players impacting team mates - passing raising fg% Topic

I've never noticed having a scoring PG  negatively effect my offensive performance useing the triangle in HD.   
2/11/2014 1:49 PM
mygeneration, i was inspired by a thread (ahalusa's compendium of HD offenses, or something like that) to research different basketball offenses and defenses, to see how they related to HD. i think game creator tarek was much like myself, a computer guy first, and a basketball fan, not really a basketball expert. i think he had some idea of the game that he tried to recreate, and in some cases, it reflects reality well... in others, not so much. i think offenses are probably the area that most poorly reflect reality. there are some similarities but i think its more of coincidence, i think they are intentionally blended to avoid "boxing players in" to certain strategies. so really, i believe pretty strongly the attempt to analyze the transitions in a real offense, like you are doing here, and then to apply to HD, is a fruitless effort. also, keep in mind there is no first pass, as there is no passing whatsoever... so trying to consider who gets the first pass and what importance that places on him, IMO, is not going to get you anywhere.

whomever told you triangle in HD just uses three players and ignores the other, is totally wrong. there is a coincidence where in HD only so many highly talented offensive players make your team better - at some point, the incremental offensive stud is wasted. you obviously get the most benefit from the first, but the second and third can be pretty useful, while the fourth (depending on the quality of the first 3) and fifth add much less value. so in most of the offenses, 3 main offenses threats on a lineup is pretty much sufficient (they don't all have to be super stars) - while in motion, 3 main scorers with 2 black holes will suffer, in my experience. so, i think there is something going on with 3 on offense, but the fact that this holds in the triangle is purely coincidental, in my opinion. it holds just as much in flex and fastbreak. you seem to need a sufficient amount of offense from your players, but i don't think it needs to be concentrated in any specific number - it just ends up being the case that focusing on offense in some guys, and not in others, lets you have players better in their respective roles. i don't think concentrating offense in 3 players has any inherent advantage in the triangle, although the indirect benefit of possibly having better defense or rebounding as a result, may still be significant.
2/11/2014 5:15 PM
im not sure how i feel about this comment about not being able to have scoring PGs in the triangle offense. in the old engine, i definitely would disagree, but there might be something funny going on there with the PG now. it might be that there is some consideration where the scorer is affected by their own ratings and the ratings of the rest of the team - by excluding the PG from the "rest of the team", there might be some negative benefit. i very much doubt this effect, if it exists at all, is limited to triangle. you can see the assist numbers vary wildly for the PG based on how much he scores, but generally assist numbers are useless. i don't know why, but i feel if i have a strong bh/pass SG, i have never had trouble with my PG scoring. however, if your PG is the only guy who can run the offense, there might be something there. *might*. i've wondered about it before, but never reached a conclusion, it seemed kind of unlikely on one hand, that such a thing would exist, but on the other hand i could also see the shooter, in the new engine, getting open to some degree based on the skills of his team mates. im not sure why this would hurt the team as a whole, though, i don't know that i've ever seen anything there... the concerns i had were always with the PG himself. but, im really not sure, its an interesting topic, anyway.
2/11/2014 5:22 PM

mygeneration, all I can tell you is that I have tested my theory of low PG distro with multiple teams with many different roster setups, and in every single case, the offense seemed to suffer the higher I set my PG's distro.  You are corrent that in the real world it should not work that way, but this isn't the real world.  I think the engine holds certain abilities in higher regard at certain positions, as in it expects a player playing PG to look and act a certain way, and when he doesn't, it throws the offense off.  From what I can tell, the engine in triangle expects the PG to be a passer, to either the wing or the post, on a certain side of the floor.  When his distro is high and he takes a lot of shots before initiating the offense, he often is ignoring the better scoring option elsewhere and the efficiency of the offense in this game engine suffers.  I have placed absolute stud scorers at PG in a triangle offense facing incredibly weak defenders when facing a M2M defense only to watch their production fall vs. when I have them at SG and see my offense as a whole suffer.  I've done this enough to be fairly certain it isn't just the randomness of the engine that is causing those outcomes, but rather the way the engine works.

I'd love to hear from someone who has a different opinion than me as to why they believe a PG in a triangle can have high distro and the offense still work, because my experience tells me that simply isn't possible.

