A basketball world simulated from the body up. Watch only.
Every engine version changes how the league plays. After each one, the same games are played on the old version and the new one and the league's numbers are set side by side. A number is called moved only when its change is well beyond its noise, the spread the games themselves show; a smaller change is real or not, and this many games cannot say.
Players without the ball who move to a new spot after a pass now keep an arms length from the line a teammate with the ball is driving along, so they no longer step into the path of the drive.
No number moved by more than 2.8 times the games' own spread. Measured over 48 games against version 115; a number is called moved only past 2.8 times its noise.
| The league | Version 115 | Version 116 | Change ± noise | |
|---|---|---|---|---|
| possessions per 48 minutes | 180.438 | 180.250 | -0.188 ± 0.791 | within the noise |
| points per possession | 1.309 | 1.287 | -0.022 ± 0.016 | within the noise |
| effective field goal % | 63.4% | 62.3% | -1.064 points ± 0.842 | within the noise |
| turnovers per possession | 0.089 | 0.090 | +0.001 ± 0.003 | within the noise |
| offensive rebound % | 26.6% | 26.5% | -0.071 points ± 0.625 | within the noise |
| fouls per 100 possessions | 0.046 | 0.012 | -0.034 ± 0.026 | within the noise |
| spread of the margin (points) | 20.247 | 22.460 | +2.213 ± 3.119 | within the noise |
| home side's win % | 47.9% | 50.0% | +2.083 points ± 8.675 | within the noise |
Defenders running to their spots now also route round the attackers who are standing or running, not only round their own teammates and the ball, so a defender and an attacker end up inside each other less often.
No number moved by more than 2.8 times the games' own spread. Measured over 48 games against version 114; a number is called moved only past 2.8 times its noise.
| The league | Version 114 | Version 115 | Change ± noise | |
|---|---|---|---|---|
| possessions per 48 minutes | 180.208 | 180.438 | +0.229 ± 0.671 | within the noise |
| points per possession | 1.307 | 1.309 | +0.002 ± 0.014 | within the noise |
| effective field goal % | 62.6% | 63.4% | +0.757 points ± 0.668 | within the noise |
| turnovers per possession | 0.086 | 0.089 | +0.002 ± 0.002 | within the noise |
| offensive rebound % | 27.1% | 26.6% | -0.523 points ± 0.560 | within the noise |
| fouls per 100 possessions | 0.047 | 0.046 | -0.001 ± 0.017 | within the noise |
| spread of the margin (points) | 20.874 | 20.247 | -0.627 ± 2.999 | within the noise |
| home side's win % | 45.8% | 47.9% | +2.083 points ± 7.585 | within the noise |
A drive that a defender cuts off now goes as far round that defender as the driver's own first step carries them in the time the defender needs to react and slide across. Before, nearly four in five cut-off drives went the same two feet, whoever drove and whoever guarded. Across 48 games nothing in the league's headline numbers moved past what chance alone explains.
No number moved by more than 2.8 times the games' own spread. Measured over 48 games against version 113; a number is called moved only past 2.8 times its noise.
| The league | Version 113 | Version 114 | Change ± noise | |
|---|---|---|---|---|
| possessions per 48 minutes | 180.479 | 180.208 | -0.271 ± 0.895 | within the noise |
| points per possession | 1.294 | 1.307 | +0.012 ± 0.017 | within the noise |
| effective field goal % | 62.3% | 62.6% | +0.349 points ± 0.857 | within the noise |
| turnovers per possession | 0.084 | 0.086 | +0.002 ± 0.003 | within the noise |
| offensive rebound % | 26.1% | 27.1% | +1.046 points ± 0.594 | within the noise |
| fouls per 100 possessions | 0.059 | 0.047 | -0.012 ± 0.026 | within the noise |
| spread of the margin (points) | 24.680 | 20.874 | -3.805 ± 3.334 | within the noise |
| home side's win % | 47.9% | 45.8% | -2.083 points ± 6.976 | within the noise |
Players off the ball who step to a new spot now get the ball on the move. Before, a pass waited until the receiver had arrived and stood still; now it leads a player running to a place they chose, and is caught no sooner than they can get there. In the league's games, catches beyond the three-point line made on the move rose from about 6% to about 22%, and catches from long two range from about 12% to about 41%. Teammates also run round each other, so two of them end up inside each other in about half as many possessions.
