Why I'm not a WP fan

Post #141 » by **floppymoose** » Wed Apr 6, 2011 5:48 pm

Welcome aboard EvanZ. Are you the same one who was posting in the warriors-stumble-on-50-wins thread? That Evanz seemed to be a WP fan. Maybe you are a different evanz, or maybe you changed your mind? Either way, welcome.

EvanZ · Post #142 » by **EvanZ** » Wed Apr 6, 2011 6:32 pm

floppymoose wrote:Welcome aboard EvanZ. Are you the same one who was posting in the warriors-stumble-on-50-wins thread? That Evanz seemed to be a WP fan. Maybe you are a different evanz, or maybe you changed your mind? Either way, welcome.

Yes, same guy. Different mind.

Post #143 » by **floppymoose** » Wed Apr 6, 2011 7:43 pm

That's awesome. I will now celebrate that someone changed their mind about something on the internet.

Idunkon1stdates · Post #144 » by **Idunkon1stdates** » Mon Apr 18, 2011 6:26 am

http://dberri.wordpress.com/2011/04/15/ ... ork-style/

Kevin Love walks away with the em vee pee

http://www.youtube.com/watch?v=yJasNo0An28#t=23s

Post #145 » by **floppymoose** » Mon Apr 18, 2011 7:36 am

I await any Kevin Love trade with glee. <insert evil cackle>

Post #146 » by **Doctor MJ** » Mon Apr 18, 2011 10:57 am

Idunkon1stdates wrote:http://dberri.wordpress.com/2011/04/15/the-regular-season-awards-wow-network-style/

Kevin Love walks away with the em vee pee

http://www.youtube.com/watch?v=yJasNo0An28#t=23s

The first comment is killer:

So without Love, the Wolves would have won -8 games. That is impressive for Love, the Wolves and Wins Produced.

mysticbb · Post #147 » by **mysticbb** » Mon Apr 18, 2011 3:32 pm

Doctor MJ wrote:The first comment is killer:

So without Love, the Wolves would have won -8 games. That is impressive for Love, the Wolves and Wins Produced.

Funny thing is that even with an average replacement the Timberwolves would be at -3 wins. And Berri finds nothing wrong with that. :lol:

More interesting: Humphries replacing Nowitzki's minutes would actually helped the Mavericks by 7.7 wins for this season. Can you imagine the Mavericks build around Humphries? According to WP48 a 64 wins team right away. BUT replacing Kidd with Derrick Rose would actually made them worse by 3 wins.

azuresou1 · Post #148 » by **azuresou1** » Mon Apr 18, 2011 3:47 pm

I would find it very enjoyable if Berri actually managed to GM a team and promptly ran it into the ground. I really wish nothing but the very worst on him.

EvanZ · Post #149 » by **EvanZ** » Mon Apr 18, 2011 4:58 pm

azuresou1 wrote:I would find it very enjoyable if Berri actually managed to GM a team and promptly ran it into the ground. I really wish nothing but the very worst on him.

He would then claim that he was tanking in order to snag a high lottery pick. Then again, he also claims that there is little value in high draft picks. So, I'm not sure how he would wiggle out of that one. :lol:

EvanZ · Post #150 » by **EvanZ** » Mon Apr 18, 2011 5:05 pm

ezPM has Kevin Love with a 16 WARP, and his teammates with 11 WARP. If you add up WARP, it should equal PyWins (in theory). Minnesota had 24 PyWins this season. They actually won 17 games.

Post #151 » by **Doctor MJ** » Mon Apr 18, 2011 10:51 pm

mysticbb wrote:
Doctor MJ wrote:The first comment is killer:

So without Love, the Wolves would have won -8 games. That is impressive for Love, the Wolves and Wins Produced.

Funny thing is that even with an average replacement the Timberwolves would be at -3 wins. And Berri finds nothing wrong with that.

More interesting: Humphries replacing Nowitzki's minutes would actually helped the Mavericks by 7.7 wins for this season. Can you imagine the Mavericks build around Humphries? According to WP48 a 64 wins team right away. BUT replacing Kidd with Derrick Rose would actually made them worse by 3 wins.

Yeah. The fact that we can get into negative wins here isn't particularly damning to me as a metric as long as you acknowledge the flaws in the stat - but of course he doesn't.

That it should occur with the player who is now the wet dream of WP on the way to claiming that a guy on the worst team in the league was the MVP is embarrassing no matter how you look at it.

