Jackson Jr.picks 4 to 10, MEM +5.2+5.2wins2018-19−0.7per 1002024-25+3.6per 100Biggest jump+2.92024-25
Second round and undrafted
- Nikola Jokićsecond round, DEN+7.7
- Tre Jonessecond round, SAS, CHI+3.9
- Collin Gillespieundrafted, PHX+3.6
- Josh Richardsonsecond round, MIA+3.6
- Isaiah Hartensteinsecond round, NYK, LAC, HOU+3.4
- Julian Champagnieundrafted, SAS+3.4
- Isaiah Joesecond round, OKC, PHI+3.2
- Fred VanVleetundrafted, TOR+3.0
- Jamal Sheadsecond round, TOR+2.6
- Davis Bertanssecond round, SAS, WAS+2.5
- Moussa Diabatésecond round, CHA+2.4
- Jalen Brunsonsecond round, DAL+2.3
- Devonte' Grahamsecond round, CHA, NOP+2.3
- Sam Hauserundrafted, BOS+2.2
- Naz Reidundrafted, MIN+2.2
Wins from growth beyond the normal path in their young seasons (25 and under, or a first four seasons).
Top 10 picks who started slow
- Darius Garlandpicks 4 to 10, CLE, rookie season −0.8 against his slot+6.9
- Stephon Castlepicks 4 to 10, SAS, rookie season −0.6 against his slot+3.5
- De'Aaron Foxpicks 4 to 10, SAC, rookie season −0.7 against his slot+3.3
- Anthony Edwardstop 3 pick, MIN, rookie season −0.3 against his slot+2.9
- Brandon Millertop 3 pick, CHA, rookie season −1.7 against his slot+2.3
- Jalen Suggspicks 4 to 10, ORL, rookie season −0.5 against his slot+2.1
- Reed Sheppardtop 3 pick, HOU, rookie season −0.8 against his slot+1.5
A rookie season below the normal one for his slot and age, then growth beyond his path after it.
Better than the age curve, 26 to 31
Nikola Jokićsecond round, DEN +19.6, 2021-22 to 2025-26+19.6
Season Age Team Possessions Level 2020-21 25 DEN 5,143 +5.9 2021-22 26 DEN 5,162 +8.6 2022-23 27 DEN 4,868 +8.6 2023-24 28 DEN 5,648 +8.6 2024-25 29 DEN 5,490 +9.4 2025-26 30 DEN 4,783 +8.6 Level: points per 100 possessions against an average player that season. Counted: 2021-22 to 2025-26. Dimmed, not counted: 2020-21 (his starting point).
Kawhi Leonardpicks 11 to 20, LAC +6.0, TOR +3.2, SAS −0.1, 2017-18 to 2020-21+9.1
Season Age Team Possessions Level 2016-17 25 SAS 4,960 +6.0 2017-18 26 SAS 427 +1.9 2018-19 27 TOR 4,327 +3.2 2019-20 28 LAC 3,909 +7.5 2020-21 29 LAC 3,649 +6.5 2022-23 31 LAC 3,637 +5.3 2023-24 32 LAC 4,809 +4.5 2024-25 33 LAC 2,422 +3.3 2025-26 34 LAC 4,299 +7.6 Level: points per 100 possessions against an average player that season. Counted: 2017-18 to 2020-21. Dimmed, not counted: 2016-17 (his starting point), 2022-23 (counted for another team) and 2023-24 to 2025-26 (after 31).
Giannis Antetokounmpopicks 11 to 20, MIL +8.8, 2020-21 to 2025-26+8.8
Season Age Team Possessions Level 2019-20 25 MIL 3,841 +8.0 2020-21 26 MIL 4,373 +5.5 2021-22 27 MIL 4,705 +6.2 2022-23 28 MIL 4,356 +4.6 2023-24 29 MIL 5,485 +5.2 2024-25 30 MIL 4,829 +6.5 2025-26 31 MIL 2,171 +5.0 Level: points per 100 possessions against an average player that season. Counted: 2020-21 to 2025-26. Dimmed, not counted: 2019-20 (his starting point).
