Punting turnovers barely moves the top 20, but Deni Avdija jumps 37 places

Deni Avdija ranks 66th of 150 in a nine-category pool. Remove turnovers from the scoring and he ranks 29th. Josh Giddey goes from 61st to 31st and Jaylen Brown from 50th to 24th. All three lose more than two standard deviations to turnovers, and only seven players in the pool are that badly off.

This column is about punting turnovers, which means giving up the category on purpose and building around the other eight. Format: nine-category head-to-head, where you win a week by taking five of nine categories. Punt turnovers and you need five of the remaining eight.

Pool: the same 150 players as in the nine-cat top 50 and the earlier columns on blocks and free throws, taken from 467 NBA players with a 2025-26 season row (a player traded during the season counts both of his teams), keeping the 247 with at least 30 games and 20 minutes a night, then the top 150 by nine-category value. Every figure is a 2025-26 per-game average at that player's own minutes. Nothing here is a projection of 2026-27.

Why turnovers are different

Turnovers are the one category where a higher number is worse. The pool averages 1.89 turnovers a game with a standard deviation of 0.77. A player at 3.7, like Nikola Jokić, sits 2.34 standard deviations on the wrong side of the pool average.

They also travel with the things that make a player good. Across the 150 players, turnovers per game correlate 0.76 with assists and 0.74 with points, and only 0.16 with rebounds and -0.11 with blocks. The players who create the most for their team are the players who give the ball away the most.

That is why the top of the board looks the way it does. The ten best players in the nine-category ranking average 2.51 turnovers a game, against 1.89 for the pool. Seventeen of the 150 players commit 3.0 or more, and five of them are in the nine-category top 20.

How the punt ranking is built

Each player has a z-score in each category, which measures how many standard deviations he sits from the pool average. For turnovers the sign is flipped, so fewer turnovers is a positive score. To build the punt ranking I subtracted each player's turnover z-score from his nine-category total and re-ranked the 150. I did not rebuild the pool without turnovers, so a player just outside the 150 could rank differently in a true punt draft. Read the ranks as a good guide to direction and size, not as exact draft positions.

What barely changes

The top holds. Seven of the ten best players in the nine-category ranking stay in the top ten: Nikola Jokić, Victor Wembanyama, Shai Gilgeous-Alexander, Kawhi Leonard, Tyrese Maxey, Luka Dončić and Donovan Mitchell. Nineteen of the top 20 hold, and 45 of the top 50. Jokić is 1st either way, and Dončić, who has the worst turnover score in the pool at -2.73, moves only from 6th to 4th.

The reason is scale. Both are so far ahead in points, rebounds, assists and shooting that removing one category does not close the gap.

The top ten does change at its edge. Jamal Murray (8th to 15th), Trey Murphy III (9th to 18th) and Lauri Markkanen (10th to 20th) drop out of the top ten, and Cade Cunningham (18th to 8th), Stephen Curry (11th to 9th) and Jalen Johnson (17th to 10th) come in. Murray, Murphy and Markkanen average 1.5 to 2.3 turnovers a game. Cunningham averages 3.7 and Johnson 3.4.

Who rises

Player	Nine-cat rank	Punt-TO rank	Turnovers per game	Assists per game	Minutes	Games
Deni Avdija	66	29	3.8	6.7	33.3	66
Josh Giddey	61	31	3.6	9.1	32.1	54
Jaylen Brown	50	24	3.6	5.1	34.4	71
Julius Randle	81	55	2.7	5.0	33.0	79
Devin Booker	59	38	3.1	6.0	33.5	64
Giannis Antetokounmpo	53	34	3.2	5.4	28.9	36
LeBron James	44	30	3.0	7.2	33.2	60
Alperen Sengun	39	26	3.2	6.2	33.3	72

Five players move into the top 50 from outside it: Avdija, Giddey, Antetokounmpo, Booker and Brandon Miller (51st to 44th, at 2.5 turnovers a game). Brown was already inside it, at 50th.

Antetokounmpo is the case to be careful with. He played 36 games, so his per-game line is a small slice of a season. A punt build makes his turnovers free, but it does not make his availability free.

Who falls

Player	Nine-cat rank	Punt-TO rank	Turnovers per game	Assists per game	Minutes	Games
Kel'el Ware	47	76	0.8	0.7	22.1	77
Neemias Queta	57	81	1.0	1.7	25.3	76
Mark Williams	62	83	1.1	1.0	23.6	60
Donovan Clingan	27	50	1.2	2.1	27.2	77
Mikal Bridges	34	56	1.0	3.7	32.8	82
Alex Sarr	43	53	1.7	2.7	27.2	48
Tyler Herro	46	52	1.9	4.1	31.3	33
Evan Mobley	45	51	1.9	3.6	31.9	65

Five players in the nine-category top 50 fall out of the punt top 50: Bridges, Sarr, Mobley, Herro and Ware. If you plan to punt turnovers, these are the players whose draft price is highest relative to what they give you. Three of the five, Sarr, Mobley and Ware, are centers who score and rebound without handling the ball much. The table also includes Queta, Williams and Clingan, who lose 21 to 24 places. Herro played 33 games, so treat his line with the same care as Antetokounmpo's.

The minutes assumption

A turnover is a per-minute rate multiplied by a guess about minutes. Take a made-up guard who commits 0.11 turnovers a minute.

• At 28 minutes a game: 3.08 turnovers, which is 1.19 above the pool average and about -1.54 standard deviations.
• At 34 minutes a game: 3.74 turnovers, which is 1.85 above the pool average and about -2.39 standard deviations.

The same player is worth 0.85 standard deviations less in this category at 34 minutes than at 28. If the minutes are a guess, the cost of the category is a guess too, and the guess matters most for the heaviest ball handlers, because their cost grows with every minute on the floor.

What this does not say

This is 2025-26 data, one season of it, and it says nothing about 2026-27 roles. A change of role changes a player's turnovers, and this list changes with it. The pool was fixed before turnovers were removed, and a player's rank in a real punt draft depends on who else is available. Punting is a choice for a team you have already built, not a rule for every draft.

The finding that holds is the size of the effect. At the top of the board it is small. In the middle of the draft, between roughly 25th and 85th, dropping one category moves the eight players in the first table by 13 to 37 places.