Does anyone have a percentage chart on the effects of a 2-3 zone and 3-2 at each setting +5 to -5, for instance what the shooting percentages from 3 pointers

What would a 3-2 zone -4 compare to with a 2-3 zone? Does a 3-2 zone have a powerful impact on stopping the 3 when playing a negative defense. I've seen many instances where this is the case.

If you play a 3-2 zone, -4 would it be effective at stopping a mid range perimeter game?

Also, if you go with a shot blocker for your center in a 2-3, would it make sense to go with 2 shot blockers in a 3-2 or does the shot blocking ability have less of an impact in a 3-2?

Gillispie, if you have time, I'd be interested in hearing your thoughts on this.

i haven't played a ton of zone, and its definitely by far my least comfortable defense, but i can at least offer some insight.

i don't have a chart or anything, but that +/- defensive setting is definitely huge (generally speaking - for any defense). it has two effects, it shifts shot selection, as well as shifting attempts (so a +3 significantly reduces both 3pta and the % shot). i think the same goes for 2-3 and 3-2, except i feel like for the 2-3 and 3-2, the shot selection part is more pronounced. so, when you run say a negative 3-2, and suppose you are about equal in the 3pt% defense you are putting out to m2m - i think the 3-2 is a little better in preventing 3pt shots. conversely, a positive 2-3 with a similar level of 3pt% defense would likely give up more 3pta.

as a result, i tend to play my 2-3 and 3-2 largely around the 3pta of my opponent, when i have the luxury of switching (some teams just aren't built for it - it primarily comes down to the 3, who is basically calculated as a guard in a 3-2, and a forward in a 2-3).

a consequence of this is something like a 3-2 -2, which is a pretty common thing i play (i also see it pretty frequently among high end zone coaches), is a REALLY good defense when it simply comes to shots. it limits 3pta fairly effectively, while still doing a good job on limiting 2pt% and 3pt%. on the whole, by those 3 attributes, no other scheme can really match it. however, you pay a very significant penalty in terms of rebounding - 3-2 is without a doubt the worst rebounding scheme. even a 3-2 with a heavy negative isn't very good rebounding wise.

i do think a 3-2 -4 would be effective for a mid range perimeter game - but i also think mid range perimeter shots are sort of a wash, when it comes to all this +/- and 3-2 vs 2-3 stuff - its really more about paint vs 3pt shots, that is where the effects are pronounced. there is some effect on mid range shots. the game's shot function doesn't have the location of the shooter, but it has distance - and i would basically think of a real long range 2 being impacted by +/- and 3-2/2-3 similarly to 3s, just a little less in quantity. on the other end, short range jumpers, i think actually get an inverse effect - basically, they are impacted similar to shots in the paint, just with a lesser magnitude. how this all shakes out on mid range shots as a class, becomes fairly nebulous.

the 2-3 is without question the best 2pt% defense in the game, and 3-2 is without question the best 3pt% and 3pta limiting defense in the game - when you only care about those singular attributes. you just pay a lot in rebounding for that 3-2, and a lot in 3pta and 3pt% from that 2-3. i think 2-3 is very slightly worse than man on rebounding, but i can't swear to it. one benefit of any zone scheme is that you foul less than anything else. you also get more blocks and fewer steals.

on the sb question, my take on the 2-3 is that shot blocking is incredibly valuable for the center - like, roughly on par with defense. it could even be a tad higher (my best guess is a tad lower, but close). this was a shock for me, because in press, i would basically ignore sb entirely (its not that sb is useless in press - granted, its at most like a third of defense, so its way less - but sb is also tightly correlated with reb (both are tightly correlated to height), so if you really care about reb, like me, you'd end up with enough sb without even trying).

anyway, in the 3-2, i don't really know how exactly this translates, i just don't have the experience. however, from a theoretical standpoint, one would expect the center in a 2-3 to value blk more than a pf/c in 3-2, just because, that's sort of how everything in this game works, you get this spectrum from the 1-5 of some sort. i do feel like blk is generally quite important for the pf/c in the 3-2, maybe like .75 of defense instead of almost on par. you see a real difference in blocks for these players - and i think that translates to team blocks and overall possessions. in short - i definitely consider blk to be a core rating for the 4/5 in a 3-2.

not to digress too badly, but many individual stats in this game are half window dressing (or all, like assists). rebounding, turnovers - things like that are half window dressing - the team total is legit, but the individual totals probably are not (rebounding for sure is not). it makes it real hard to figure stuff out, because the individual stats become highly misleading - but i think individual blocks are pretty legit. regardless, i see a very significant impact on team blocks, from blocking at the 4/5, and even at the 3 in a 2-3, and blocks obviously are pretty valuable from a possessions standpoint (sadly, less than steals - a steal gives you the ball all the time - a block, its more like a coin flip from there, who gets the shot. blocking does also impact fg%, so its not just about blocks, but still).

