So, I recall reading more than once that passing works as follows: the sim adds up the total passing on the floor, and that determines how adequately the ball is moved from player to player.  So, two questions:

A) In the past, I've given less distribution in certain situations to players with better passing ratings, so that they would pass more.  Am I stupid?

B) Is there any weighting given to what position those ratings are coming from?  That is, if a PG has 90 PA and a C has 2, does that result in the same effectiveness of ball movement as if the C had the 90, and the PG had the 2?  That seems unlikely.  But again, I might be stupid.  Sub-question: if there is indeed weighting, any idea how significant it is?  

A.   Yes, I think you are stupid.  I don't think distro affects the "passing multiplier".  :)

B.  Good question that Ihave also wondered.  not just with passing but other ratings as well (ie rebounding).   (example - is a 60 REB PF and a 40 REB SF more effective than a 40 REB PF and a 60 REB SF?).    My intuition says that guards with higher passing ratings are more imprtant than LPs with higher passing ratings because the engine assumes they handle the ball more.   So all else being equal I'd put a better passer at SG rather than SF.

Well, we always knew you thought I was stupid :)

But here's where it gets counterintuitive for me -- many coaches are happy to take a PG with high PA/BH and very low LP/PE and put their distribution to 0, with the thinking that they're playing him as a pure PG, distributing the ball to the 4 scorers on the court.  But the fact is, by not shooting, this player is not "distributing" the ball any more than he would be if he had a 20 distribution.  Just because you're shooting less does not mean you're passing more.  In effect, everyone on the court is actually in a constant state of "passing."  Which is weird.

I think the key is understanding the engine breaks down the posession into 2 distinct areas.  (IMO)
1) Does the posession end in a made FG?  It takes into accounts all the individual ratings of both offense (including passing) and defense of the players on the floor.  it may also decide here if an assist was credited. (although I'm not sure if that would be in step #2 or not)
2) If the engine determines a FG is made it decides if there was an assist and who go the assist.  For this step it takes into account the individual passing ratings of the players and their distributions to assign the "assist man".

So in essence, if you are shooting less then you ARE passing more, but I don't think that (in itself) has anything to do with how well the offense performs.

Thing is, my understanding is that assists are just cosmetic -- they're awarded after the fact.  So non-shooters will get more assists just because there are more made baskets per game not made by them.  But it isn't because their passing "did the job" any more often.  So really, even if they took half the team's shots, they're still "passing" the same amount -- which is always; they're just getting fewer assists because there are fewer made baskets for which they are assist-elegible. 

Technically, they are assisting on their own made baskets just as much as they are on other people's -- the sim just has a line of code saying "if basket made by player X, player X cannot be cosmetically credited with that assist".

If the sim works by using a single constant passing rating based on the 5 players on the floor and their positions, then technically you are always passing -- even when you're shooting.

Posted by mullycj on 10/7/2011 1:00:00 PM (view original):
I think the key is understanding the engine breaks down the posession into 2 distinct areas.  (IMO)
1) Does the posession end in a made FG?  It takes into accounts all the individual ratings of both offense (including passing) and defense of the players on the floor.  it may also decide here if an assist was credited. (although I'm not sure if that would be in step #2 or not)
2) If the engine determines a FG is made it decides if there was an assist and who go the assist.  For this step it takes into account the individual passing ratings of the players and their distributions to assign the "assist man".

So in essence, if you are shooting less then you ARE passing more, but I don't think that (in itself) has anything to do with how well the offense performs.

I think #2 makes more sense.  Since a PG has 0 distrobution he wont be the awarded scorer.  His high pass rating gives him a high probability of being awarded the assist.

The better question imo is whether or not passing creates scoring plays, or if it is simply there to determine if there is a turnover before a scoring attempt. 

Posted by jeffdrayer on 10/7/2011 2:07:00 PM (view original):
Thing is, my understanding is that assists are just cosmetic -- they're awarded after the fact.  So non-shooters will get more assists just because there are more made baskets per game not made by them.  But it isn't because their passing "did the job" any more often.  So really, even if they took half the team's shots, they're still "passing" the same amount -- which is always; they're just getting fewer assists because there are fewer made baskets for which they are assist-elegible. 

