As of Oct 21, 3:05 PM ET, before tonight’s 11 games (first tip 7:00 PM ET). Injuries checked 3:05 PM ET. Projection snapshot ba352b68. Injury news can change this before lock.
The model’s top two value picks both come with role-change warnings. Micah Potter ($3,000, 29.7 projected points) changed teams and is listed Questionable. Julian Reese ($3,000, 29.2) also changed teams and came off the bench in all three of his preseason games, which the projection does not see. Moussa Cisse ($3,300, 32.0) is third, with 3.6 more minutes than his recent average. The biggest minutes story is the Clippers: five players are Out for 118.6 combined average minutes, and Jalen Pickett, who joined from Denver, is expected to play 27.6 minutes against a weighted average of 13.5.
All 10 board picks rest on a baseline from last season, and 320 of the 382 players in the pool have baselines more than 45 days old. Rosters and minutes may have changed since.
Who is out and where the minutes go
49 players are listed Out. These ten have the highest average minutes:
- Maxime Raynaud, SAC, $5,500, averages 31.8 minutes
- Brandon Ingram, LAC, $5,600, averages 31.5
- Moses Moody, GSW, $3,000, averages 31.1
- Kon Knueppel, CHA, $6,800, averages 30.9
- Brandin Podziemski, GSW, $5,900, averages 30.8
- Aaron Nesmith, IND, $4,900, averages 30.0
- Jimmy Butler, GSW, $6,000, averages 29.7
- Jusuf Nurkic, UTA, $5,900, averages 28.4
- Max Strus, LAC, $3,800, averages 27.9
- Santi Aldama, DAL, $4,700, averages 27.5
DraftKings lists Blake Hinson (UTA, $3,000, averages 21.1 minutes) Out and the second injury source returned no status for him, so the two sources do not agree and his absence is unconfirmed. The Utah figures below count his minutes as out.
Two teams stand out. Golden State has Moody, Podziemski, Butler and Kristaps Porzingis out, 117.4 rotation minutes in all. Charles Bassey, who averages 11.7 minutes, is also Out but falls under the model’s 12-minute cutoff for counting an absence. Los Angeles has Ingram, Strus, Jordan Miller, Bradley Beal and Yanic Konan Niederhauser out, 118.6 minutes. The model moves minutes to the players who remain, by different amounts depending on how many minutes they were playing. Expected minutes against the recency-weighted average:
- Jalen Pickett, LAC: averages 9.8 (13.5 weighted), expected 27.6, with 118.6 teammate-minutes out. His history is from Denver, and he played 2 preseason games for the Clippers at 15.0 minutes
- Cam Christie, LAC: averages 6.2 (4.6 weighted), expected 18.7, with 118.6 out
- Gradey Dick, LAC: averages 6.2 (4.3 weighted), expected 18.4, with 118.6 out
- Isaiah Jackson, LAC: averages 18.2 (17.6 weighted), expected 25.3, with 118.6 out
- Kobe Sanders, LAC: averages 14.4 (16.3 weighted), expected 24.0, with 118.6 out
- L.J. Cryer, GSW: averages 18.9 (19.5 weighted), expected 27.2, with 117.4 out
- Malevy Leons, GSW: averages 14.0 (16.5 weighted), expected 24.3, with 117.4 out
- Will Richard, GSW: averages 16.1 (11.2 weighted), expected 18.9, with 117.4 out
- Harrison Ingram, UTA: averages 3.7 (4.1 weighted), expected 9.1, with 61.8 out, and limited history (under 10 games)
- Mo Bamba, UTA: averages 11.0 (10.9 weighted), expected 15.9, with 61.8 out, and limited history (under 10 games)
The gains are +14.1 minutes for Pickett, Christie and Dick, +7.7 for Isaiah Jackson, Sanders, Cryer, Leons and Richard, and +5.0 for the two Utah players. Several of these start from very small bases, so a jump from 4.3 to 18.4 minutes is a large assumption. For how much minutes move when a rotation player sits, see the minutes and injuries data.
The value board
The board is the top 10 by projected points per $1,000 among players projecting 15 or more. Minutes show the recency-weighted average the model used, then expected minutes tonight. No team has published a lineup as of 3:05 PM ET, so no player below is described as a starter or a backup.
