Posted by ncmusician_7 on 8/26/2010 5:45:00 PM (view original):

Posted by seble on 8/26/2010 4:47:00 PM (view original):
On a related note, I mentioned once that fouls are not being factored in.  Do you guys think it's necessary to do that?

I think they should affect salary a little bit (but have less weight on that attribute than the other ones).  If everything else is equal, a player who fouls once per 15 minutes is worth slightly more than one who fouls once per 10 minutes.

Agree.

Posted by slymonium on 8/26/2010 7:07:00 PM (view original):

Posted by ncmusician_7 on 8/26/2010 5:45:00 PM (view original):

Posted by seble on 8/26/2010 4:47:00 PM (view original):
On a related note, I mentioned once that fouls are not being factored in.  Do you guys think it's necessary to do that?

I think they should affect salary a little bit (but have less weight on that attribute than the other ones).  If everything else is equal, a player who fouls once per 15 minutes is worth slightly more than one who fouls once per 10 minutes.

Agree.

definitely agree, assuming fouls are normalized for pace across eras.

let's put it this way - people dont use Charles Oakley for a reason

These salaries are good.  I would be happy moving forward with these.

Very thoughtful and fruitful inputs, All.  An artist needs to know when to stop dabbing the canvas, and I'd say that, for this release, that time has come.  I like the latest salaries.  No doubt there will be arguable issues and exploits, but let's face it, no what if formula will achieve perfection until Pearl Washington reigns supreme, and frankly, I'm not even sure Einstein could come up with an equation for that.  Meanwhile, I say the time is now ... to RELEASE THE KRAKEN!

holy crap.  These salaries are MUCH better.  I still think you're double dipping with ts% & ft% and it penalizes people in the wrong place, but I can live with this (aside from the greatest defender in nba history, Troy Murphy).

Something I'm mystified by, however... what is Usage Count - and will it be a searchable stat?  Are you changing the way usage works or is this something in the formula we've never seen?  I can't figure it out... 87-88 Barkley has a UC of 47 (26.7% usage%) while 89-90 Mailman has a UC of 6 (32.6% usage%).

Posted by longtallbrad in phase 1:
 

I don't agree that having no defense rating at all would be preferable. Then, as has been noted elsewhere, the wide world of players renowned more for man defense rather than help defense would truly be nerved by the system. So much for Rodman, Moncrief, Michael Cooper, etc. There must be a way to do some common sense quality control of the quantifications.

seble, would you all consider taking suggestions regarding defence ratings that just don't pass the straight face test and exercising judgment to make at least a few manual adjustments accordingly? I understand you can't afford to go very far down the slippery slopes of either second-guessing the system you've established or letting the inmates run the asylum, but some of the ratings (and you've heard us complain about them) are silly enough to detract from satisfaction with the game. You could focus on our top 10 suggestions for D-rating adjustment regarding specific player seasons...something like that.

 

FirstName	LastName	Season	DefRating
Carlos	Boozer	2009-10	96
Troy	Murphy	2008-09	96
Chris	Kaman	2007-08	96
Carlos	Boozer	2006-07	95
Ersan	Ilyasova	2009-10	95
Vladimir	Stepania	2000-01	95
Carlos	Boozer	2007-08	94
David	Lee	2006-07	94
Chris	Dudley	1995-96	94
DeJuan	Blair	2009-10	94
Manu	Ginobili	2006-07	93
Kurt	Thomas	2009-10	93
Robert	Traylor	2002-03	93
Loren	Woods	2005-06	93
David	Lee	2009-10	92
Charlie	Villanueva	2008-09	92
Carlos	Delfino	2007-08	92
Darius	Miles	2008-09	92
David	Lee	2008-09	91
Zach	Randolph	2007-08	91
Carlos	Boozer	2008-09	91
Dirk	Nowitzki	2002-03	90
Antonio	McDyess	2006-07	90
Vlade	Divac	1989-90	90
Greg	Ostertag	1999-00	90
Jermaine	O'Neal	2006-07	89
Jamaal	Magloire	2009-10	89
James	Posey	2008-09	88
Zach	Randolph	2008-09	88
Jahidi	White	2000-01	86

This is a great suggestion by longtall. Seble would you be willing to do this? I have started creating alist of players who I believe are overrated defensivly? Of course this up for debate of who and who shouldnt be added? But maybe lets average the ratings out of what people think?

I believe that is how many times they are being used in leagues.

Yes, Ash, I took that column to be an indicator of how popular they are among NBA WIS users as well.

better but Baylor is still unplayable as is Hayes - oddly Hondo did see some cuts

Posted by ashamael on 8/27/2010 1:53:00 AM (view original):
holy crap.  These salaries are MUCH better.  I still think you're double dipping with ts% & ft% and it penalizes people in the wrong place, but I can live with this (aside from the greatest defender in nba history, Troy Murphy).

Something I'm mystified by, however... what is Usage Count - and will it be a searchable stat?  Are you changing the way usage works or is this something in the formula we've never seen?  I can't figure it out... 87-88 Barkley has a UC of 47 (26.7% usage%) while 89-90 Mailman has a UC of 6 (32.6% usage%).

Its how many teams currently have that player on their squad.

I know we probably want to be done with this but one thing I do see is a guy like Adrian Dantley 80-81 salary go up from $7m to $9.288m.  Now of course he has a high TS% but he's inherently less effective than a Ray Allen becuase he doesn't shoot many threes.  Is there a way to deal with this?  Don't want to raise the salaries any higher but would like to see discounted salaries  (for PG, SG and SF only) for players who don't take many threes.

Posted by ncmusician_7 on 8/26/2010 5:45:00 PM (view original):

Posted by seble on 8/26/2010 4:47:00 PM (view original):
On a related note, I mentioned once that fouls are not being factored in.  Do you guys think it's necessary to do that?

I think they should affect salary a little bit (but have less weight on that attribute than the other ones).  If everything else is equal, a player who fouls once per 15 minutes is worth slightly more than one who fouls once per 10 minutes.

agree

I'm with abrondon at this point.  Time to run the experiment. 

I do agree with ncmusician's point about the ideal relative weighting of fouls, so if that's an easy tweak that doesn't slow down the release significantly, that's great.  But I'd love to throw the lever and start tinkering with rosters.  Hell, I'd even join an open league or two.

Posted by ashamael on 8/27/2010 1:53:00 AM (view original):
holy crap.  These salaries are MUCH better.  I still think you're double dipping with ts% & ft% and it penalizes people in the wrong place, but I can live with this (aside from the greatest defender in nba history, Troy Murphy).

Something I'm mystified by, however... what is Usage Count - and will it be a searchable stat?  Are you changing the way usage works or is this something in the formula we've never seen?  I can't figure it out... 87-88 Barkley has a UC of 47 (26.7% usage%) while 89-90 Mailman has a UC of 6 (32.6% usage%).

I'm not using both FT% and TS%, just TS%.  But TS% includes FT efficiency in it, whereas the old formula did not in any way factor in FT efficiency.

Usage count is the number of teams that currently have drafted that player.  It's only there to identify which players are most popular, and thus most likely underpriced.  Probably not the best name for that value.