No number moved by more than 2.8 times the games' own spread. Measured over 48 games against version 112; a number is called moved only past 2.8 times its noise.
| The league | Version 112 | Version 113 | Change ± noise | |
|---|---|---|---|---|
| possessions per 48 minutes | 181.458 | 180.479 | -0.979 ± 1.006 | within the noise |
| points per possession | 1.304 | 1.294 | -0.010 ± 0.015 | within the noise |
| effective field goal % | 63.0% | 62.3% | -0.765 points ± 0.803 | within the noise |
| turnovers per possession | 0.083 | 0.084 | +0.001 ± 0.004 | within the noise |
| offensive rebound % | 25.4% | 26.1% | +0.744 points ± 0.738 | within the noise |
| fouls per 100 possessions | 0.033 | 0.059 | +0.027 ± 0.028 | within the noise |
| spread of the margin (points) | 22.100 | 24.680 | +2.580 ± 3.417 | within the noise |
| home side's win % | 45.8% | 47.9% | +2.083 points ± 8.675 | within the noise |
Defenders running to their spots now route round their own teammates, so two defenders end up inside each other far less often: in about a tenth of possessions, down from a quarter.
No number moved by more than 2.8 times the games' own spread. Measured over 48 games against version 111; a number is called moved only past 2.8 times its noise.
| The league | Version 111 | Version 112 | Change ± noise | |
|---|---|---|---|---|
| possessions per 48 minutes | 180.271 | 181.458 | +1.188 ± 0.668 | within the noise |
| points per possession | 1.297 | 1.304 | +0.007 ± 0.011 | within the noise |
| effective field goal % | 62.0% | 63.0% | +1.004 points ± 0.555 | within the noise |
| turnovers per possession | 0.081 | 0.083 | +0.001 ± 0.002 | within the noise |
| offensive rebound % | 26.7% | 25.4% | -1.347 points ± 0.568 | within the noise |
| fouls per 100 possessions | 0.023 | 0.033 | +0.009 ± 0.020 | within the noise |
| spread of the margin (points) | 19.763 | 22.100 | +2.337 ± 3.058 | within the noise |
| home side's win % | 43.8% | 45.8% | +2.083 points ± 6.976 | within the noise |
Each defender now stands off their own man by their own distance, drawn from how loose or tight they play, instead of everyone sharing one stance.
No number moved by more than 2.8 times the games' own spread. Measured over 48 games against version 110; a number is called moved only past 2.8 times its noise.
| The league | Version 110 | Version 111 | Change ± noise | |
|---|---|---|---|---|
| possessions per 48 minutes | 178.958 | 180.271 | +1.312 ± 0.975 | within the noise |
| points per possession | 1.281 | 1.297 | +0.015 ± 0.017 | within the noise |
| effective field goal % | 61.5% | 62.0% | +0.564 points ± 0.834 | within the noise |
| turnovers per possession | 0.085 | 0.081 | -0.003 ± 0.003 | within the noise |
| offensive rebound % | 26.1% | 26.7% | +0.637 points ± 0.737 | within the noise |
| fouls per 100 possessions | 0.059 | 0.023 | -0.035 ± 0.035 | within the noise |
| spread of the margin (points) | 21.211 | 19.763 | -1.448 ± 2.990 | within the noise |
| home side's win % | 50.0% | 43.8% | -6.250 points ± 7.536 | within the noise |
Every player now has their own reaction time, so a quicker, more aware defender stands closer to the player they guard and a slower one gives more room. The advance up the floor also lets the other team clear first.
No number moved by more than 2.8 times the games' own spread. Measured over 48 games against version 109; a number is called moved only past 2.8 times its noise.
| The league | Version 109 | Version 110 | Change ± noise | |
|---|---|---|---|---|
| possessions per 48 minutes | 178.979 | 178.958 | -0.021 ± 0.873 | within the noise |
| points per possession | 1.287 | 1.281 | -0.005 ± 0.013 | within the noise |
| effective field goal % | 61.6% | 61.5% | -0.150 points ± 0.668 | within the noise |
| turnovers per possession | 0.084 | 0.085 | +0.001 ± 0.003 | within the noise |
| offensive rebound % | 26.0% | 26.1% | +0.081 points ± 0.682 | within the noise |
| fouls per 100 possessions | 0.046 | 0.059 | +0.013 ± 0.042 | within the noise |
| spread of the margin (points) | 20.081 | 21.211 | +1.131 ± 3.013 | within the noise |
| home side's win % | 54.2% | 50.0% | -4.167 points ± 6.630 | within the noise |