The season, simulated 20,000 times.Here is the distribution.
The free NBA Season Playbook: every team's projected record and range, the title and playoff race next to the market, and the opening-week games, straight from the simulation that powers the Olympus NBA page.
Built from data updated Oct 15, 2026 · 9:18 AM ET. A research guide, not betting advice.
One season, played out again and again
A projection is not one number. The simulation runs the 2026-27 season 20,000 times and reports the spread of what happens, so you can see how sure it is.
Game nights use a second engine. Each matchup runs through a possession-level simulation (10,000 runs per game, engine 5.0.2-calibrated-possession) built on current rosters, the injury report and each team's four factors. That is where the opening-week projected scores in this guide come from.
The honest scorecard
Judge the simulation on seasons it never saw.
Before the season, the simulation was rebuilt as it would have stood on opening day of two seasons it never saw, 2024-25 and 2025-26, and scored against what happened. To pass it had to beat simple baselines in both seasons: guessing 41 wins for everyone, a regressed copy of last season's record, and equal odds for every team on playoffs, conference and title.
2024-25: a typical miss of 10.6 wins; it beat the 41-win guess, the regressed copy of last season, equal playoff odds and equal title odds, but not equal conference odds.
2025-26: a typical miss of 12.3 wins; it beat the 41-win guess and equal playoff odds, but not the regressed copy of last season, equal conference odds and equal title odds.
The eventual champion got 8% (OKC, 2024-25) and 5% (NYK, 2025-26) of the simulation's title odds on opening day.
All 30 teams, every simulated season
Each bar is a team's middle 80% of simulated win totals. The royal dot is the simulation's average; the white tick is the market's median. Ordered by projected seed.
NYKDET
BOS
CLE
CHA
ATL
ORL
TOR
PHI
MIA
MIL
CHI
IND
BKN
WASOKCSAS
DEN
HOU
LAL
MIN
LAC
POR
PHX
GSW
NOP
DAL
SAC
MEM
UTAMarket: Kalshi exchange prices, de-vigged; exchange fees are excluded. These are the market's numbers, not the model's. Playoffs column: chance of a top-6 seed or winning the play-in.
Who wins it all, and who gets there
The simulation's championship and conference odds, with the market's de-vigged numbers beside them. Royal is the model; white is the market.
OKC20%Market 25%NYK14%Market 8%SAS14%Market 25%DET7%Market 3%BOS7%Market 5%DEN5%Market 3%CHA4%Market <1%LAL4%Market 2%CLE4%Market 3%MIN4%Market 3%| Team | Model | Market | Model: conf. finals |
|---|---|---|---|
| NYK | 26% | 20% | 42% |
| DET | 15% | 6% | 29% |
| BOS | 15% | 13% | 28% |
| CHA | 9% | 1% | 18% |
| CLE | 9% | 7% | 19% |
| ATL | 6% | 2% | 14% |
| Team | Model | Market | Model: conf. finals |
|---|---|---|---|
| OKC | 32% | 33% | 51% |
| SAS | 23% | 33% | 41% |
| DEN | 10% | 4% | 21% |
| LAL | 8% | 6% | 18% |
| HOU | 8% | 3% | 17% |
| MIN | 7% | 10% | 17% |
Market title odds: 4 sportsbooks, de-vigged. Conference market: Kalshi, de-vigged.
Where the two views split
Both directions, the four widest gaps each way. The simulation sees results; the market also sees rosters and injuries. These gaps are tracked forward against real results. They are not bets.
Simulation above the market
Market above the simulation
Where every team finishes
Each row is a team's spread of final seeds across every simulated season, brighter where it lands more often. Top 6 go straight to the playoffs; seeds 7 to 10 go to the play-in.
NYK84%13%DET73%22%BOS72%22%CLE60%30%CHA59%30%ATL50%35%ORL48%36%TOR44%38%PHI43%38%MIA38%39%MIL11%31%CHI9%26%IND4%18%BKN2%14%WAS1%9%OKC92%7%SAS86%12%DEN67%25%HOU61%29%LAL60%30%MIN58%31%LAC41%38%POR40%38%PHX39%38%GSW22%37%NOP13%31%DAL9%27%SAC6%21%MEM4%18%UTA3%16%Wins, then head-to-head win% among tied teams, then conference win%, then coin flip. Division-leader criterion not reproduced.
