forgive me if i don't take everything seble says as gospel. 

Posted by bow2dacowz on 2/17/2011 1:14:00 AM (view original):
youre an idiot.  the title of the thread asked if people noticed any effects of the new engine and i stated my experience and my observation.  if you're SO sensitive to the issue as to think i was speaking on behalf of the entire WIS universe, then that is your perogative.  next time ill be sure to qualify every post with "this is my opinion" or "here is my dissertation and scientific analaysis on why this is unequivocally a fact"

to anyone who is actually NOT a troll, i will throw this out there.  my best D3 team, which in theory should be the most consistent and high performing, lost only 2 players from last years team that shot nearly 56% from the field.  this year's team is shooting 49% through non-conference while attempting a similar # of 3's.   last season those players shot close to 75% from the free throw line.  this season it is at 65% (one of the graduated players was a very good free throw shooter at about 80% for 150 attempts... which accounts for some but certainly not all of the variance).  i suspect the fg% will go up some during conference play because my conference is not very good, nor was it last year...but i think that is mitigated by the fact this team is FAR superior in terms of both perimeter and lp compared to last season.  In theory we should be more dominant offensively than we were last year, but we are obviously significantly less so to the tune of about 7% from the field and 10% from the line, which supposedly wasn't touched at all.

I'm a troll because I don't think you say things that are categorically untrue in a way that represents them as facts.

Yeah, I'm the idiot.  Good point...

Maybe, just maybe, you answered your own question about FG% already, and it will come back up during conference.  Maybe, just maybe, it's a small sample size to this point.  Maybe, just maybe, you're mistaken about your team's per/lp being "far superior" at this point. 

Or, maybe you're right, and maybe seble is lying about the across-world numbers for no reason.  Yeah, I'm the idiot.

bow since your team is shooting a lot lower % what you should also look at is the teams your playing are they also shooting about 7% less, even if your team was shooting 25% less if the teams your playing are shooting the same way then the change is working the way it should. it is a across the board reduction in fg%

Posted by bow2dacowz on 2/17/2011 1:14:00 AM (view original):
youre an idiot.  the title of the thread asked if people noticed any effects of the new engine and i stated my experience and my observation.  if you're SO sensitive to the issue as to think i was speaking on behalf of the entire WIS universe, then that is your perogative.  next time ill be sure to qualify every post with "this is my opinion" or "here is my dissertation and scientific analaysis on why this is unequivocally a fact"

to anyone who is actually NOT a troll, i will throw this out there.  my best D3 team, which in theory should be the most consistent and high performing, lost only 2 players from last years team that shot nearly 56% from the field.  this year's team is shooting 49% through non-conference while attempting a similar # of 3's.   last season those players shot close to 75% from the free throw line.  this season it is at 65% (one of the graduated players was a very good free throw shooter at about 80% for 150 attempts... which accounts for some but certainly not all of the variance).  i suspect the fg% will go up some during conference play because my conference is not very good, nor was it last year...but i think that is mitigated by the fact this team is FAR superior in terms of both perimeter and lp compared to last season.  In theory we should be more dominant offensively than we were last year, but we are obviously significantly less so to the tune of about 7% from the field and 10% from the line, which supposedly wasn't touched at all.

Comment. . .you are comparing your team through non conference play to your other team through the end of the season. . . (I know you mentioned it in passing, only to dismiss it. . but I think you are vastly undervaluing that difference in your post.

Not the same sample. . . for example, looking at ECSU's schedule, if you scheduled your non conference the same last year and the rest of the conference was about the same toughness, I can easily believe that your shooting percentage went up considerably during conference play to get to 56%.  Four games in, its already up to 51.6% Meaning right now, if you take the three percent mentioned you are effectively less than two percent behind last year's final shooting percentage with twelve conference games left to raise it still.


In non con, you have 19,115,62,18,23,33,195, 10, 38, 45

Coming up in conference play. . .  119, 68, 275, 306, 235, 204, 119, 165, 167, 371, 306, 215, 235, 204.

Looking at your upcoming schedule. . quite likely you aren't going to be far under 56% by the time you get to the end of the season>  Heck, you have already risen to .516 just in the four games since conference play started.

Again, if your conference is about the same quality as last season and you scheduled your non conference about the same than comparing your team shooting percentage at the end of non conference play to your total end of season percentages is apples and oranges.  I don't think, looking at your schedule that its going to go up 'some' but 'allot'.

Posted by mullycj on 2/21/2011 9:02:00 PM (view original):
Seble said he "attempted" a 2% reduction which, if true, was a pretty good figure. unfortunately my gut is telling my its more in the 5% range, which again would be overkill back to the old engine again. I havent seen enough games yet to know if this is true or not, but it IS my initial thinking.

He said in the thread about the latest update that it has resulted in a 3% reduction and now mirrors the FG% in real life.