I play a 3-2 (-2) pretty often. I agree with Billy on most of his analysis.

the one thing I would add is as the zone reduces the chance of a foul, playing extreme - defenses becomes easier to stomach. Even with a base 3-2 team I have no problem dropping to a 2-3 (-5) against a Dickinson style team.

random unrelated note - a few posts over the past few months (by others) have made me question how long i waste on these boards, typing these long responses. i stop and start and use this as a distraction between other stuff i do (or the reverse) so much, i've been struggling to get a sense. i've been paying some attention, this one here was a 15-20m post, counting an after-post review and edit. so, quite a lot, but also not that bad?

Posted by gillispie1 on 1/24/2020 9:24:00 PM (view original):
random unrelated note - a few posts over the past few months (by others) have made me question how long i waste on these boards, typing these long responses. i stop and start and use this as a distraction between other stuff i do (or the reverse) so much, i've been struggling to get a sense. i've been paying some attention, this one here was a 15-20m post, counting an after-post review and edit. so, quite a lot, but also not that bad?

You do you, man. Many of us, myself included, have learned a lot from you. Sometimes I'll put a pot of coffee on and make some popcorn before sitting down to see what Uncle Gil has to teach today.

I'm only a little kidding.

Posted by pallas on 1/25/2020 12:31:00 AM (view original):

Posted by gillispie1 on 1/24/2020 9:24:00 PM (view original):
random unrelated note - a few posts over the past few months (by others) have made me question how long i waste on these boards, typing these long responses. i stop and start and use this as a distraction between other stuff i do (or the reverse) so much, i've been struggling to get a sense. i've been paying some attention, this one here was a 15-20m post, counting an after-post review and edit. so, quite a lot, but also not that bad?

You do you, man. Many of us, myself included, have learned a lot from you. Sometimes I'll put a pot of coffee on and make some popcorn before sitting down to see what Uncle Gil has to teach today.

I'm only a little kidding.

Agreed! Be yourself! No need to re-think things because of others comments. When I post here, it's 50% real thoughts on HD, and 50% poking the bear. Whoever that bear may be at the time. But we're all just characters on the internet. Who really cares what anyone thinks? As long as your input to HD is beneficial, and how you feel about the topic at hand, here's the place for it! Whether it's 2 words or 2k words..... someone is reading it!

And sorry for going off topic, I do not have zone data. But I think this is an important question. I don't see a lot of successful teams that play a 2-3 zone mainly. I feel it's the least played defense in HD. Would be nice to have more info/stats/data on that. I've tried 2-3 and failed. So I gave up on it

Posted by gillispie1 on 1/24/2020 9:24:00 PM (view original):
random unrelated note - a few posts over the past few months (by others) have made me question how long i waste on these boards, typing these long responses. i stop and start and use this as a distraction between other stuff i do (or the reverse) so much, i've been struggling to get a sense. i've been paying some attention, this one here was a 15-20m post, counting an after-post review and edit. so, quite a lot, but also not that bad?

I really appreciate your time and insight. Thank you.

On one occasion., I was worried you were going to give that same 15 to 20 minute response to Fake Wardo, not realizing he was a troll.

I've been playing this game since 2004 and one thing I'm quite confused about is what is cosmetic in the box score and pbp. In the pbp, if you're scouting a team you're playing in your next game, if you see they scored lots of easy baskets, layups etc, should we assumed they actually scored those points in the paint or again is it just cosmetic and we don't actually know, all we know is they scored a basket that wasn't a 3?

Posted by thewizard17 on 1/25/2020 2:01:00 PM (view original):
I've been playing this game since 2004 and one thing I'm quite confused about is what is cosmetic in the box score and pbp. In the pbp, if you're scouting a team you're playing in your next game, if you see they scored lots of easy baskets, layups etc, should we assumed they actually scored those points in the paint or again is it just cosmetic and we don't actually know, all we know is they scored a basket that wasn't a 3?

good question - i definitely don't know all the answers, if you were to pose this question for every number on the box score or every outcome in the pbp.

however, i will say this - most of the team stats are legitimate, with the major exception being assists, which are meaningless garbage. the individual stats are way more suspect, really the only stuff i know is 100% rock solid on the individuals is shooting - so that is the general rule of thumb i'd apply - team stats are good, individual shooting #s are good, other individual #s range from suspect to window dressing.

for this specific question - i'm not exactly sure. i suspect the part about points in the paint, or a long jumper, both the stats and the pbp, i think that stuff is probably legit - the shooting function takes a distance from the basket, so every shot, while not taken from a specific point on the floor, is taken from a specific distance. i assume then, that the pbp notes and box score stuff that would reasonably be based on distance, is (is based on distance - as opposed to the distance being ignored and random BS being made up).