Post #152 » by **Paydro70** » Tue Apr 19, 2011 6:31 pm

God those comments are embarrassing... people agreeing that the T-Wolves would win zero games if Love were replaced with an average player. Stunning.

mysticbb · Post #153 » by **mysticbb** » Wed Apr 20, 2011 6:23 am

And he even knows that it would need an average of -33 MOV for them. Heck, it is not like we have no games played without Love. Even though the Timberwolves weren't any good, but they at least managed to keep it close in some games and had even the chance to win it in Phoenix. They looked like they "tanked" games away in order to have a greater chance in the draft, something we have to keep in mind.

Overall it is no secret how the Timberwolves played without Love, we have the data, we know that. And they are very much in-line with a 19 wins team when Love wasn't playing. They are at 26 wins pace with Love. That's 7 wins above his replacement on the Timberwolves. We are not talking about a player who gave a team the impact of an nearly 20 wins above average replacement, the value WP is giving him. That is seen NOWHERE. WS48 gives him 6, my rating gives him 9, APM gives him 10, Net gives him 7, seriously, that is a whole lot of different ways to determine that and all are relatively close with 6 to 10 wins above replacement, WP comes up with 20, double the amount.

EvanZ · Post #154 » by **EvanZ** » Wed Apr 20, 2011 1:29 pm

mysticbb wrote:We are not talking about a player who gave a team the impact of an nearly 20 wins above average replacement, the value WP is giving him. That is seen NOWHERE. WS48 gives him 6, my rating gives him 9, APM gives him 10, Net gives him 7, seriously, that is a whole lot of different ways to determine that and all are relatively close with 6 to 10 wins above replacement, WP comes up with 20, double the amount.

I give him 16 WARP and my replacement value is -3.35. What is the replacement value that you use?

EvanZ · Post #155 » by **EvanZ** » Wed Apr 20, 2011 1:35 pm

The adjusted +/- of the most used units without Love are really, really awful, generally much worse than -9.

mysticbb · Post #156 » by **mysticbb** » Wed Apr 20, 2011 2:15 pm

EvanZ wrote:I give him 16 WARP and my replacement value is -3.35. What is the replacement value that you use?

That's the replacement value adjusted for Love's minutes, I guess. So, +5.4 is the value of an average replacement for WS and for my value. That makes 11.4 WS, 14.4 overall according to my rating.

If we take all the units with APM numbers assigned to, we are getting 385 minutes with an APM of -11.9 (just minuted weighted average). That makes it a 14 wins team without Love. We are still getting 26 wins with Love, which again makes 12 wins over replacement.

There is only one rating which thinks Love is clearly the best and most valuable player in the league, WP48. Now, either everything else is not capable of capturing the "true" value of Kevin Love or WP48 is bogus. Well, take your pick.

EvanZ · Post #157 » by **EvanZ** » Wed Apr 20, 2011 2:23 pm

Just wanted to see if our numbers are on the same scale, and it seems they are. I went back and checked and I have Love closer to 15.5 WARP, with his teammates combined for about 8 WARP. I have no reason to believe that the T-Wolves would be anything other than vying for the worst record of all time with Love being replaced by a -3.5 player. They're really bad no matter how you slice it. Still, WP obviously has problems with rebounds and that is inflating Love's value over what it really is in any universe of reality.

Post #158 » by **floppymoose** » Wed Apr 20, 2011 7:00 pm

Is there any difference if you use raw +-? I don't yet trust APM.

mysticbb · Post #159 » by **mysticbb** » Wed Apr 20, 2011 7:13 pm

Well, for "raw+/-" we can use the On- OffCourt ratings. Put those into the pythagorean expectation (exponent 14) and we get a difference of 7.6 wins, which means he gets 5 above replacement for his playing time. Adding the 5.4 of an average player (assuming, the replacement is just average) would give us 10.4 wins.
For APM (only Love's) it would become 10.4, which gives us 6.9 above average for his playing time. So, overall we are getting 12.3 wins. (It would be 13.4 wins with the other methods, comparing how the team plays without Love according to APM with the APM with him)

So, with the whole lot of different methods (6) we are getting between 10.4 and 14.4 wins, in average it is 12.3 wins, about half of what WP gives him.

Well, if I combine my rating (which I devided into a offensive and defensive rating) with RAPM, I get 10.1 wins for Love. He is 24th in that list. The Top5 for this season: James, Howard, Rose, Nowitzki, Wade.

EvanZ · Post #160 » by **EvanZ** » Wed Apr 20, 2011 7:55 pm

mystic, you have your own version of SPM? How does it work?