Post by eric on Jul 16, 2015 19:34:53 GMT -6
What does it mean to have a position? What does it mean to change a position? What does it mean to play out of position? Awaken your spirit! Now be rid of that filthy body!
On the eighth day I made fifteen copies of Raef LaFrentz. Centers, power forwards, small forwards, shooting guards, and point guards; all with 50 in everything. They had identical grades in Inside Scoring, Outside Scoring, Handling, and Rebounding. This was expected but still nice to see. Their grades in defense ranged from 59 (C+) to 70 (B): behold, the +1s etc. of position changes. It only ever has to do with defense (so long as you aren't hard capping yourself for some insane reason), which in turn only ever has to do with the soft (and meaningless) caps on each position's defense formula.
Position changes' +1 and -1 mean nothing.
Okay, but what about on the court? So I simulated 20 seasons of the identical teams with everyone in their proper position on the depth chart (center at center, etc.), then 20 more with centers playing at point guard and vice versa. Nothing changed.
The top two are the results of players playing center on the depth chart, the bottom two as point guards. The values are per 36 minutes per 20 seasons, so centers were getting roughly 9 boards 1 block a game, point guards were getting 15 points 5 assists, and so on. There are dramatic differences between the positions being played on the court, but none between the positions written on the roster sheet.
Roster position means nothing.
Okay, but depth chart position clearly means something. So I ran 60 seasons in each of inside, balanced, and outside offenses to see what. Here are the raw results, 5/4/3/2/1, inside/balanced/outside, team totals at the end.
There's a lot of stuff going on here, so let's go from most fixed to least.
Constant wrt Offense and Position
Free Throw % (makes sense)
Constant wrt Offense
Rebounds: bigs 520, wings 420, point 390
Blocks: bigs 60, wings 30, point 25
Fouls: bigs 210, wings 110, point 90
Steals: shooting guards 90(!), centers 60, everyone else 70
Three Point %: bigs 30%, small forwards 27%, guards 25%
Not Constant but Same Pattern
FTA/use: bigs, wings, point. In order of offense the splits were 10/7/6, 8.5/6.5/5, 8/6/5.
TOV/use: bigs, wings, point. 14/16/23, 14/16/19, 15/15/16
3PA/use: small forward, shooting guard, point, bigs. 14/12/11/7, 16/15/13/10, 19/17/15/13
Uses
Again going 5/4/3/2/1 i/b/o and in % of total team uses, we have...
22-22-17-17-22
17-17-19-19-28
15-15-17-22-31
Outside is the really weird one. I would have thought SF would have kept being the same as SG, but it actually goes DOWN. I think this explains a lot of why PGs have such good stats in our league: everyone runs outside offense and point guards use a comically high number of touches in those sets, and even if people do run inside bigs don't get THAT many touches.
Assists are also a weird one. They are roughly correlated with uses, but ast/tov ratios are all over the place. The point guard had an awful 1.22 for inside offense and a still pretty terrible but markedly better 1.45 in outside, but their decrease in tov/use far outpaced the ast/use increase. Point guards like everyone else foul the same in all offenses, so it's not a case of barreling into the lane out of control and racking up charges. My guess is the system is simulating difficulties in feeding the post or trying to drive and kick out of a dangerously congested post area.
Here are the team-wide trends going from inside to outside offense:
FGA/use up and 3PA/use up even more (so taking more FGAs and taking a higher percentage of them from three)
FTA/use and TOV/use down
Overall scoring efficiency unchanged (but a team that shoots above friggin' 27% from three and 40% from the field would probably see it go up)
AST/TOV up
Huge increase in PG uses
Large increase in SG uses
No change in SF uses
Large decrease in PF and C uses
.
It makes a lot of sense to run outside, then get defensive bigs on the assumption that a one-way player is in higher supply than a two-way player. Note finally that roster position does matter for training camps and what positions they're allowed to play in real (sim) life, so keep an eye on those.
On the eighth day I made fifteen copies of Raef LaFrentz. Centers, power forwards, small forwards, shooting guards, and point guards; all with 50 in everything. They had identical grades in Inside Scoring, Outside Scoring, Handling, and Rebounding. This was expected but still nice to see. Their grades in defense ranged from 59 (C+) to 70 (B): behold, the +1s etc. of position changes. It only ever has to do with defense (so long as you aren't hard capping yourself for some insane reason), which in turn only ever has to do with the soft (and meaningless) caps on each position's defense formula.
Position changes' +1 and -1 mean nothing.
Okay, but what about on the court? So I simulated 20 seasons of the identical teams with everyone in their proper position on the depth chart (center at center, etc.), then 20 more with centers playing at point guard and vice versa. Nothing changed.
