Adjust the game engine so that defensive positioning actually means something. As it is, it affects the # of 3pt attempts by the opponent and that's about it.

One of the changes you made when you came in was to nerf the affect of positioning and it really sucks. Minus defenses vs teams that don't shoot any threes are worthless, and that's not very realistic.

Posted by mullycj on 10/22/2015 12:25:00 PM (view original):
Adjust the game engine so that defensive positioning actually means something. As it is, it affects the # of 3pt attempts by the opponent and that's about it.

One of the changes you made when you came in was to nerf the affect of positioning and it really sucks. Minus defenses vs teams that don't shoot any threes are worthless, and that's not very realistic.

+1

In favor of realism, I agree. 

Posted by mullycj on 10/22/2015 12:25:00 PM (view original):
Adjust the game engine so that defensive positioning actually means something. As it is, it affects the # of 3pt attempts by the opponent and that's about it.

One of the changes you made when you came in was to nerf the affect of positioning and it really sucks. Minus defenses vs teams that don't shoot any threes are worthless, and that's not very realistic.

+1

yes there should be a noticeble difference between  -3 & 0 but right now I dont see much that I definilty say that positioning was a determing factor in the outcome.  Maybe if they just doubled the current +/- deviations then that might be a place to start?

Agree. I don't see much of a strategy other than 3's. 

Mully two comments.  First off, didn't tarek once run 500 sims at each level for the same two evenly matched teams?  I tried to google it too see if I could find the results but could not find it.

Second, I one time played a season with all ten of my teams set to -4.  The wins and losses didn't change real much, if I recall I got a few more boards, gave up a few more threes and at a higher %, maybe fouled a little bit more, defended two a little bit better.  The % change from a my expected defense was in the 4% range - i.e. I defended the three 4% worse, say from 32% to 36% (which one could say was really a 12% change or 36/32*100).

I did decide I lost a few games as result, and quit the experiment b4 the national tournament for all my teams, my guess is vs the high end of d1 teams, the change would even be greater, but not that great, say from 32% to 40%.  This was very much in line with what I recall seeing tarek report back to us when he actually ran the experiment.

I must admit, if this is about the change between 0 and -4 today, lower % on 2's, higher on 3's, fewer attempted 2's more attempted 3's, 1-2 extraboards, I'm ok with it.  I don't want -4 to cause on averge 20% shooting from 2 and 66% shooting from 3 for example, as I want realistic box scores, and no games in real college basketball between 2 even teams results in 20% shooting from 2 or 66% from 3, based on the defensive settings.

I might consider making a more limited, maybe more controlled experiement with some of my teams if anyone is interested.  It does take a little guesswork to determine the real results, as one really doesn't know what the results would be if I did play the 'right' defense instead of -4.

What would you or anyone else want to see in such an experment?

i think the ideal experiment to run would be to run it in conf against your 10 divisional games, playing a -4 for half home/away games and something else (maybe -4/+4) for the other half (playing each setting in one game vs a divisional rival). this would have to be done across like 10 worlds or 10 seasons to get enough info to be meaningful though. would be a pain in the but to aggregate all the data too.

Posted by gillispie1 on 10/26/2015 12:16:00 PM (view original):
i think the ideal experiment to run would be to run it in conf against your 10 divisional games, playing a -4 for half home/away games and something else (maybe -4/+4) for the other half (playing each setting in one game vs a divisional rival). this would have to be done across like 10 worlds or 10 seasons to get enough info to be meaningful though. would be a pain in the but to aggregate all the data too.

nice suggestion gil  I might fool around with it a bit, prob just with a few teams.  won't announce it or anything, too easy for oppoents to game it.  In some ways, it might be best done against 10 sims all on the road, teams that play the same off and def and are somewhat equal?

A ? for you gil, what effect in terms of 1000 or 100,000 sims should + or - 4 have on shooting defense and other things like rebounds and such, vs 0 in your mind, same team playing all 300,000 games, all else equal?  I mean some of the fix could be if you play -4 and the other team still shoots all 2's, they make more TO's along with shooting crappy.  I don't think that is in the formula, at all, or only to a limited amount?

This should be something SEBLE should be doing.