| Rank | Player | Team | Salary | Weighted avg to expected minutes | Projection | 80% range | Value | Note |
|---|---|---|---|---|---|---|---|---|
| 1 | Micah Potter | POR | $3,000 | 25.4 to 26.3 | 29.7 | 17.3 to 43.1 | 9.9x | Listed Questionable, projection assumes he plays; joined POR from IND; baseline from last season |
| 2 | Julian Reese | NOP | $3,000 | 30.2 to 29.6 | 29.2 | 16.9 to 42.6 | 9.7x | Joined NOP from WSH; off the bench in all 3 preseason games (6.0 min); minutes still from last season; low confidence |
| 3 | Moussa Cisse | DAL | $3,300 | 25.8 to 29.3 | 32.0 | 18.1 to 46.2 | 9.7x | Baseline from last season; model 32.0 vs DraftKings average 15.0; low confidence |
| 4 | Cormac Ryan | MIL | $3,000 | 31.3 to 31.0 | 29.0 | 16.6 to 42.3 | 9.7x | Baseline from last season |
| 5 | Oscar Tshiebwe | HOU | $3,000 | 24.9 to 24.6 | 28.1 | 15.8 to 41.5 | 9.4x | Baseline from last season; joined HOU from UTA |
| 6 | Cody Williams | MIN | $3,100 | 36.0 to 36.2 | 28.3 | 15.9 to 41.7 | 9.1x | Baseline from last season; joined MIN from UTA; averaged 16.3 min in 6 preseason games, started 2; model 28.3 vs DraftKings average 17.5; low confidence |
| 7 | Quenton Jackson | IND | $3,000 | 26.1 to 27.6 | 27.0 | 14.6 to 40.4 | 9.0x | Baseline from last season |
| 8 | Javon Small | MEM | $3,100 | 25.0 to 24.6 | 26.3 | 13.9 to 39.6 | 8.5x | Baseline from last season; off the bench in all 4 preseason games (10.5 min), possibly team-wide rest |
| 9 | Jericho Sims | MIL | $3,000 | 31.0 to 30.5 | 24.8 | 12.4 to 38.2 | 8.3x | Baseline from last season; model 24.8 vs DraftKings average 14.9, minutes do not explain it; low confidence |
| 10 | Brandon Williams | GSW | $3,900 | 24.0 to 27.4 | 32.0 | 18.1 to 46.2 | 8.2x | Baseline from last season; joined GSW from DAL (17.0 min in 4 preseason games) |
One of the 10 is listed Questionable, so the board is shown again below without him.
Micah Potter, POR, $3,000. He is listed Questionable, and the projection assumes he plays. He averages 21.5 minutes over his last 10 games and 25.4 weighted, and the model expects 26.3, a gain of 0.9 minutes with 30.3 teammate-minutes out. That history is from Indiana, and the model has not seen him with Portland’s roster. In 3 preseason games for Portland he averaged 13.0 minutes and did not start any of them. At 1.13 points per minute he projects to 29.7, range 17.3 to 43.1. The 5x benchmark is 15.0 and his DraftKings average is 20.2, which makes him a low-confidence pick.
Julian Reese, NOP, $3,000. He averages 30.3 minutes, 30.2 weighted, and the model expects 29.6, with no teammate-driven change. At 1.01 points per minute he projects to 29.2, range 16.9 to 42.6. The 5x benchmark is 15.0 and his DraftKings average is 31.1. He moved from Washington, where he started 7 of his last 10 games last season. For New Orleans he came off the bench in all 3 preseason games and averaged 6.0 minutes. The projection uses last season’s minutes, so 29.6 expected minutes sits far above anything he played this preseason, and he is a low-confidence pick.
Moussa Cisse, DAL, $3,300. His 10-game average is 19.8 minutes and the weighted figure is 25.8. The model expects 29.3, a gain of 3.6 with 61.9 teammate-minutes out (Aldama at 27.5, Caleb Martin at 18.0 and Dereck Lively II at 16.4). At 1.09 points per minute he projects to 32.0, range 18.1 to 46.2. The 5x benchmark is 16.5 and his DraftKings average is 15.0. The minutes gain covers only part of the distance between 32.0 and 15.0, and the data gives no other reason, so this is a low-confidence pick on a last-season baseline.
Cormac Ryan, MIL, $3,000. He averages 26.2 minutes and 31.3 weighted, with 31.0 expected, so the two averages differ by 5.1 minutes. At 0.94 points per minute he projects to 29.0, range 16.6 to 42.3. The 5x benchmark is 15.0 and his DraftKings average is 23.2.
Oscar Tshiebwe, HOU, $3,000. He averages 20.7 minutes and 24.9 weighted, with 24.6 expected and no teammate-driven gain. That history is from Utah, and in 3 preseason games for Houston he averaged 11.3 minutes with no starts. At 1.16 points per minute he projects to 28.1, range 15.8 to 41.5. The 5x benchmark is 15.0 and his DraftKings average is 19.2.
