Posted by kowboykoop on 8/28/2010 10:29:00 PM (view original):

Posted by hughesjr on 8/28/2010 9:37:00 PM (view original):

Posted by kowboykoop on 8/28/2010 9:22:00 PM (view original):
If a guy is a 45% shooter, than each shot he takes (not considering all other possible variables -- defense, positioning, type of shot, etc.) should have a 45% shot of going in. Simple as that. If he makes one shot, that shouldn't have a damn thing to do with the next shot. If you are seriously telling us that if a player has a hot 1st half, then the engine will FORCE that player's probability of making a shot to go down in the 2nd half until he reaches a predetermined percentage....than that's just absolutely insane and wrong.

If I am missing something...please correct me..because as I understand it now...this game has just gotten significantly worse (unless it's been this way all the time...in which game this game is unplayable).

First off, I do not think there would be any first half and second half arbitrary distinction.

Secondly, it is not as simple as a 45% chance of going in.  There is fatigue, defense, ath/spd of the shooter and defender, +/- rating of the defense, etc.

So, what he is saying he did is put in a "Minimal" feedback so that if a guy misses 10 shots in a row, he will be "Slightly" more likely to hit a shot.  If he makes 10 in a row, he is "Slightly" more likely to miss one.

This should prevent a team shooting 0-20 from 3 or shooting 18-20 from 3 ... we should have more 30%-50% and less 10% and 90% from 3 .etc.

Again ,,, if there is currently too much variability in the current game, then the only way to combat variability is to feedback previous results into the current try.  There is no other way to do it.  Was there too much variability before ... according to everyone and their brother, yes.

Why don't we play a couple of games and see what happens before we declare that the change has made the game unplayable, shall we?

I already said that in my example I was not considering all other factors....but anything that makes a guy a better shooter magically on one shot just because he missed some shots in the beginning of the game is BS. If a guy is a 45% shooter on a wide-open, standard 15 foot jump shot, then every wide-open, standard 15 foot jump shot should have a 45% chance of going in, regardless of previous shots made or missed.

That is your opinion .. and mine is I want some feed back for closer results.

What makes you right and me wrong?

I Very rarely comment on the game engine. For me the greatest strength of the game engine was each event was handled individually. That is how I understood statistics to work. Almost all players and teams will perform better or worse than I feel they should. The oddity would be the player who is dead-on.

The race isn't always won by the fastest horse but that's the way to bet.
Well, I don't want my horse who should win 65% of the time to be "helped" if he's only winning 30%.
Or worse, I don't want the other guys horse to be "helped" when he's been underperforming.

The simulation aspect just got lessened I fear.

Posted by oldresorter on 8/28/2010 10:08:00 PM (view original):
I'm asking a simple ?:

was the change a swag, or based on variance data from HD vs real life -

seble obviously is monitoring this - and has been unwilling to address the question -

hughes - km - cb - a in the b - aren't you concerned that such an easy ? can't / won't be answered? I am not bashing the change, I am really curious.

I will even tell you why, for years I have heard from tarek that he did monitor real life, and that the variance does match real life hence he was unwilling to change, has something changed that now seble is willing, or are we deviating from this course of following real life variance - that seems pretty legitimate doesn't it too you fellows?  Would you not want this question asked, discussed and answered?

I have a pretty good feel for real life d1 stats, and indeed I have always been impressed with how real life our stats have been (at times, certainly we lost that in the slow down era for example), to the point this past season I was able to talk to the then assistant now head coach at GB about certain HD success ratios I use vs ones he used.

This probably is not a huge change, it is more an area of interest to me than concern, you guys seem to be very sensative?

OR honestly I didn't even look at your ?, and my post wasn't directed towards you at all, just a general observation. For example look at the Tea Party movement to use political example of what I said. I know you are cautious about the new engine and rightfully so, and I take all of your opinions seriously, probably as seriously as anybody else on this site. I'm pretty sure you are or used to be a programmer and have been playing this game forever so all of your concerns and opinions are very legitamate.

1. so long as this adjustment is small - a tweak around the edges - then I think it likely will be fine - and we likely never would have noticed it

2.  if it is large, then a major departure from having each event, possession action independent seems bad

3. I can rationalize that this is something like an adjustment to reflect what would happen during a game - good FT shooter is missing too many, coach calls him aside, tells him "kid it is just you and the net, no one else is here, dont look at your girlfriend, dont hear the crowd, just you and your net outside your garage back home....:" or whatever

4. better be small - just enuf to reduce the odds of extremes and NOT enough to make most, many, all performances standardized

Posted by jenningss on 8/28/2010 10:43:00 PM (view original):
I Very rarely comment on the game engine. For me the greatest strength of the game engine was each event was handled individually. That is how I understood statistics to work. Almost all players and teams will perform better or worse than I feel they should. The oddity would be the player who is dead-on.