MP FG FGA FT FTA 3P 3PA REB AST STL BLK TOV PF PT uses
31622 73 185 28 45 8 25 174 45 23 18 35 66 182 240
31414 74 192 29 48 7 25 173 44 19 20 35 74 184 248
29754 127 311 31 50 13 53 134 95 23 8 79 29 297 412
29582 126 309 30 49 12 53 128 100 24 8 78 30 295 408The top two are the results of players playing center on the depth chart, the bottom two as point guards. The values are per 36 minutes per 20 seasons, so centers were getting roughly 9 boards 1 block a game, point guards were getting 15 points 5 assists, and so on. There are dramatic differences between the positions being played on the court, but none between the positions written on the roster sheet.
Roster position means nothing.
Okay, but depth chart position clearly means something. So I ran 60 seasons in each of inside, balanced, and outside offenses to see what. Here are the raw results, 5/4/3/2/1, inside/balanced/outside, team totals at the end.
MP FG FGA FT FTA 3P 3PA REB AST STL BLK TOV PF PT uses
95243 277 727 139 226 19 66 518 159 55 61 132 221 713 959
95596 299 777 140 229 20 71 521 167 68 55 139 198 759 1017
89323 228 573 74 122 27 102 394 137 69 34 120 108 557 747
89386 239 572 77 126 22 90 429 138 90 28 120 115 577 748
90005 283 676 86 138 26 104 394 270 69 24 222 87 679 959
MP FG FGA FT FTA 3P 3PA REB AST STL BLK TOV PF PT uses
94590 222 570 87 142 22 76 521 132 63 58 105 209 553 738
94956 227 577 89 146 23 77 525 128 63 58 105 202 566 746
88223 275 679 76 121 38 142 404 158 70 34 137 107 664 869
88769 271 670 80 132 34 131 434 165 89 29 135 117 656 864
89109 380 927 92 148 39 161 391 294 70 23 234 89 891 1226
MP FG FGA FT FTA 3P 3PA REB AST STL BLK TOV PF PT uses
92920 193 486 71 116 26 86 528 119 53 63 94 221 483 631
93426 211 535 75 123 26 91 527 129 68 57 96 200 524 684
87330 238 608 60 95 40 147 407 138 69 32 114 107 576 764
87718 315 801 91 147 43 177 444 194 89 31 150 120 764 1015
88672 439 1084 94 150 54 210 394 325 68 25 224 87 1026 1374
FG FGA FT FTA 3P 3PA REB AST STL BLK TOV PF PT uses
1327 3325 517 841 115 432 2256 871 352 202 733 728 3285 4429
1374 3424 425 690 156 588 2275 877 355 201 716 724 3330 4443
1397 3513 391 631 189 711 2300 906 346 208 677 735 3373 4468There's a lot of stuff going on here, so let's go from most fixed to least.
Constant wrt Offense and Position
Free Throw % (makes sense)
Constant wrt Offense
Rebounds: bigs 520, wings 420, point 390
Blocks: bigs 60, wings 30, point 25
Fouls: bigs 210, wings 110, point 90
Steals: shooting guards 90(!), centers 60, everyone else 70
Three Point %: bigs 30%, small forwards 27%, guards 25%
Not Constant but Same Pattern
FTA/use: bigs, wings, point. In order of offense the splits were 10/7/6, 8.5/6.5/5, 8/6/5.
TOV/use: bigs, wings, point. 14/16/23, 14/16/19, 15/15/16
3PA/use: small forward, shooting guard, point, bigs. 14/12/11/7, 16/15/13/10, 19/17/15/13
Uses
Again going 5/4/3/2/1 i/b/o and in % of total team uses, we have...
22-22-17-17-22
17-17-19-19-28
15-15-17-22-31
Outside is the really weird one. I would have thought SF would have kept being the same as SG, but it actually goes DOWN. I think this explains a lot of why PGs have such good stats in our league: everyone runs outside offense and point guards use a comically high number of touches in those sets, and even if people do run inside bigs don't get THAT many touches.
Assists are also a weird one. They are roughly correlated with uses, but ast/tov ratios are all over the place. The point guard had an awful 1.22 for inside offense and a still pretty terrible but markedly better 1.45 in outside, but their decrease in tov/use far outpaced the ast/use increase. Point guards like everyone else foul the same in all offenses, so it's not a case of barreling into the lane out of control and racking up charges. My guess is the system is simulating difficulties in feeding the post or trying to drive and kick out of a dangerously congested post area.
Here are the team-wide trends going from inside to outside offense:
FGA/use up and 3PA/use up even more (so taking more FGAs and taking a higher percentage of them from three)
FTA/use and TOV/use down
Overall scoring efficiency unchanged (but a team that shoots above friggin' 27% from three and 40% from the field would probably see it go up)
AST/TOV up
Huge increase in PG uses
Large increase in SG uses
No change in SF uses
Large decrease in PF and C uses
.
It makes a lot of sense to run outside, then get defensive bigs on the assumption that a one-way player is in higher supply than a two-way player. Note finally that roster position does matter for training camps and what positions they're allowed to play in real (sim) life, so keep an eye on those.