now, the part about it being a wide open shot and whatnot - im really not sure, but guessing its garbage. old admin caused some uproar once with a comment like, the pbp was just meant to be fun, its basically generated after the fact just to give coaches something fun to look at. i think people were like dude wtf, this is total bullshit, how could you feed us such a pile of crap?? but it was sort of clarified the pbp wasn't just like totally made up where the only correlation with the actual game was score - its just that the pbp is based on limited underlying data about how the game went, and the rest is filled in.

so, where's the line of what is based on something real, and what is filled in? hard to say. the engine very possibly computes shooting in 2 steps (2 dice rolls), how open is the guy, then actually taking the shot. in that case, it would be reasonable for the engine to say stuff like 'X makes the wide open 3'. but it is also is very possible this is just arbitrary BS. in general, i assume anything that is strictly an outcome, is going to be in line - like the state of fatigue, which guy took a shot and if it actually went in, who got a rebound - but i think the rest should mostly be assumed to be garbage. especially if it doesn't show on the box score (like, if the pbp is in theory providing info beyond what is then aggregated and put in the box score), you should probably assume its garbage (so 'wide open looks' would be garbage, but distance from basket could very possibly be legit - but also could go either way).

When running doubleteams, I always assumed a "wide open shot" was the result of one player left wide open while another was doubleteamed. I would then compare the number of those occurrences with the number of effective doubleteam plays. Not sure whether a wide open shot occurring without doubleteams is the result of a speed or athleticism mismatch.Or if it is just purely cosmetic.

" Not sure whether a wide open shot occurring without doubleteams is the result of a speed or athleticism mismatch."

I just looked at three game logs where I didn't double anyone. There were no "wide open" shots.

I don't believe assists are "purely window dressing".

Last season my Tennessee St team had the highest "pass rating" in the nation. We averaged over 20 ast/game playing normal tempo and had assists on over 70% of our shots.

Posted by iamjoeyd on 1/27/2020 9:31:00 PM (view original):
I don't believe assists are "purely window dressing".

Last season my Tennessee St team had the highest "pass rating" in the nation. We averaged over 20 ast/game playing normal tempo and had assists on over 70% of our shots.

there's a lot of stuff about how the engine works where we all throw out models, maybe its this, maybe its that - but its not altogether that often we sort of know the actual fact of the matter, of how it actually works. so usually we are just sharing opinions of what the facts may be - but in this case, i think we actually know.

first, let me quickly address what i mean by 'pure window dressing'. the most obvious case of pure window dressing is a value that is uncorrelated with anything else in the game. let's suppose recruits came in with a minutes played per game stat (i can't remember, do we even still get stats on players? oh i found them - bottom right corner of the recruit page). oh, so there even is a minutes per game. let's assume that is based on nothing - its pure RNG (random number generator), its not based on the recruit's stamina or work ethic or skills of any kind. that would be pure window dressing type A - complete random bullshit.

however, there is also pure window dressing type B, which is just as much, in my opinion, pure window dressing. in those recruit stats, there is also a 3pt%. we think (we could be wrong, but let's assume we are right) that this is somewhat based on the ratings of the player. yet, the HD community has basically considered all the recruit stats to be pure window dressing since the beginning of time, except for gpa and ft% (now we get a FT rating display too - that didn't used to exist). the reason for this is, even though the item in question (freshman recruit's previous season 3pt%) is correlated with something that matters (best guess is its PER plus some RNG) - because we get no additional information about that thing that matters, its window dressing. that is, we already know what the PER rating is, if the previous season 3pt% is some mapping of PER rating plus some RNG, its not giving us any info about PER - we already know PER. if we didn't know PER, then the 3pt% wouldn't be window dressing. so, window dressing type B is window dressing that is correlated with something real, but is nonetheless completely useless - because it gives us no new information.

assists are window dressing type B. they are correlated loosely with stuff that matters, but they provide no information about anything. the reason we know this is as follows - for most of the existence of HD, a player's bh and passing (and anything else) has absolutely no impact on the shooting % of another guy on the court. if you were a big man with whatever ratings, and you went to take a shot against a roughly even defender, and as a result your odds of making this 2 is 50% - then that was the whole story - it was 50% with a walkon pg, it was 50% with the best pg ever. the game just randomly decided to give assists or not, after made shots.

this random factor isn't 100% random - it basically wanted to make player stats look real - so it randomly picks in a way that makes sense, it basically gives guards more assists and guys with more bh/pass more assists. but neither individual assists nor team assists tell you anything about anything (at least, not anything you don't already know). specifically, they tell you nothing about the impact your guys by, whatever, assisting?

now, in today's game a small factor does exist whereby a great passer can get a team mate a more open look. however, this was added fairly superficially, long, long after the sim engine was in place. i am positive assists were not updated at that time. could they have been updated since then - i severely doubt it, but you know, its not impossible? but even if it was updated, its very unlikely the update was comprehensive enough to make assist stats really matter - assist stats would really matter if 1) 'assisting' was a major part of the game, like it is in real life, and 2) assists were based on that 'assisting'. but neither one of these is true.