2/11/2014 5:24 PM
Posted by gillispie1 on 2/11/2014 5:23:00 PM (view original):
im not sure how i feel about this comment about not being able to have scoring PGs in the triangle offense. in the old engine, i definitely would disagree, but there might be something funny going on there with the PG now. it might be that there is some consideration where the scorer is affected by their own ratings and the ratings of the rest of the team - by excluding the PG from the "rest of the team", there might be some negative benefit. i very much doubt this effect, if it exists at all, is limited to triangle. you can see the assist numbers vary wildly for the PG based on how much he scores, but generally assist numbers are useless. i don't know why, but i feel if i have a strong bh/pass SG, i have never had trouble with my PG scoring. however, if your PG is the only guy who can run the offense, there might be something there. *might*. i've wondered about it before, but never reached a conclusion, it seemed kind of unlikely on one hand, that such a thing would exist, but on the other hand i could also see the shooter, in the new engine, getting open to some degree based on the skills of his team mates. im not sure why this would hurt the team as a whole, though, i don't know that i've ever seen anything there... the concerns i had were always with the PG himself. but, im really not sure, its an interesting topic, anyway.

You might be onto something here.  I think in general the triangle works much, much better if you have one primary scoring option outside and one primary scoring option in the post.  When I move my stud scorer to PG to face a weak M2M PG defender, I usually also slide a decent scoring option into the SG spot.  Maybe what I need to be doing in these situations is putting a pass-first guard into the SG slot to see if that impacts the outcome any.  I think there is definitely something going on in the engine with how player abilities impact everyone on the floor.  There seems to very much be a symbiotic relationship among the 5 players on the floor in this engine.  seble has told us that TEAM rebounding ability impacts who gets a rebound in addition to individual matchups, so perhaps the same holds true when looking at passing and the ability of an offense to be efficient?

2/11/2014 5:29 PM
Great discussion, guys.
2/11/2014 5:58 PM
Posted by mduncanhogs on 2/11/2014 5:29:00 PM (view original):
Posted by gillispie1 on 2/11/2014 5:23:00 PM (view original):
im not sure how i feel about this comment about not being able to have scoring PGs in the triangle offense. in the old engine, i definitely would disagree, but there might be something funny going on there with the PG now. it might be that there is some consideration where the scorer is affected by their own ratings and the ratings of the rest of the team - by excluding the PG from the "rest of the team", there might be some negative benefit. i very much doubt this effect, if it exists at all, is limited to triangle. you can see the assist numbers vary wildly for the PG based on how much he scores, but generally assist numbers are useless. i don't know why, but i feel if i have a strong bh/pass SG, i have never had trouble with my PG scoring. however, if your PG is the only guy who can run the offense, there might be something there. *might*. i've wondered about it before, but never reached a conclusion, it seemed kind of unlikely on one hand, that such a thing would exist, but on the other hand i could also see the shooter, in the new engine, getting open to some degree based on the skills of his team mates. im not sure why this would hurt the team as a whole, though, i don't know that i've ever seen anything there... the concerns i had were always with the PG himself. but, im really not sure, its an interesting topic, anyway.

You might be onto something here.  I think in general the triangle works much, much better if you have one primary scoring option outside and one primary scoring option in the post.  When I move my stud scorer to PG to face a weak M2M PG defender, I usually also slide a decent scoring option into the SG spot.  Maybe what I need to be doing in these situations is putting a pass-first guard into the SG slot to see if that impacts the outcome any.  I think there is definitely something going on in the engine with how player abilities impact everyone on the floor.  There seems to very much be a symbiotic relationship among the 5 players on the floor in this engine.  seble has told us that TEAM rebounding ability impacts who gets a rebound in addition to individual matchups, so perhaps the same holds true when looking at passing and the ability of an offense to be efficient?

i would say without question the same holds of passing, well, really in the sense of how a team gets team mates better or worse looks, which plays out in fg%. i think this is something we "know" not think because seble confirmed it *and* the high end coaching community recognizes it. i think it takes both to know anything, but here, i think we have both. now from seble we have good reason to believe the same sort of thing plays into turnovers. 