The simulation's rating, tiered
Ranked by the season simulation's team rating, the same order as the Power Rankings tab. A rating of 1500 is a league-average team.
OKC1657NYK1632SAS1631DET1592BOS1588DEN1575MIN1558CLE1558HOU1558LAL1558CHA1557ATL1538ORL1530TOR1528PHI1523POR1516LAC1516MIA1510PHX1510GSW1466NOP1433MIL1428DAL1413CHI1412SAC1396MEM1380IND1380UTA1375BKN1364WAS1335What the market thinks, for now
The award model is experimental. MVP and Sixth Man passed their pre-registered backtests and switch on once teams have played enough games; the other awards failed theirs and stay market-only. Below: the exchange's de-vigged top five for each.
| Market top 5 | Market |
|---|---|
| Victor Wembanyama | 28% |
| Shai Gilgeous-Alexander | 12% |
| Nikola Jokić | 11% |
| Luka Doncic | 11% |
| Anthony Edwards | 5% |
| Market top 5 | Market |
|---|---|
| Cameron Boozer | 23% |
| Caleb Wilson | 17% |
| AJ Dybantsa | 16% |
| Darryn Peterson | 14% |
| Yaxel Lendeborg | 11% |
| Market top 5 | Market |
|---|---|
| Victor Wembanyama | 53% |
| Anthony Davis | 12% |
| Ausar Thompson | 4% |
| Chet Holmgren | 3% |
| Rudy Gobert | 3% |
| Market top 5 | Market |
|---|---|
| Payton Pritchard | 15% |
| Ajay Mitchell | 12% |
| Dylan Harper | 9% |
| Ayo Dosunmu | 6% |
| Collin Gillespie | 2% |
| Market top 5 | Market |
|---|---|
| Pelle Larsson | 5% |
| Derik Queen | 5% |
| Khaman Maluach | 5% |
| Ajay Mitchell | 5% |
| Matas Buzelis | 5% |
| Market top 5 | Market |
|---|---|
| JJ Redick | 10% |
| Sean Sweeney | 9% |
| Will Hardy | 8% |
| Rick Carlisle | 6% |
| Erik Spoelstra | 6% |
| Market top 5 | Market |
|---|---|
| Shai Gilgeous-Alexander | 26% |
| Jalen Brunson | 26% |
| Stephen Curry | 24% |
| Jayson Tatum | 23% |
Kalshi exchange prices, power de-vigged across the field; exchange fees are excluded.
The first 16 games, simulated
Projected scores and win probabilities from the possession simulation, next to the market line. These re-run every morning until tip, so the live page will move as injury news breaks.
Tuesday, October 20
Wednesday, October 21
Thursday, October 22
Simulated Oct 15, 2026 · 7:17 AM ET. Lines: the market's, as stored with each simulation. Team ranks use last season until this season's games are played.
All 30, in projected order
Each page: projected record and range, the market's view, last season's ratings, top scorers, the schedule, who is out, and the simulation's read, written from that team's own numbers.
New York Knicks
A real contender in the simulation: 14% to win the title and 26% to come out of the East.
Model and market nearly agree here, 54.1 simulated wins to a 51.9 market median.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Miles McBride | 12.0 | 2.4 | 2.6 | 41 |
| Landry Shamet | 9.3 | 1.8 | 1.4 | 51 |
| Jordan Clarkson | 8.6 | 1.8 | 1.3 | 72 |
Detroit Pistons
Last season's 60 wins get trimmed to 50.4: the simulation pulls every team toward the middle, and the further out a team finished, the harder the pull.
There is little daylight between the two views: 50.4 from the simulation, 49.0 from the market.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Cade Cunningham | 23.9 | 5.5 | 9.9 | 64 |
| Jalen Duren | 19.5 | 10.5 | 2.0 | 70 |
| John Collins | 13.6 | 5.3 | 1.0 | 69 |
Boston Celtics
Last season's 56 wins get trimmed to 50.2: the simulation pulls every team toward the middle, and the further out a team finished, the harder the pull.
Last season's profile is lopsided: #19 offense, #1 defense. The defense carried the rating.