Technically, they are assisting on their own made baskets just as much as they are on other people's -- the sim just has a line of code saying "if basket made by player X, player X cannot be cosmetically credited with that assist".

If the sim works by using a single constant passing rating based on the 5 players on the floor and their positions, then technically you are always passing -- even when you're shooting.

Yes, that is what I am saying.

I think the game is more complicated than the 2 steps ... more like:

1. Moving the ball to the frontcourt and starting the offense: PG's SPD/BH taken into account (or Passing, BH, SPD and maybe more if facing a press) ... results in a turnover, or a quick transition basket, or move to step 2.

2. Looking for a shot: this one is a combination of the Distro settings, the individual defense/offense matchup/differentials, team passing, team offense vs team defense, etc. ... results in a turn over, a foul, a quick transition basket, or a shot (step 3) and who is taking the shot.

3. Determine the outcome of the "normal shot": This would be based on the actual offensive player selected for the shot and the defensive player (or average of players in a zone). This is a foul, a turnover, a made shot or a missed shot (step 4)

4. Determine the outcome of the rebound .... result: Offensive rebound, defensive rebound, foul.

5. If Offensive rebound, either loop back to #2 (take it back out ... maybe for teams that are SLOW for tempo), or loop back to #3 ... OR ... if Defensive Rebound, move back to #1 with the other team.

====
I would think that different things are important in different steps (and all things are important in an entire game).

So, when determining "Step 2 outcome" then overall team passing is likely important, though it would not be hard to subtract out the shooter's passing if you pick the shooter at the beginning of the step ... so maybe only the OTHER 4 guys passing matters if player 1 gets the "normal shot".

In that case, the high PASS/BH, low LP/PER PG having a lower distribution would have a larger impact in passing (since he is 1 of 4 more often, taking less shots). I think everyone knows that seble said assists are cosmetic ... but having more means you have a better passing rating, which is helping the overall team passing.

Does tempo play a roll in passing? By that I mean if you play Slowdown you take up most of the shot-clock allowing for more movement of the ball and theoreticaly more passing to take place. Whereas with Uptempo you take a faster shot which would result in fewer passes. 

If that is the case, would one expect to see more a team with a lower relative passer rating commit fewer turnovers (based on bad passes, i.e. throw-aways and picked off pases) playing uptempo rather than Slowdown? For this instance lets assume that  Ath/Spd for the offense and defense are similar so as to not introduce a different variable.

Posted by jeffdrayer on 10/7/2011 12:34:00 AM (view original):
So, I recall reading more than once that passing works as follows: the sim adds up the total passing on the floor, and that determines how adequately the ball is moved from player to player.  So, two questions:

A) In the past, I've given less distribution in certain situations to players with better passing ratings, so that they would pass more.  Am I stupid?

B) Is there any weighting given to what position those ratings are coming from?  That is, if a PG has 90 PA and a C has 2, does that result in the same effectiveness of ball movement as if the C had the 90, and the PG had the 2?  That seems unlikely.  But again, I might be stupid.  Sub-question: if there is indeed weighting, any idea how significant it is?  

"So, I recall reading more than once that passing works as follows: the sim adds up the total passing on the floor, and that determines how adequately the ball is moved from player to player."

jeff, I don't recall ever reading anything like that, and I don't believe that's how it works. Would be curious to see where you'd read that, I could be mistaken.

I'm not much of an engine wonk but I feel like hughesjr is on track.

Daalt -- can't remember whether it was in the forums or a developer chat, but I've heard it said more than once, and know it was from a trusted source since I only keep things like this in my head if I hear them from OR, Rails, you, etc.  No way I'll be able to go back and find it -- I just remember reading it and thinking "wow, if Coach X says he's been told it works this way, that's surprising, but I guess it's true."

Anyone else remember hearing this?  I know I'm not making it up, because it always a) seemed like a weird way to structure things, and b) I'm not that creative.