If the Questionable players sit
Removing Potter and applying the same rule to everyone else brings in one name. Kobe Brown (IND, $3,000) is expected to play 29.9 minutes and projects to 24.5, 8.2x. He came off the bench in all 4 preseason games and averaged 21.8 minutes, though the team’s starters sat across the preseason, so it may be rest.
| Rank | Player | Team | Salary | Expected minutes | Projection | Value |
|---|---|---|---|---|---|---|
| 1 | Julian Reese | NOP | $3,000 | 29.6 | 29.2 | 9.7x |
| 2 | Moussa Cisse | DAL | $3,300 | 29.3 | 32.0 | 9.7x |
| 3 | Cormac Ryan | MIL | $3,000 | 31.0 | 29.0 | 9.7x |
| 4 | Oscar Tshiebwe | HOU | $3,000 | 24.6 | 28.1 | 9.4x |
| 5 | Cody Williams | MIN | $3,100 | 36.2 | 28.3 | 9.1x |
| 6 | Quenton Jackson | IND | $3,000 | 27.6 | 27.0 | 9.0x |
| 7 | Javon Small | MEM | $3,100 | 24.6 | 26.3 | 8.5x |
| 8 | Jericho Sims | MIL | $3,000 | 30.5 | 24.8 | 8.3x |
| 9 | Brandon Williams | GSW | $3,900 | 27.4 | 32.0 | 8.2x |
| 10 | Kobe Brown | IND | $3,000 | 29.9 | 24.5 | 8.2x |
The public record so far has 2 graded players who were listed Questionable, and 1 of them did not play, against 3% of healthy players. Two players is too few to put a rate on.
Decisions still pending
Ten rotation players are listed Questionable: Deni Avdija (POR, $8,200, averages 35.9 minutes), Luke Kennard (PHX, $3,900, 32.5), Stephen Curry (GSW, $7,500, 29.4), Collin Sexton (LAL, $4,400, 29.2), Leonard Miller (CHI, $4,600, 28.8), Isaac Okoro (CHI, $3,200, 27.3), Steven Adams (HOU, $4,500, 25.8), Micah Potter (POR, $3,000, 21.5), Malik Monk (SAC, $5,000, 21.0) and Zach Collins (CHI, $3,900, 18.4).
If the Questionable players on their teams sit, projections move as follows:
- Donovan Clingan, POR: 31.9 to 35.1
- Tre Jones, CHI: 28.9 to 32.1
- Jalen Smith, CHI: 22.0 to 25.1
- Buddy Hield, CHI (averages 9.2 minutes): 12.0 to 23.0, a large move from a small base
- Zach Collins, CHI, himself Questionable: 23.7 to 31.0
- Micah Potter, on the board: 29.7 to 32.3
- Oscar Tshiebwe, on the board: 28.1 to 28.8
How the projections have done
One night of results is graded: Oct 20, covering 2 slates, 141 projections and 104 players who played. On points, the model missed by 7.09 on average, against 7.12 for DraftKings’ season average and 6.91 for the model with its injury logic turned off. On minutes, the model missed by 3.75 against 3.80 for a plain recent average. The 80% range held the actual result 79.8% of the time.
The value board did worse. Its 20 picks across the 2 slates hit 5x salary 45% of the time and averaged 26.3 points. Ranking by DraftKings’ season average instead gave 70% and 33.5 points. That is one night, so it says little either way, but it is the record. Method details are in how the projections are made and graded.
Before you use any of this
Everything here is as of the snapshot time above and can change before lock, and the 11 games tip between 7:00 PM and 10:30 PM ET. No lineups are confirmed yet. The 80% range means roughly four in five outcomes land inside it, so one in five do not. Two players in the pool have no baseline and no projection: Alexandre Sarr (WAS) and Gregory Jackson (MEM), whose names did not match in our player records. A range is not a guarantee, and with last season’s numbers underneath and preseason roles the model cannot see, tonight’s projections deserve extra suspicion.
If you want to talk through tonight’s slate, the DFS Engineer Playbook community on Skool covers NBA as well as NFL: join here.
FAQ
What does the 80% range mean?
The range runs from the 10th to the 90th percentile of past misses at the same projection level. About four in five outcomes should land inside it and about one in five fall outside, above or below. A player projected at 29.7 with a range of 17.3 to 43.1 can finish outside that window without the projection being wrong.
Why do a player’s minutes change when a teammate is out?
The model starts from a recent-average baseline and adjusts it for how many rotation minutes teammates are newly missing. Those minutes go to the players who remain, and the size differs for bench players, role players and high-minute players. Tonight Golden State and the Clippers each lose more than 117 minutes, so the adjustments there reach 14.1 minutes for some players.
How are the projections made and graded?
Projected points equal points per minute over the last 30 games times expected minutes. Expected minutes start from a recent-average baseline and shift with teammate absences. Projections are saved before lock and later graded against actual results, including the value board shown above. Early in the season the inputs come from last season, and only one night has been graded so far.