it is reasonable to me that the case you are showing would be treated the same as having the same guy at the 2 and another guy at the 3 at both spots. the sim engine is thinking - is it better two have the second man at pf and 4th at sf or 3rd at both? programmer hack would just have some crude value assigned to playing the nth man on the depth chart, so you could do a simple ordering, and its probably linear. so 2/4 vs 3/3 is probably no different than 2/3 vs 3/2. personally i would make it scale down (assuming the diff between 1 and 2 is bigger than 2 and 3, which is bigger than 3 and 4). but its extremely reasonable that it would have been made linearly.

i have noticed in the old engine that it was not deterministic which guy subbed in where in these tie situations (hence the method to get around it i mentioned above). i never noticed if they were swapping without notice - i never really paid attention either. anyway, in both of the situations you listed, it seems to me there is no obvious tie breaker that would be used. in both cases it was sanders and dale at pf and sf - which is arbitrary, to the sim engine, i think.

i do think its messed up it would just switch it for no apparent reason. if they added a 5th position, it would let the user make the choice. i suspect that is easier than actually allowing you to specify in all the cases (think about when the same guy is 2nd at all 3 spots and another is 3rd about all 3 spots, or other variations - could get nasty). i would also change the calculation of the best depth chart to say 2/4 is better than 3/3 so it was deterministic to the user, and the only case they had to resort to using position 5 for is if it was a difference between 2/3 and 2/3 etc.

billyg, I had a hunch that my depth chart settings were responsible for allowing Dale to play SF and Sanders PF when both were on the court. I infrequently recognized this occurrence so I assumed the position equivalency was inlfuenced by specific scenerios within the game that only occurred seldomly, so I wasn't very concerned.

I think your most likely correct that the position determination between the two players is completely arbitrary. Although maybe unrelated, I think the in-game position swapping shows that (in this specific case) its more of a rodeo than a direct outcome of engine logic.         

The ticket that I submitted a few days ago is still pending. If anyone was still awaiting a response.

Here is the response to the ticket I submitted:

12/9/2010 1:32 PM

Customer Support

Nathan, I've investigated this more. It is possible for positions to change without players entering or leaving the game, which is what happened in your first example. Sanders came in at SF and Dale at PF at 4:00. They were then swapped at the 3:44 mark when there was a foul. This is happening because the engine sees either option as equal. In a simplified explanation of how it works, it tries to find the lowest combined depth value. So Sanders at SF and Dale at PF would equal 6 (Sanders is depth 3 at SF and Dale is depth 3 at PF). Sanders at PF and Dale at SF would also equal 6 (Sanders is depth 2 at SF and Dale is depth 4 at PF). So those two are interchangeable between SF and PF when they're both in the game.

Just as suspected (good call billyg).

 

well, being a software developer does have its advantages :) 

would you consider responding to the ticket asking if it would be easy to expand the depth chart to 5, which would allow users to customize their depth chart to avoid these kinds of ties? in your case, a trivial scaling like i mentioned above would make it deterministic. but not it in the situation where player A is 2nd on both positions and player B is third on both.

billyg, I will submit that question as soon as I hear back about a ticket I submitted early this morning. I've submitted more tickets in the past week than I did in all my previous seasons, haha.