Paul Georgepicks 4 to 10, OKC +7.9, LAC +0.6, 2017-18 to 2021-22+8.5
Season Age Team Possessions Level 2016-17 26 IND 5,486 +2.9 2017-18 27 OKC 5,984 +1.2 2018-19 28 OKC 6,227 +7.0 2019-20 29 LAC 3,063 +3.1 2020-21 30 LAC 3,662 +4.9 2021-22 31 LAC 2,267 +2.4 2022-23 32 LAC 4,037 +3.5 2023-24 33 LAC 5,179 +4.9 2024-25 34 PHI 2,764 0.0 2025-26 35 PHI 2,398 +1.4 Level: points per 100 possessions against an average player that season. Counted: 2017-18 to 2021-22. Dimmed, not counted: 2016-17 (his starting point) and 2022-23 to 2025-26 (after 31).
Joel Embiidtop 3 pick, PHI +8.4, 2020-21 to 2025-26+8.4
Season Age Team Possessions Level 2019-20 25 PHI 3,184 +3.4 2020-21 26 PHI 3,368 +6.0 2021-22 27 PHI 4,703 +7.1 2022-23 28 PHI 4,702 +7.3 2023-24 29 PHI 2,770 +6.4 2024-25 30 PHI 1,197 +0.9 2025-26 31 PHI 2,540 +3.0 Level: points per 100 possessions against an average player that season. Counted: 2020-21 to 2025-26. Dimmed, not counted: 2019-20 (his starting point).
Derrick Whitepicks 21 to 30, BOS +7.7, SAS +0.7, 2021-22 to 2025-26+8.4
Season Age Team Possessions Level 2020-21 26 SAS 2,224 +1.5 2021-22 27 SAS, BOS 4,623 +1.8 2022-23 28 BOS 4,858 +3.8 2023-24 29 BOS 4,920 +3.9 2024-25 30 BOS 5,238 +2.4 2025-26 31 BOS 5,343 +4.1 Level: points per 100 possessions against an average player that season. Counted: 2021-22 to 2025-26. Dimmed, not counted: 2020-21 (his starting point).
James Hardendrafted before 2010, HOU +8.3, BKN −0.9, 2017-18 to 2020-21+7.4
Season Age Team Possessions Level 2016-17 27 HOU 6,282 +5.4 2017-18 28 HOU 5,291 +6.2 2018-19 29 HOU 5,988 +6.7 2019-20 30 HOU 5,339 +6.9 2020-21 31 BKN, HOU 3,387 +3.6 2021-22 32 BKN, PHI 5,076 +3.2 2022-23 33 PHI 4,373 +4.0 2023-24 34 LAC 5,108 +2.8 2024-25 35 LAC 5,787 +2.6 2025-26 36 LAC, CLE 5,027 +2.1 Level: points per 100 possessions against an average player that season. Counted: 2017-18 to 2020-21. Dimmed, not counted: 2016-17 (his starting point) and 2021-22 to 2025-26 (after 31).
Donovan Mitchellpicks 11 to 20, CLE +7.4, 2022-23 to 2025-26+7.4
Season Age Team Possessions Level 2021-22 25 UTA 4,709 +2.1 2022-23 26 CLE 4,968 +3.5 2023-24 27 CLE 4,078 +4.9 2024-25 28 CLE 4,755 +3.4 2025-26 29 CLE 5,002 +4.9 Level: points per 100 possessions against an average player that season. Counted: 2022-23 to 2025-26. Dimmed, not counted: 2021-22 (his starting point).
Shai Gilgeous-Alexanderpicks 11 to 20, OKC +7.3, 2024-25 to 2025-26+7.3
Season Age Team Possessions Level 2023-24 25 OKC 5,436 +7.6 2024-25 26 OKC 5,546 +9.3 2025-26 27 OKC 4,671 +9.0 Level: points per 100 possessions against an average player that season. Counted: 2024-25 to 2025-26. Dimmed, not counted: 2023-24 (his starting point).
