Post by eric on Oct 20, 2015 13:41:23 GMT -6
I've always said that usage-increasing stats are underrated by my analysis, so I decided to do the analysis from two sides:
first with a mediocre player (50 below the max in every attribute, then +50 to a single attribute, record, do the next attribute)
then with a great player (max every attribute, then -50 to a single attribute, record, do the next attribute)
There was also an odd wrinkle for point guards specifically that didn't occur for either of the other positions I checked where increasing a defensive attribute made the team's offense dramatically worse. The first time I did the analysis I just chalked that up to small sample size and normalized, but now I think it's a real effect. Check it:
There's a lot going on here. The #1 takeaway is that Inside Scoring is still awesome, always, forever, and that Shot Blocking is still the best defensive upgrade for everyone everywhere too, but adding defense to a bad point guard always makes the team worse. My theory is this: in the same way that adding to a player's Jump Shot will make him want to take proportionally more jumpers even if he's still terrible at them, adding to a player's defense will make him want to expend proportionally more energy on the defensive end even if he's still terrible at it. This gives him less energy for his already terrible offense, so the team loses more on offense than it gains on defense, and the effect is more pronounced the more useless the defensive attribute is (hey there point guard Post Defense).
Okay. Now we have to talk about Passing. If we regress the "good" results against the "bad" we get pretty good linear correlation. The two that stand way out in opposite directions are Inside Scoring and Passing, and we'll get to the reasons for Inside Scoring in a second. It's pretty obvious that if I have an A+/A+ scoring grades guy I want him passing somewhere between never and not at all, that makes sense. What doesn't make sense is that both his possessions and the team's possessions go dramatically down when his Passing increases from 50 to 100. Thousands of team possessions per season just vanish, I can't find them anywhere. It creates the biggest gap in opponent vs. team possessions of any non-rebounding attribute, and all I can figure is the system has the player's team go deeper into the shot clock on average. This strategy works insofar as the ORtg goes up, but nowhere near enough to balance out the possessions lost.
Usage-wise it's the same story as always on the bad side. Inside Scoring the most, Strength at 2/3rds, Jumping at 1/3rd; Handling the only one to slightly reduce usage, a slight bump in turnovers from Defensive Rebounding. Like the SF, there are Jumping-sized bumps for Jump Shot, Three Shot, and Quickness. Unlike SF, there is only a Jumping-sized bump for Passing as opposed to a Strength-sized. On the good side, Inside Scoring had the same sized effect, then nobody else but Strength had an appreciable effect and even it was small, and as mentioned above Passing was a disaster. My feeling here is that usage has a cap, and so if a guy already has 100 Inside Scoring and even mediocre secondary attributes his usage just can't go up that much. Like Shot Blocking and defense, Inside Scoring so thoroughly dominates a player's usage rate that it's impossible to be an elite offensive player without it because the player just won't shoot enough.
Minutes: annoyingly, neither Inside Scoring nor Strength increase a PG's minutes. The only ones that moved the needle were Jump Shot, Handling, Quickness, and Passing. This makes sense, obviously, the coach wants his PG to have good Handling grades and an outside shot would be nice too, so those are the attributes he stresses. I suspect we will see a mix of PG and SF attributes for the SG ones that increase MP. Minutes never changed for the good player. Like usage, a player can only play so many minutes, and having three of the minutes played attributes maxed was enough to redline it. I don't know if it's additive or multiplicative, but either way two attributes won't be enough to max MP. If you're really desperate to bump minutes do Quickness, never Passing.
.