The schedule helps: #28 of 30 by opponent rating, among the softest slates on the calendar.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Paul George | 17.3 | 5.3 | 3.6 | 37 |
| Payton Pritchard | 17.0 | 3.9 | 5.2 | 79 |
| Derrick White | 16.5 | 4.4 | 5.4 | 77 |
Cleveland Cavaliers
From 52 wins a year ago to 47.1 in the simulation, a 4.9-win haircut that is regression to the mean rather than a verdict on the roster.
Model and market nearly agree here, 47.1 simulated wins to a 48.7 market median.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Donovan Mitchell | 27.9 | 4.5 | 5.7 | 70 |
| James Harden | 23.6 | 4.8 | 8.0 | 70 |
| Evan Mobley | 18.2 | 9.0 | 3.6 | 65 |
Charlotte Hornets
The simulation lands 9.0 wins above the market on the Hornets (46.9 vs a 37.9 market median). The simulation rates the Hornets from last season's 44 wins and nothing else; the market is pricing something those results don't show yet.
The simulation adds 2.9 wins to last season's 44, landing at 46.9: the point margin was better than the record.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Brandon Miller | 20.2 | 4.9 | 3.3 | 65 |
| Naz Reid | 13.6 | 6.2 | 2.2 | 77 |
| Dennis Schroder | 10.8 | 2.7 | 4.9 | 70 |
Atlanta Hawks
The simulation and the market are close on the Hawks: 45.0 wins against a 44.1 median.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Nickeil Alexander-Walker | 20.8 | 3.4 | 3.7 | 78 |
| Jock Landale | 10.6 | 5.7 | 1.7 | 68 |
| Corey Kispert | 9.2 | 2.3 | 1.6 | 58 |
Orlando Magic
Model and market nearly agree here, 44.3 simulated wins to a 44.7 market median.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Paolo Banchero | 22.2 | 8.4 | 5.2 | 72 |
| Franz Wagner | 20.6 | 5.2 | 3.3 | 34 |
| Desmond Bane | 20.1 | 4.1 | 4.1 | 82 |
Toronto Raptors
The simulation sits 5.4 wins under the market (43.7 vs 49.1). The simulation is anchored to last season's 46 wins; the market expects this roster to finish above that.
The play-in (38%) is the most common path; the playoffs come up in 64% of simulated seasons.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Kawhi Leonard | 27.9 | 6.4 | 3.6 | 65 |
| RJ Barrett | 19.3 | 5.3 | 3.3 | 57 |
| Scottie Barnes | 18.1 | 7.5 | 5.9 | 80 |
Philadelphia 76ers
The market is the optimist on the 76ers: its median is 51.1 wins, the simulation's is 43.4. The simulation is anchored to last season's 45 wins; the market expects this roster to finish above that.
The play-in (38%) is the most common path; the playoffs come up in 63% of simulated seasons.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Tyrese Maxey | 28.3 | 4.1 | 6.6 | 70 |
| VJ Edgecombe | 16.0 | 5.6 | 4.2 | 75 |
| Caleb Love | 10.4 | 2.3 | 2.5 | 49 |
Miami Heat
The market is the optimist on the Heat: its median is 48.0 wins, the simulation's is 42.1. The simulation is anchored to last season's 43 wins; the market expects this roster to finish above that.
The offense did the heavy lifting last season (#2 offense, #22 defense by rating).
Close to a coin flip: the Heat make the playoffs in 58% of simulated seasons, with a 39% chance of the play-in.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Giannis Antetokounmpo | 27.6 | 9.8 | 5.4 | 36 |
| Andrew Wiggins | 15.4 | 4.8 | 2.7 | 68 |
| Bobby Portis | 13.7 | 6.4 | 1.6 | 67 |
Milwaukee Bucks
Jaime Jaquez Jr., Kyle Kuzma and Tyler Herro are listed out on the current injury report. The season simulation rates teams from results, so that absence is not in its number yet; the market can price it today.
