Posted by beefburglar on 5/16/2014 11:25:00 AM (view original):
Posted by Trentonjoe on 5/16/2014 10:58:00 AM (view original):
Posted by guyo26 on 5/16/2014 10:40:00 AM (view original):
I don't see it as a bonus, I see it as upping the average.
I don't know that spacing or defenders in the area are taken into account in the engine. Well, too an extent that's +/-, but in theory a man playing + and a 3-2 playing + or 0 is the bonus. That brings up the debate of is a 3-2 really man +2 or whatever.
I think I just talked on both sides lol.
(I just realized I quoted the wrong post, this is more in response to beefburglars)
This is what I think. I don't know if this is general consensus but some off it probably is:
1. A 3-2 limits 3FGA. It also decrease the rate they go in.
2. The 2-3 increases 3FGA. It also increases their success rate.
3. The 3-2 has a positive defensive effect against guards who score.
4. The 2-3 has a negative defensive effect for guard scoring.
5. Number 1 may just be an example of #3.
6. A 3-2 (0) and a 2-3 (+2) are about the same. I think there will be less 3FGA's in the 3-2 but the success rate will probably be lower in the 2-3.
7. IIn a nut shell, I think just having the extra defender prevents shots. His defensive metrics influence if the shot goes in.
Thanks Joe. Yeah my thoughts definitely leaned in a lot of those directions.
3 & 7 are definitely the heart of what I was trying to dig at - The concept of a "positive defensive effect against guards who score". Poking around to see how people interpret the weight of the "effect"
The part in 7 about "the extra defender prevents shots". This is probaby part of the answer i was searching for. Somehow I had never really noticed. Makes a lot of sense. And seems to be very true.
i actually completely disagree with this mentality... that the extra defender prevents shots, but his defense metrics influence if the shot goes in. the latter part, yeah, i think that is going to be true regardless. but i don't think the extra defender is just preventing shots, neither from 2 or 3, as much as shifting the scale for fg%/3pt%. so like, for example, a defense value of 5 vs an offense value of 6 (in an incredibly simplistic view of the world - although the engine likely gets to that point eventually, itself), you might see a 2pt% of 53% if you play 3-2, and 47% in the 2-3. whatever the numbers are, i have no idea, but i think even with a worse defender being thrown into the 3 part, whichever it is, within reason, hes going to help defense, in terms of fg% or 3pt%. maybe a walkon would hurt, but even players of the wrong "type" (guard in 2-3 or big in 3-2) seem to help on the %.
i had theorized similarly, but it never seemed to work out like that for my teams. also, when i dug into the data deeply for our oh st team last season (i had a few drinks and wrote a program to facilitate pulling out certain data from games, i still copy paste parts in, its really an ugly hack, but it did let me look at statistics about the game that are not evident just from looking at the stat line - i wouldn't call it "advanced statistics", but still, stuff you can't readily see, to facilitate comparisons), it just didn't look that way. i plan to look at future seasons the same way, with different sets of players it might play out differently, but we had a good variety of 2/3 3/2 and guard/big playing that 3 spot, it was pretty interesting. also, the rebounding impact of the 3/2 vs 2/3 was not what i expected (i expected it to be substantial, it was not, but this is definitely looking at limited sample size, i didn't even put in all the crappy games vs sims, i looked at like maybe low 20s number of games, thats it)