Bottom line, here's where I would put as upgrades for a PG:
I will be updating the new guys article, then moving on to SG. Eventually I will come back around to SF and C to repeat the "good" analysis for them as well.
first with a mediocre player (50 below the max in every attribute, then +50 to a single attribute, record, do the next attribute)
then with a great player (max every attribute, then -50 to a single attribute, record, do the next attribute)
There was also an odd wrinkle for point guards specifically that didn't occur for either of the other positions I checked where increasing a defensive attribute made the team's offense dramatically worse. The first time I did the analysis I just chalked that up to small sample size and normalized, but now I think it's a real effect. Check it:
bad good attribute
-0.5 4.7 inside scoring
1.3 6.6 jump shot
0.5 3.4 three shot
1.0 5.7 handling
0.3 1.4 quickness
-1.5 -1.4 passing
-1.5 0.6 stealing
-1.7 2.6 shot blocking
-2.4 -0.3 post defense
-0.6 0.1 perimeter defense
-1.9 0.0 drive defense
-1.4 1.9 offensive rebounding
-0.8 2.0 defensive rebounding
-1.0 0.9 strength
-2.8 -0.3 jumpingThere's a lot going on here. The #1 takeaway is that Inside Scoring is still awesome, always, forever, and that Shot Blocking is still the best defensive upgrade for everyone everywhere too, but adding defense to a bad point guard always makes the team worse. My theory is this: in the same way that adding to a player's Jump Shot will make him want to take proportionally more jumpers even if he's still terrible at them, adding to a player's defense will make him want to expend proportionally more energy on the defensive end even if he's still terrible at it. This gives him less energy for his already terrible offense, so the team loses more on offense than it gains on defense, and the effect is more pronounced the more useless the defensive attribute is (hey there point guard Post Defense).
Okay. Now we have to talk about Passing. If we regress the "good" results against the "bad" we get pretty good linear correlation. The two that stand way out in opposite directions are Inside Scoring and Passing, and we'll get to the reasons for Inside Scoring in a second. It's pretty obvious that if I have an A+/A+ scoring grades guy I want him passing somewhere between never and not at all, that makes sense. What doesn't make sense is that both his possessions and the team's possessions go dramatically down when his Passing increases from 50 to 100. Thousands of team possessions per season just vanish, I can't find them anywhere. It creates the biggest gap in opponent vs. team possessions of any non-rebounding attribute, and all I can figure is the system has the player's team go deeper into the shot clock on average. This strategy works insofar as the ORtg goes up, but nowhere near enough to balance out the possessions lost.
Usage-wise it's the same story as always on the bad side. Inside Scoring the most, Strength at 2/3rds, Jumping at 1/3rd; Handling the only one to slightly reduce usage, a slight bump in turnovers from Defensive Rebounding. Like the SF, there are Jumping-sized bumps for Jump Shot, Three Shot, and Quickness. Unlike SF, there is only a Jumping-sized bump for Passing as opposed to a Strength-sized. On the good side, Inside Scoring had the same sized effect, then nobody else but Strength had an appreciable effect and even it was small, and as mentioned above Passing was a disaster. My feeling here is that usage has a cap, and so if a guy already has 100 Inside Scoring and even mediocre secondary attributes his usage just can't go up that much. Like Shot Blocking and defense, Inside Scoring so thoroughly dominates a player's usage rate that it's impossible to be an elite offensive player without it because the player just won't shoot enough.
Minutes: annoyingly, neither Inside Scoring nor Strength increase a PG's minutes. The only ones that moved the needle were Jump Shot, Handling, Quickness, and Passing. This makes sense, obviously, the coach wants his PG to have good Handling grades and an outside shot would be nice too, so those are the attributes he stresses. I suspect we will see a mix of PG and SF attributes for the SG ones that increase MP. Minutes never changed for the good player. Like usage, a player can only play so many minutes, and having three of the minutes played attributes maxed was enough to redline it. I don't know if it's additive or multiplicative, but either way two attributes won't be enough to max MP. If you're really desperate to bump minutes do Quickness, never Passing.
.
Bottom line, here's where I would put as upgrades for a PG:
gain growth mp usage attribute
highest slowest yes medium Jump Shot
high slow no highest Inside Scoring
high fast yes none Handling
medium slow no none Shot Blocking
low slowest no high Strength
medium fastest no medium Three Shot
low slowest yes medium Quickness
medium slow no none ReboundingsI will be updating the new guys article, then moving on to SG. Eventually I will come back around to SF and C to repeat the "good" analysis for them as well.