If anything, looking in the world I am in, it might still be a smidge HIGH, not low. (At least if the real world numbers are the basic metric used)

i'm seeing changes since the 2-16 update........that are having a very negative effect upon gameplanning.

my question is whether he's taking into account how most teams run a "minus" defense

Posted by dahsdebater on 2/17/2011 2:29:00 AM (view original):
Seble gave us the numbers...  It was about a 3% change...  That's a fact, not an opinion.  He has access to all the numbers from all the games from all the worlds.  Overall 3% decrease.  Since it was applied to the base rate it shouldn't really impact any one type of player more than others, either.  Maybe you've just been unlucky?  The free throw abilities of existing players weren't touched, so that dropoff is probably just random and unfortunate.

pretty dangerous assumption you are making there, if you ask me (not impacting any one type of player more than others).

Posted by gillispie1 on 2/22/2011 4:17:00 PM (view original):

Posted by dahsdebater on 2/17/2011 2:29:00 AM (view original):
Seble gave us the numbers...  It was about a 3% change...  That's a fact, not an opinion.  He has access to all the numbers from all the games from all the worlds.  Overall 3% decrease.  Since it was applied to the base rate it shouldn't really impact any one type of player more than others, either.  Maybe you've just been unlucky?  The free throw abilities of existing players weren't touched, so that dropoff is probably just random and unfortunate.

pretty dangerous assumption you are making there, if you ask me (not impacting any one type of player more than others).

+1

So I guess the real question is which attributes became more important towards determining a made FG?  In other words, if 1 player saw his FG% drop 8% and another saw it drop just 1%, which attributes does the second guy have more of than the first?  Is it ATH?  Is it SPD?  Is it PER/LP?  Is it BH?  I simply don't have the time to look at enough data to answer that question, but it has one or more of those.

gill, how would you program a base rate such that it WOULDN'T impact all shots equally?  You like to imply that you know how you'd program this game.  Maybe you'd set up a base rate very differently than I would.  But it seems to me that at the very least all 2-point shots and all 3-point shots should be equally impacted (with the possible exception of special circumstances like uncontested/breakaway layups and dunks which may have a separate base rate which may or may not have been changed).

Posted by mduncanhogs on 2/22/2011 8:36:00 PM (view original):
So I guess the real question is which attributes became more important towards determining a made FG?  In other words, if 1 player saw his FG% drop 8% and another saw it drop just 1%, which attributes does the second guy have more of than the first?  Is it ATH?  Is it SPD?  Is it PER/LP?  Is it BH?  I simply don't have the time to look at enough data to answer that question, but it has one or more of those.

You're forgetting the impact of luck.  7% shooting difference over the span of half a season or less is really well within the bounds of normal variance.  If the actual change is actually in the vicinity of 3-4%, as we have been told, it's not at all unlikely that over the small sample period one guy is shooting a few % better than his "true value" and another guy a few % worse...  Also worth pointing out that it's a small enough sample size a few very strong or very weak opposing defenders could significantly impact the shooting percentage a player put up.  Also worth pointing out that these % change numbers assume that the numbers before the change were from a reasonably large sample and accurate (which is probably not true to within a few % given the fluctuation we see from season to season in players whose ratings have basically maxed out) and that we should expect the same %s before and after the update, which would only be true if they were facing the exact same opponents pre- and post- change.  Which is obviously not the case.

Posted by dahsdebater on 2/22/2011 8:45:00 PM (view original):

Posted by mduncanhogs on 2/22/2011 8:36:00 PM (view original):
So I guess the real question is which attributes became more important towards determining a made FG?  In other words, if 1 player saw his FG% drop 8% and another saw it drop just 1%, which attributes does the second guy have more of than the first?  Is it ATH?  Is it SPD?  Is it PER/LP?  Is it BH?  I simply don't have the time to look at enough data to answer that question, but it has one or more of those.

You're forgetting the impact of luck.  7% shooting difference over the span of half a season or less is really well within the bounds of normal variance.  If the actual change is actually in the vicinity of 3-4%, as we have been told, it's not at all unlikely that over the small sample period one guy is shooting a few % better than his "true value" and another guy a few % worse...  Also worth pointing out that it's a small enough sample size a few very strong or very weak opposing defenders could significantly impact the shooting percentage a player put up.  Also worth pointing out that these % change numbers assume that the numbers before the change were from a reasonably large sample and accurate (which is probably not true to within a few % given the fluctuation we see from season to season in players whose ratings have basically maxed out) and that we should expect the same %s before and after the update, which would only be true if they were facing the exact same opponents pre- and post- change.  Which is obviously not the case.

I agree with everything you just said, but for the purposes of my question if we assume that the sample sizes were big enough and all of the other variables were taken away, the importance of which attributes changed to account for the 3% drop?