Mully, OR alias here. way back when tarek ran the game and -5 was a big debate, tarek ran a sim of something like 500 games each with the team vs the same team, all setups the same, except he varied the def level, -5, -4, etc all the way to +5. I can't recall the exact numbers, but he posted a bunch. This is based on an old man's memory so forgive me if I have some of it wrong, but from what I recall the base def level, 0 came out to the NCAA avg more or less. I looked right now this seasons real life stats, first 3p def is 28.5% and 25th is 30.8, that doesn't seem so far off from what we see. Off team leaders first is 41.4 / 25th 38.9, again that seems about right, more or less?
As far as the rest of his findings, from what I recall with each 1 increment of level, defense % changed about 1.2%, or -5 was about 6% worse than zero, and 12% worse than +5. The defense of making the defense that limited, was otherwise, the results of the sims would change too dramatically from real life. Teams with the right setup might win 100-10 for example, which was not Tarek's vision for the game.
Also, I'm pretty sure he reported wins and losses, and the best defense was zero, but the difference in wins and losses was minimal all the way to +/- 5. I'm testing my memory, but somewhere in the 15-20 range difference, i.e. you won 3-4% fewer games when playing -5 or +5 vs the standard defense in a standard game.
I hope nobody can dig out the old post, as I'm sure most of these memories are off a bit. Actually, If anyone can dig it up, it would be immensely helpful to the community, and it would not hurt my feelings one bit to see the actual correct information. I'm guessing the post is ten years old, no idea the title?
I have no idea if seble runs sims to verify the sim turns out real life accuracy, but IMO most of what I see in terms of box scores and deviations seem about right, better than other sims I've used, seen, tried.
4/9/2014 3:02 PM (edited)