The simulation lands 5.7 wins above the market on the Bucks (33.6 vs a 27.9 market median). The simulation rates the Bucks from last season's 32 wins and nothing else; the market is pricing something those results don't show yet.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Kel'el Ware | 11.1 | 9.0 | 0.7 | 77 |
| Ousmane Dieng | 7.6 | 3.2 | 2.4 | 57 |
| Caris LeVert | 7.4 | 2.0 | 2.7 | 60 |
Chicago Bulls
Last season's profile is lopsided: #12 offense, #28 defense. The offense carried the rating.
The simulation and the market are close on the Bulls: 32.1 wins against a 30.9 median.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Norman Powell | 21.7 | 3.5 | 2.5 | 58 |
| Josh Giddey | 17.0 | 8.3 | 9.1 | 54 |
| Matas Buzelis | 16.3 | 5.8 | 2.1 | 77 |
Indiana Pacers
This is the league's widest split between the two views, and the market is the higher one: 46.7 wins against the simulation's 28.8. The simulation is anchored to last season's 19 wins; the market expects this roster to finish above that.
The simulation hands back 9.8 wins to a team that won 19 last year, landing at 28.8, mostly regression toward the middle.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Pascal Siakam | 24.0 | 6.6 | 3.8 | 62 |
| Kelly Oubre Jr. | 14.1 | 5.0 | 1.6 | 50 |
| Obi Toppin | 11.6 | 4.4 | 2.3 | 24 |
Brooklyn Nets
After 20 wins last season, the simulation climbs to 27.0: records that far below average tend to drift back up.
Outcomes cluster: 80% of simulated seasons land between 17 and 38 wins.
The schedule does them no favours: by the simulation's own team ratings, it is the #2 hardest in the league.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Julius Randle | 21.1 | 6.7 | 5.0 | 79 |
| Noah Clowney | 12.3 | 4.1 | 1.6 | 66 |
| Nolan Traore | 8.9 | 1.8 | 3.8 | 56 |
Washington Wizards
An 11.9-win gap runs the other way here: 24.3 simulated wins, 36.2 at the market median. The simulation is anchored to last season's 17 wins; the market expects this roster to finish above that.
17 wins last season, 24.3 in the simulation. Bad records regress up the same way good ones regress down.
The simulated range is tight by NBA standards, 14 to 35 wins in 80% of seasons.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Alex Sarr | 16.3 | 7.4 | 2.7 | 48 |
| Deandre Ayton | 12.5 | 8.0 | 0.8 | 72 |
| Tre Johnson | 12.2 | 2.8 | 2.0 | 60 |
Oklahoma City Thunder
The Thunder win the championship in 20% of simulated seasons and the West in 32%.
From 64 wins a year ago to 57.0 in the simulation, a 7.0-win haircut that is regression to the mean rather than a verdict on the roster.
The simulation sits 5.8 wins under the market (57.0 vs 62.8). Both views expect a step back from last season's 64 wins; the market simply keeps more of that record than the simulation's regression does.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Shai Gilgeous-Alexander | 31.1 | 4.3 | 6.6 | 68 |
| Jalen Williams | 17.1 | 4.6 | 5.5 | 33 |
| Isaiah Hartenstein | 9.2 | 9.4 | 3.5 | 47 |
San Antonio Spurs
Last season's 62 wins get trimmed to 54.4: the simulation pulls every team toward the middle, and the further out a team finished, the harder the pull.
The market is the optimist on the Spurs: its median is 60.5 wins, the simulation's is 54.4. Both views expect a step back from last season's 62 wins; the market simply keeps more of that record than the simulation's regression does.
A real contender in the simulation: 14% to win the title and 23% to come out of the West.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Victor Wembanyama | 25.0 | 11.5 | 3.1 | 64 |
| De'Aaron Fox | 18.6 | 3.8 | 6.2 | 72 |
| Stephon Castle | 16.7 | 5.3 | 7.4 | 68 |
Denver Nuggets
The Nuggets won 54 games last season. The simulation's 48.5 gives back 5.5 of them, the usual cost of finishing that far above average.
The offense did the heavy lifting last season (#1 offense, #21 defense by rating).
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Nikola Jokic | 27.7 | 12.9 | 10.7 | 65 |
| Jamal Murray | 25.4 | 4.4 | 7.1 | 75 |
| Aaron Gordon | 16.2 | 5.8 | 2.7 | 36 |
Houston Rockets
From 52 wins a year ago to 47.1 in the simulation, a 4.9-win haircut that is regression to the mean rather than a verdict on the roster.