i think there is something similar in play with offense and defense comparisons. i don't think the underlying mechanism is the same, where some team scoring ability is paired against some team defense ability (just talking man to man - clearly in press and zone, there is team defense). but even against man defense, i definitely lean towards believing its not as simple as one on one - even outside of the openness of the looks from team mates. i often believe there is some sort of defensive prioritization, where a team guards certain players at higher priority than others. i think this explains various effects. players who aren't that good are generally efficient at lower distribution. one could say, its just the lower distribution, but if your guys at the top are not very good, you don't see this same effect, at least in my experience. also, the "senior slump", which isn't talked about much today, but was a hugely discussed topic back in the day - i think this is because as a junior, the guy (typically a guard) is ranked as a lower threat, and the defense is more focused on other guys. as a senior, without much increase in talent, the guy has moved up the priority scale, and now hes going to perform worse. i definitely think randomness plays into it but i don't think its enough to explain it. these reasons are both tough ones to swallow, IMO, because there are clearly other effects that could explain them - lower distro in and of itself could explain low distro guys being more efficient than expected - randomness could explain a decent % of juniors performing lower as seniors. but its just a feeling i have, over time.

also, one other thing that makes me believe this is, sometimes really good players - especially with gaudy lp and per numbers - have just struggled to the point i struggled to blame it on an outlier. recently, coaching with jjwarden @ ohio state, we had this juco senior, so like b+/a- iq, but he was like 90s ath, 70s speed, with like 90 lp and per, and solid bh, and he basically sucked. in my entire career, i've only coached a few really gaudy lp/per players (i don't covet them like some people), i mean both, not just one - and they have always disappointed me. not because they were great and i wanted exceptional - because they straight up played worse than i'd expect if you took away a substantial portion of the lp or per (depending on which i valued less). its too many times now for me not to be seeking another explanation - and one of these times the guy wasn't even my leading scorer, he was one of 5 on one of my all time great d2 championship teams, and i just wanted to pull my hair out, in a time when very little on my best teams surprised me, i could not figure this guy out. he (and the team as a whole) threw me off so bad i eventually set all my distro to 0 and ran uptempo in the final 4 and championship game. i've never done anything like that, ever, its way out of character. it took a long time before i thought i had something that could explain that team - its a combo of diminishing returns of offensive abilities (i had 5 really strong offensive starers, a total waste), and also, i believe that sf in particular was guarded at priority and it just couldn't be made up for. also i think there are various equations for offense, and in those, lp and per both may factor in, but i don't think in any situation, both factor in heavily - which really limits the ability for gaudy lp and per to come through. still, it makes no sense those guys would actually be worse than you'd expect with 50 less lp or 50 less per, unless the defense is prioritizing them. or, really bad luck?

along similar lines, i have seen lots of not-too-impressive guys perform way too efficiently, even factoring in SOS, and i attribute this largely to the other team not considering them a threat. without this defensive prioritization, the only thing that would explain the "reality", i call it that, that changing distro by performance actually yields good results, would be not understanding how to set distro or not understanding what makes players good. frankly, i buy that for lots of coaches, but i think many of us know enough about what makes players good, to be able to set distro better off ratings, than off performance. but consistently i find that fighting the performance to go with what intuitively makes sense off ratings, yields bad results. i think this can only be explained by something situational, that is not intuitive, that we cannot really predict. at least, that's always been my theory. but lately, i've felt more and more the most likely culprit is defensive prioritization.
2/11/2014 6:55 PM (edited)
Completely situational, but both of my championship teams - both of which ran a triangle offense - featured double-digit scorers at PG.  One of them started a 24 PPG guy at the 1.  And the offensive efficiencies of both teams were extremely solid.  I think you can run into trouble if you start a scorer with mediocre passing.  I don't see why that would impact his own FG% UNLESS the engine actually factors a players' own passing rating into his shooting percentage.  That would certainly be the easiest way to do it - have a running value comprised of weighted passing ratings from each player on the floor that weights the expectation value for anyone's shot - but it really isn't that hard to generate a team passing factor based on the other 4 players on the floor at the time.  I really have considered this an open question since the update that started allowing passing ratings to impact FG%.  I hope that a player's passing doesn't impact his own FG%, but it might.  I guess you could make an argument that it's harder to shade a second defender toward the PG if you know that he's a great passer and can punish you for sagging off your own man, but really, I don't like that...