Rudy Gobertpicks 21 to 30, UTA +7.0, MIN −0.1, 2018-19 to 2023-24+6.9
Season Age Team Possessions Level 2017-18 25 UTA 3,738 +2.9 2018-19 26 UTA 5,496 +3.9 2019-20 27 UTA 4,544 +3.0 2020-21 28 UTA 4,593 +5.6 2021-22 29 UTA 4,403 +4.5 2022-23 30 MIN 4,611 −0.1 2023-24 31 MIN 5,375 +2.4 2024-25 32 MIN 4,957 +1.7 2025-26 33 MIN 5,113 +1.2 Level: points per 100 possessions against an average player that season. Counted: 2018-19 to 2023-24. Dimmed, not counted: 2017-18 (his starting point) and 2024-25 and 2025-26 (after 31).
Jrue Holidaydrafted before 2010, MIL +3.9, NOP +3.0, 2017-18 to 2021-22+6.9
Season Age Team Possessions Level 2016-17 26 NOP 4,563 +1.1 2017-18 27 NOP 6,265 +2.7 2018-19 28 NOP 5,260 +4.1 2019-20 29 NOP 4,677 +1.3 2020-21 30 MIL 4,154 +3.2 2021-22 31 MIL 4,686 +3.8 2022-23 32 MIL 4,662 +4.4 2023-24 33 BOS 4,681 +0.1 2024-25 34 BOS 3,842 −0.1 2025-26 35 POR 3,366 +2.4 Level: points per 100 possessions against an average player that season. Counted: 2017-18 to 2021-22. Dimmed, not counted: 2016-17 (his starting point) and 2022-23 to 2025-26 (after 31).
Fred VanVleetundrafted, TOR +4.2, HOU +2.4, 2020-21 to 2024-25+6.6
Season Age Team Possessions Level 2019-20 25 TOR 3,966 +1.7 2020-21 26 TOR 4,001 +2.9 2021-22 27 TOR 5,041 +3.0 2022-23 28 TOR 5,236 +3.0 2023-24 29 HOU 5,668 +4.3 2024-25 30 HOU 4,457 +2.0 Level: points per 100 possessions against an average player that season. Counted: 2020-21 to 2024-25. Dimmed, not counted: 2019-20 (his starting point).
Wins from seasons at 26 to 31 against the normal path for a player his age and level, after the young seasons the teams are credited with. Open a player for his level each season. A season above the path adds wins and one below it takes them away, so a jump that fell back the next season nets out. Seasons after 31 aren’t counted: a player who fell off later keeps what he added from 26 to 31.
How it’s measured, and what it can’t see
- A season, on its own. His lineups’ results that season, centered on what his own box score that season says (offense less points allowed per 100 possessions). Nothing carries over from the season before, so growth shows the season it happens.
- The normal path is fitted on every player with 200 or more possessions in a season since 2016-17. It explains 46% of the spread in next season’s level.
- How he got here. Each player’s draft pick, every trade his pick or his draft rights went through, his own trades, signings and waiver claims (from July 2015) and the team he played each game for. Drafted here: the team made the pick, with its own pick or one it acquired in a trade. Draft rights acquired: another team made the pick and traded his rights here, on draft night or later, before his first NBA game. First NBA team: another team made the pick, and he played his first NBA game here. In a team’s report, select how he got here where it is underlined for the trades behind it.
- The tests. Team totals in the first seasons against the last. Team labels shuffled at random within each season, 2,000 times. The same study with the box score alone and with the site’s chained ratings.
- What it can’t see. Growth in the G League or overseas shows only once a player gets NBA minutes. A team’s staff changes over ten seasons, and a team is credited as one. The seasons start in 2016-17, so earlier growth (Jokić’s first seasons, say) isn’t here.