A 4.1-win gap runs the other way here: 47.1 simulated wins, 51.2 at the market median. Both views expect a step back from last season's 52 wins; the market simply keeps more of that record than the simulation's regression does.
Split personality in the ratings: #18 on offense and #4 on defense.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Amen Thompson | 18.3 | 7.8 | 5.3 | 79 |
| Jabari Smith Jr. | 15.8 | 6.9 | 1.9 | 77 |
| Reed Sheppard | 13.5 | 2.9 | 3.4 | 82 |
Los Angeles Lakers
The Lakers won 53 games last season. The simulation's 46.8 gives back 6.2 of them, the usual cost of finishing that far above average.
The simulation and the market are close on the Lakers: 46.8 wins against a 47.6 median.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Austin Reaves | 23.3 | 4.7 | 5.5 | 51 |
| Quentin Grimes | 13.4 | 3.6 | 3.3 | 75 |
| Sandro Mamukelashvili | 11.2 | 4.9 | 1.9 | 80 |
Minnesota Timberwolves
A 3.9-win gap runs the other way here: 46.4 simulated wins, 50.3 at the market median. The simulation is anchored to last season's 49 wins; the market expects this roster to finish above that.
Last season's 49 wins get trimmed to 46.4: the simulation pulls every team toward the middle, and the further out a team finished, the harder the pull.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Anthony Edwards | 28.8 | 5.0 | 3.7 | 61 |
| LaMelo Ball | 20.1 | 4.8 | 7.1 | 72 |
| Jaden McDaniels | 14.8 | 4.2 | 2.7 | 73 |
LA Clippers
Here the market median is 27.1 wins and the simulation says 42.7, a 15.6-win gap in the model's direction. The simulation rates the Clippers from last season's 42 wins and nothing else; the market is pricing something those results don't show yet.
Brandon Ingram is listed out on the current injury report. The season simulation rates teams from results, so that absence is not in its number yet; the market can price it today.
Last season's profile is lopsided: #24 offense, #10 defense. The defense carried the rating.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Darius Garland | 18.8 | 2.4 | 6.7 | 45 |
| Rui Hachimura | 11.5 | 3.3 | 0.8 | 68 |
| Derrick Jones Jr. | 10.1 | 3.5 | 1.4 | 50 |
Portland Trail Blazers
The injury report lists Shaedon Sharpe as out. Results-based ratings can't see that yet — watch how the first weeks of games move this number.
Close to a coin flip: the Trail Blazers make the playoffs in 60% of simulated seasons, with a 38% chance of the play-in.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Deni Avdija | 24.2 | 6.9 | 6.7 | 66 |
| Jrue Holiday | 16.3 | 4.6 | 6.1 | 53 |
| Scoot Henderson | 14.2 | 2.7 | 3.7 | 30 |
Phoenix Suns
Collin Gillespie, Jalen Green and Miles Bridges are listed out on the current injury report. The season simulation rates teams from results, so that absence is not in its number yet; the market can price it today.
Split personality in the ratings: #26 on offense and #6 on defense.
This is a bubble team: 59% to reach the playoffs and 38% to land in the play-in.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Jordan Goodwin | 8.7 | 4.9 | 2.2 | 70 |
| Luke Kennard | 8.4 | 2.3 | 2.2 | 78 |
| Pat Spencer | 7.2 | 2.4 | 3.5 | 66 |
Golden State Warriors
Jimmy Butler III and Moses Moody are listed out on the current injury report. The season simulation rates teams from results, so that absence is not in its number yet; the market can price it today.
This is a bubble team: 39% to reach the playoffs and 37% to land in the play-in.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Stephen Curry | 26.6 | 3.6 | 4.7 | 43 |
| Brandin Podziemski | 13.8 | 5.1 | 3.7 | 82 |
| Brandon Williams | 13.0 | 2.9 | 3.9 | 66 |
New Orleans Pelicans
The simulation hands back 7.9 wins to a team that won 26 last year, landing at 33.9, mostly regression toward the middle.