2/11/2014 8:40 PM
Posted by dahsdebater on 2/11/2014 8:40:00 PM (view original):
Completely situational, but both of my championship teams - both of which ran a triangle offense - featured double-digit scorers at PG.  One of them started a 24 PPG guy at the 1.  And the offensive efficiencies of both teams were extremely solid.  I think you can run into trouble if you start a scorer with mediocre passing.  I don't see why that would impact his own FG% UNLESS the engine actually factors a players' own passing rating into his shooting percentage.  That would certainly be the easiest way to do it - have a running value comprised of weighted passing ratings from each player on the floor that weights the expectation value for anyone's shot - but it really isn't that hard to generate a team passing factor based on the other 4 players on the floor at the time.  I really have considered this an open question since the update that started allowing passing ratings to impact FG%.  I hope that a player's passing doesn't impact his own FG%, but it might.  I guess you could make an argument that it's harder to shade a second defender toward the PG if you know that he's a great passer and can punish you for sagging off your own man, but really, I don't like that...

i agree, could go either way... would be nice to know. but i basically agree with your premise that triangle PGs can score a lot without a problem, i haven't played much triangle in the new engine, so glad to hear from someone who has... i think there are a lot of things that can cause a player to underperform, so its hard to isolate what specifically was the trouble for people who have struggled.

i think you are one of the few, maybe only, veteran coach proponents for assists being meaningful. given the decrease in assists for a scoring PG, what is your take on that possibly reducing the efficiency of the rest of the team? also, do you generally have quality bh/pass on your SGs on those high end teams? i think the point you are touching on with the PG passing having an impact on his own scoring is slightly different... i think the question here was if a PG scoring could impact the efficiency of his team mates. 

i think regardless of opinions here, a theory like this is essentially impossible to prove, its at such a level of detail. one concern i always have on issues like this is that the difference in how people play their teams, the players they recruited and the settings they set, has the potential to make the situations different enough to really change how things like this seem to shake out. there are a bunch of ways this team efficiency through passing/iq thing could be implemented, and it seems somewhat unlikely how much your pg shooting would really factor in, but its certainly possible. hard to guess, really.
2/11/2014 10:10 PM (edited)

My experience is that lower distribution for my guards relative to my scorers does help with assists. 

My rationale in assuming the correlation is that my guards are going to assume the majority of the ball handling because their ratings are higher. Since distribution is really treated like plays run for a player (or at least that's how I understand it from the player guide) I'm assuming that even without the ball a play can be run for a player. So my hope and belief is that while plays are being run for my scorers one of my guards still have the ball and are making the final pass to complete the play. 

Not sure if this is completely accurate but I have noticed that usually when my players do get assists it is to one of the players with the highest distribution, with the lower distribution players usually not being associated with an assist when they score. I.e. my PG has been getting what I consider more assists with a lower distribution, he is not assisted as much in the play by play as the high distribution scorers.

2/12/2014 12:37 AM
Posted by mikvitu on 2/12/2014 12:37:00 AM (view original):

My experience is that lower distribution for my guards relative to my scorers does help with assists. 

My rationale in assuming the correlation is that my guards are going to assume the majority of the ball handling because their ratings are higher. Since distribution is really treated like plays run for a player (or at least that's how I understand it from the player guide) I'm assuming that even without the ball a play can be run for a player. So my hope and belief is that while plays are being run for my scorers one of my guards still have the ball and are making the final pass to complete the play. 

Not sure if this is completely accurate but I have noticed that usually when my players do get assists it is to one of the players with the highest distribution, with the lower distribution players usually not being associated with an assist when they score. I.e. my PG has been getting what I consider more assists with a lower distribution, he is not assisted as much in the play by play as the high distribution scorers.

it does help with assists, because scoring lowers assists and assists usually go to guards. however, assists don't really have a direct correlation with success. think of assists as window dressing - they are only assigned with random probability to random players, after the basket has been made or not.

having guards with low distribution, on the whole, is going to hurt, not help, your team. also, don't let this ruin your notion of the game, but there is no passing in this game. the passing rating factors in, in a meaningful way, in the newer version of the sim (it was really screwed up previously). but there is no passing action. however, i do think there are two phases of shooting, or at least, two equations - how open are you, and does the shot go in. i think its the former which relies on the passing and iq of the rest of the team (and also, IMO, the defense and defensive settings, and even the scoring abilities and distribution of other players). so you can sort of think of that as passing. but the sim does not ever explicitly "pass", so it seems to me to be dangerous/counterproductive to over think passing and assists. just think of passing as resulting in less turnovers and better looks for your players, don't try to think of distro and passers or distro and who is the ball handler, those concepts do not seem to be linked (bh and pass will reduce turnovers so a guy with better bh and pass who shots a lot will see less of an increase in turnovers - so the bh and passing ratings can factor in to how you set distro - but dont try to think about links between distro and WHO is passing or ball handling - distro relates to shot taking and thinking about it more than that doesn't seem to get you anywhere, i just don't think the sim is sophisticated enough to get past that - which may be a good thing?)
2/12/2014 1:43 AM
That's depressing, gillispie. It seems something I said once in a cynical moment is true:

This is not a basketball game played on computers; it is a computer game set in a basketball milieu. Or as you might put it: the basketball part of HD is merely window dressing.
2/12/2014 2:12 AM
i think you are one of the few, maybe only, veteran coach proponents for assists being meaningful. given the decrease in assists for a scoring PG, what is your take on that possibly reducing the efficiency of the rest of the team? also, do you generally have quality bh/pass on your SGs on those high end teams? i think the point you are touching on with the PG passing having an impact on his own scoring is slightly different... i think the question here was if a PG scoring could impact the efficiency of his team mates.
I definitely wouldn't go so far as to say that assists are meaningful, but I think you can learn something from them.  I think there is very close to a linear relationship between the percentage of teammates' baskets that are assisted on by a player and the influence he has on teammates' shooting percentages.  Of course there's a lot of very fuzzy math in there, that above all other issues assumes I can estimate reasonable shooting percentages for players in a vacuum.  The 24 PPG guy I referenced above also had nearly 4 assists per game.  I don't have the full stats on my home computer - I'm not sure I have them anywhere anymore, although I had saved them at one point - but IIRC he was assisting on very close to 20% of all teammates' made field goals while he was on the floor.  That is a very solid number.  I don't think he was hurting the rest of the team at all - he had something like 91 BH, 78 pass; that's exceptional for a D3 guard.

As far as your second question, my BH is usually solid amongst my 2 guards, the passing can be spotty.  I'd love for everybody to be able to pass since it does help with team FG%, but you can't have everything you want at the lower levels.  I know the team that had the 13 PPG PG had a 20+ PPG 2-guard who had very good BH but low-40s passing.  The team with the elite scorer at the 1 was starting a sophomore at the 2 who started the year with BH in the 50s and passing in the high 40s but saw near 10 points of growth in each over the course of the season.  He was a defensive stud, not just for a sophomore but for a D3 guard.
2/12/2014 3:53 AM
Posted by MyGeneration on 2/12/2014 2:12:00 AM (view original):
That's depressing, gillispie. It seems something I said once in a cynical moment is true:

This is not a basketball game played on computers; it is a computer game set in a basketball milieu. Or as you might put it: the basketball part of HD is merely window dressing.
the basketball part is definitely not window dressing... its just that certain aspects are more complex and detailed, and others are less. knowing how the sim works behind the scenes to some degree gives us some idea on where to dig deep, and where not to dig deep. i was simply pointing out a couple areas that one should not dig deep. other things are very complex, like what all factors into how many shots a player takes and how many go in. nobody really knows how that all works, we speculate, but there isn't enough reason to be convinced one way or another.

there's a great dilemma in making any simulation. if you make it too complex, with so many inputs, to allow the kind of control and detail the real sport has - its too confusing for the user. if you dumb it down too much, it doesn't behave realistically enough. personally, given the difficulty of this dilemma, i think its one of the areas where HD excels. yeah, offenses are kind of fudged and players don't really pass, but its not like passing has no value - the TSF decision (turnover shot fouled) flattens much of the actual passing into the common outcomes, which is a fairly reasonable way to approach it. thing is, you could simulate 5 passes and the probabilities of each outcome could be the same as it is represented in 1 simple equation. of course, its unlikely its really quite the same, but you always have to juggle the difficulty of balancing a complex system against the richness and detail added by making the system more complex. 

like i said in the post where i mentioned that passing didn't really happen - you shouldn't let this discourage you too much. the sim is still rich and does a good job emulating most aspects of basketball. at first, it threw me for a loop, but i definitely understand why they did what they did, and how they limit the negative impact of not fully simulating certain parts of the game. its hard to walk through all that but if you do, i think you would also conclude they made pretty reasonable decisions there.
2/12/2014 11:23 AM
Posted by gillispie1
 i think you would also conclude they made pretty reasonable decisions there.......



unless your name starts with ett


2/12/2014 12:35 PM
◂ Prev 1234 Next ▸
players impacting team mates - passing raising fg% Topic

Search Criteria

Terms of Use Customer Support Privacy Statement

© 1999-2024 WhatIfSports.com, Inc. All rights reserved. WhatIfSports is a trademark of WhatIfSports.com, Inc. SimLeague, SimMatchup and iSimNow are trademarks or registered trademarks of Electronic Arts, Inc. Used under license. The names of actual companies and products mentioned herein may be the trademarks of their respective owners.