The race isn't always won by the fastest horse but that's the way to bet.
Well, I don't want my horse who should win 65% of the time to be "helped" if he's only winning 30%.
Or worse, I don't want the other guys horse to be "helped" when he's been underperforming.

The simulation aspect just got lessened I fear.

jennings - it is a little disconcerting - i.e. 'fudging' of random number theory - but if it produces more accurate bell shaped curve HD results vs real life, I am all for it, which is why I keep asking, how indeed was the change arrived at - I'm out of here for the night, I guess we will know how round 1 goes in the morning, or sooner for those who stay up for the nightly results.

Posted by oldresorter on 8/28/2010 10:08:00 PM (view original):
I'm asking a simple ?:

was the change a swag, or based on variance data from HD vs real life -

seble obviously is monitoring this - and has been unwilling to address the question -

hughes - km - cb - a in the b - aren't you concerned that such an easy ? can't / won't be answered? I am not bashing the change, I am really curious.

I will even tell you why, for years I have heard from tarek that he did monitor real life, and that the variance does match real life hence he was unwilling to change, has something changed that now seble is willing, or are we deviating from this course of following real life variance - that seems pretty legitimate doesn't it too you fellows?  Would you not want this question asked, discussed and answered?

I have a pretty good feel for real life d1 stats, and indeed I have always been impressed with how real life our stats have been (at times, certainly we lost that in the slow down era for example), to the point this past season I was able to talk to the then assistant now head coach at GB about certain HD success ratios I use vs ones he used.

This probably is not a huge change, it is more an area of interest to me than concern, you guys seem to be very sensative?

UM. . I have only made one comment in this thread at all, OR. . and I don't know that I would call it 'sensitive"  To whit:

 

"SO performance within a game will tend to regress towards the mean?"
 

ok, what about the stamina change?

I've never had a problem with any change to the game until now.  I understand that they are all attempts to improve the game and give everyone a better experience but this is a real disappointment.  Someone becomes a better shooter by missing his first 8 shots?  Really?  I hate losing to a weaker team just as much as the next guy but I'd rather lose that game than feel like the second half was fixed just to make sure that I didn't.  Maybe it won't be that noticeable but that's not really the point.  I just think that the events need to be independent.  That's just my two cents but like I said earlier, it's a real disappointment.

I still really don't like this.  In reality if a 40% shooter is 2/11 in a game I would expect him to be less than 40% likely to make his next shot, not more.  I'm at least assuming that the percentage odds of any given shot going in are calculated based on the defender and positioning, etc.  What I want to know is what happens when the conditions change.  Do results of previous shots impact the shot under different conditions?  Based on the normalization of the shooting % for the new conditions?  Or just the shooting % so far in spite of the fact that the expected % may have changed?  If a guy has an expected percentage of 20% against one defender and goes 1 for 5 and then faces a weaker defender against whom he should shoot 40%, is he more than 40% likely to hit because he's only shooting 20% on the game?  Even though he's actually performing as expected?

Posted by cal_bears on 8/28/2010 11:18:00 PM (view original):
ok, what about the stamina change?

It sounds to me like the changed a purely linear stamina hit to one where the end result "at tired" is the same (or a bit more), but the curve starts out flat and increases exponentially as you get closer to tired.

It sounds to me like it would mean fatigue is less of an issue at Fairly Fresh and Getting Tired ... but there is an impact at tired and very tired.  Something I wish they would also do in GD.

Since no games have played I am still kind of wait and see, but I have a possible concern as others do with the shooting % thing.  So if I get a real favorable matchup, say someone is starting a frosh with mediocre DEF and a poor IQ, and I pour on the distro to eat the kid up, and my guy goes 7-9 in the first half does that mean he'll go 2-11 in the second half to "even it up"?  

Or. maybe its my 95 ATH 96 SPD 90 DEF A- IQ SF and he's locked down on the opponents best scorer, and holds him to a low % in the first half - now the other guy will get magic gift % to overcome my guy's D?

Posted by antonsirius on 8/28/2010 4:39:00 PM (view original):
Color me very nervous about this.

+1

Early impressions based on 2 games:

The changes are minor.

Shooting %s seem very close to the same.  

Playing while fatigued is not a good idea any longer.

..........if you're 0/6 the defense will probably not guard you quite as closely. you don't get better but the defense adjusts accordingly. if you're 6/6 the defense goes "oh crap" and its tougher to get shots off