Posted by gillispie1 on 1/28/2020 8:18:00 PM (view original):

Posted by iamjoeyd on 1/27/2020 9:31:00 PM (view original):
I don't believe assists are "purely window dressing".

Last season my Tennessee St team had the highest "pass rating" in the nation. We averaged over 20 ast/game playing normal tempo and had assists on over 70% of our shots.

there's a lot of stuff about how the engine works where we all throw out models, maybe its this, maybe its that - but its not altogether that often we sort of know the actual fact of the matter, of how it actually works. so usually we are just sharing opinions of what the facts may be - but in this case, i think we actually know.

first, let me quickly address what i mean by 'pure window dressing'. the most obvious case of pure window dressing is a value that is uncorrelated with anything else in the game. let's suppose recruits came in with a minutes played per game stat (i can't remember, do we even still get stats on players? oh i found them - bottom right corner of the recruit page). oh, so there even is a minutes per game. let's assume that is based on nothing - its pure RNG (random number generator), its not based on the recruit's stamina or work ethic or skills of any kind. that would be pure window dressing type A - complete random bullshit.

however, there is also pure window dressing type B, which is just as much, in my opinion, pure window dressing. in those recruit stats, there is also a 3pt%. we think (we could be wrong, but let's assume we are right) that this is somewhat based on the ratings of the player. yet, the HD community has basically considered all the recruit stats to be pure window dressing since the beginning of time, except for gpa and ft% (now we get a FT rating display too - that didn't used to exist). the reason for this is, even though the item in question (freshman recruit's previous season 3pt%) is correlated with something that matters (best guess is its PER plus some RNG) - because we get no additional information about that thing that matters, its window dressing. that is, we already know what the PER rating is, if the previous season 3pt% is some mapping of PER rating plus some RNG, its not giving us any info about PER - we already know PER. if we didn't know PER, then the 3pt% wouldn't be window dressing. so, window dressing type B is window dressing that is correlated with something real, but is nonetheless completely useless - because it gives us no new information.

assists are window dressing type B. they are correlated loosely with stuff that matters, but they provide no information about anything. the reason we know this is as follows - for most of the existence of HD, a player's bh and passing (and anything else) has absolutely no impact on the shooting % of another guy on the court. if you were a big man with whatever ratings, and you went to take a shot against a roughly even defender, and as a result your odds of making this 2 is 50% - then that was the whole story - it was 50% with a walkon pg, it was 50% with the best pg ever. the game just randomly decided to give assists or not, after made shots.

this random factor isn't 100% random - it basically wanted to make player stats look real - so it randomly picks in a way that makes sense, it basically gives guards more assists and guys with more bh/pass more assists. but neither individual assists nor team assists tell you anything about anything (at least, not anything you don't already know). specifically, they tell you nothing about the impact your guys by, whatever, assisting?

now, in today's game a small factor does exist whereby a great passer can get a team mate a more open look. however, this was added fairly superficially, long, long after the sim engine was in place. i am positive assists were not updated at that time. could they have been updated since then - i severely doubt it, but you know, its not impossible? but even if it was updated, its very unlikely the update was comprehensive enough to make assist stats really matter - assist stats would really matter if 1) 'assisting' was a major part of the game, like it is in real life, and 2) assists were based on that 'assisting'. but neither one of these is true.

I have no clue which way I am facing after being spun around in circles so many times hahaha. I don't agree that assists are just window dressing either. But I DO believe I understand what gil is trying to say!

In short..... if you ran a box score and take away every statistical assist in that column, everything else could remain the exact same, and the game would've played out the same. Or vice versa, if that box score originally had 25 FGs made for each team, you could add in that assists happened on all 25 FGs for both teams, and again the outcome would remain the same.

Taking away a rebound can't work that way because that would change possession count. Taking away a FG can't happen because it would change the score. But taking away an assist (or all of them) doesn't change the game.

Simplified?

"the game just randomly decided to give assists or not, after made shots."

I'm trying to uunderstand how you're saying this works. So you're saying that a team with bad passing would have the same number of assists (or maybe assists/FGA) as a team with awesome passing?

Where does passing come into the equation? How does the game determine WHO passed the ball. Or is it a cumulative passing "score" that is a weighted average by position?