Here the market median is 29.9 wins and the simulation says 33.9, a 4.0-win gap in the model's direction. The simulation rates the Pelicans from last season's 26 wins and nothing else; the market is pricing something those results don't show yet.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Trey Murphy III | 21.5 | 5.7 | 3.8 | 66 |
| Zion Williamson | 21.0 | 5.7 | 3.2 | 62 |
| Saddiq Bey | 17.7 | 5.6 | 2.5 | 72 |
Dallas Mavericks
Cooper Flagg is listed out on the current injury report. The season simulation rates teams from results, so that absence is not in its number yet; the market can price it today.
After 26 wins last season, the simulation climbs to 32.1: records that far below average tend to drift back up.
The market is the optimist on the Mavericks: its median is 35.2 wins, the simulation's is 32.1. The simulation is anchored to last season's 26 wins; the market expects this roster to finish above that.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Naji Marshall | 15.2 | 4.7 | 3.3 | 74 |
| P.J. Washington | 14.2 | 7.0 | 1.8 | 56 |
| Max Christie | 12.3 | 3.2 | 2.0 | 77 |
Sacramento Kings
The simulation lands 7.2 wins above the market on the Kings (30.0 vs a 22.8 market median). The simulation rates the Kings from last season's 22 wins and nothing else; the market is pricing something those results don't show yet.
After 22 wins last season, the simulation climbs to 30.0: records that far below average tend to drift back up.
The injury report lists Maxime Raynaud as out. Results-based ratings can't see that yet — watch how the first weeks of games move this number.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Keegan Murray | 14.0 | 5.7 | 1.7 | 23 |
| Daeqwon Plowden | 10.8 | 3.0 | 1.3 | 32 |
| Nique Clifford | 8.6 | 3.8 | 2.4 | 75 |
Memphis Grizzlies
The schedule does them no favours: by the simulation's own team ratings, it is the #3 hardest in the league.
25 wins last season, 28.6 in the simulation. Bad records regress up the same way good ones regress down.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Jerami Grant | 18.6 | 3.5 | 2.1 | 57 |
| Cedric Coward | 13.6 | 5.9 | 2.8 | 62 |
| GG Jackson | 12.5 | 4.3 | 1.5 | 55 |
Utah Jazz
The market is the optimist on the Jazz: its median is 39.9 wins, the simulation's is 27.8. The simulation is anchored to last season's 22 wins; the market expects this roster to finish above that.
The simulation hands back 5.8 wins to a team that won 22 last year, landing at 27.8, mostly regression toward the middle.
Last season's profile is lopsided: #9 offense, #30 defense. The offense carried the rating.
| Player | PTS | REB | AST | GP |
|---|---|---|---|---|
| Lauri Markkanen | 26.7 | 6.9 | 2.1 | 42 |
| Keyonte George | 23.6 | 3.7 | 6.1 | 54 |
| Brice Sensabaugh | 14.9 | 3.1 | 1.9 | 75 |
Olympus NBA, tab by tab
This playbook is a snapshot. The NBA page re-runs every morning on the latest rosters, injury report and market lines.
Free every morning, or everything for $1
The free daily email carries the morning's free board when qualified projections are available. Members see every edge, stake and player projection.
$1 today · 3-day trial · card required · cancel anytime
The words in this guide
- Projection
- Any number the model produces for a game or a season. A projection has no result yet.
- Model view
- What the simulation says, shown next to what the market says. A view is not advice.
- Market median
- The win total the market splits 50/50 on, read from de-vigged exchange prices.
- 80% range
- The 10th to 90th percentile of simulated wins: four in five simulated seasons land inside it.
- De-vigged
- Market prices with the bookmaker's or exchange's margin removed, so they sum to 100%.
- Regression to the mean
- Extreme records drift back toward average the next season; the simulation builds that pull in.
- Glicko rating
- A team strength number updated after every game, with a built-in measure of how uncertain it is.
- Net rating
- Points scored minus points allowed per 100 possessions.
- Pace
- Possessions per game. Faster teams create more scoring chances for both sides.
- Play-in
- Seeds 7 to 10 in each conference compete for the last two playoff spots.
- Held-out season
- A season the model never saw while it was being fit, used to test it honestly.
- RMSE
- Root mean squared error: the typical miss, in wins, with big misses counted more.
- Brier score
- The average squared gap between a probability and what happened. Lower is better.