ESPN: Real +/-
Old 04-07-2014, 12:30 PM   #1 (permalink)
is praying Ross makes us forget Drummomd so people stop whining

Senior Member
 
jeffb's Avatar
 
Join Date: May 2008
Location: YO MAMMA
Posts: 73,225
Representing:
Raptors ESPN: Real +/-

NBA Real Plus-Minus - National Basketball Association - ESPN

12. Amir Johnson

32. Kyle Lowry

Quote:
Drawing on advanced statistical modeling techniques (and the analytical wizardry of RPM developer Jeremias Engelmann, formerly of the Phoenix Suns), the metric isolates the unique plus-minus impact of each NBA player by adjusting for the effects of each teammate, opposing player and coach.

The RPM model sifts through more than 230,000 possessions each NBA season to tease apart the "real" plus-minus effects attributable to each player, employing techniques similar to those used by scientific researchers when they need to model the effects of numerous variables at the same time.

RPM estimates how many points each player adds or subtracts, on average, to his team's net scoring margin for each 100 possessions played. The RPM model also yields separate ratings for the player's impact on both ends of the court: offensive RPM (ORPM) and defensive RPM (DRPM).
NBA: Introducing Real Plus-Minus - ESPN
jeffb is online now   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 12:40 PM   #2 (permalink)
doesn't care where LeBron plays. Or Melo.

OY!!
 
niggles's Avatar
 
Join Date: Dec 2012
Location: Loch Ness
Posts: 2,741
Representing:
Default

Show's how impressive the team was in winning the last 2 games without these 2 in the line-up.
niggles is offline   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 12:54 PM   #3 (permalink)
whatever

Senior Member
 
Join Date: Dec 2007
Location: toronto
Posts: 9,280
Representing:
Default

have a feeling many Raptors fans will HATE this new stat ...
moremilk is offline   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 12:56 PM   #4 (permalink)
doesn't care where LeBron plays. Or Melo.

OY!!
 
niggles's Avatar
 
Join Date: Dec 2012
Location: Loch Ness
Posts: 2,741
Representing:
Default

Quote:
Originally Posted by moremilk View Post
have a feeling many Raptors fans will HATE this new stat ...
Why?
niggles is offline   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 12:59 PM   #5 (permalink)
whatever

Senior Member
 
Join Date: Dec 2007
Location: toronto
Posts: 9,280
Representing:
Default

derozan is 142 ....
moremilk is offline   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 01:04 PM   #6 (permalink)
LX
with pink peppercorns

In the Paint


 
LX's Avatar
 
Join Date: Dec 2007
Location: Toronto
Posts: 27,655
Representing:
Default

I'll hate unless someone can explain it. Seems like a bunch of numbers put into a blender. Yawn to the power of 100.
LX is online now   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 01:08 PM   #7 (permalink)
whatever

Senior Member
 
Join Date: Dec 2007
Location: toronto
Posts: 9,280
Representing:
Default

doubt anybody can explain it, seems like a lot of advanced math went into that

like clarke used to day, if it's too advanced, it will look like magic ...
moremilk is offline   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 01:15 PM   #8 (permalink)
is pounding the rock!

Senior Member
 
MBailey85's Avatar
 
Join Date: Feb 2012
Location: Toronto,Ontario
Posts: 2,377
Representing:
Default

This was created to promote LeBron for MVP over KD.......................j/k

I don't know what to make of this but as long as the Raptors keep winning and advance past the first round,I'll be happy
MBailey85 is offline   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 01:19 PM   #9 (permalink)
LX
with pink peppercorns

In the Paint


 
LX's Avatar
 
Join Date: Dec 2007
Location: Toronto
Posts: 27,655
Representing:
Default

Quote:
Originally Posted by moremilk View Post
doubt anybody can explain it, seems like a lot of advanced math went into that

like clarke used to day, if it's too advanced, it will look like magic ...
I think people and computers strained themselves to make +/- a useful metric. When you start with a bunch of problems with a stat, it probably isn't going to end well.
LX is online now   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 01:27 PM   #10 (permalink)
sticking to his guns... is stuck to his guns...

Senior Member

 
DanH's Avatar
 
Join Date: Jul 2010
Location: Montreal, QC
Posts: 4,126
Representing:
Default

Quote:
Originally Posted by LX View Post
I think people and computers strained themselves to make +/- a useful metric. When you start with a bunch of problems with a stat, it probably isn't going to end well.
Sure, but if you start with a stat that has primarily one problem, then remove that problem, it's probably a good stat.

RPM is just a more heavily iterated version of APM or RAPM - statistical methods of filtering out the impact of opponents and teammates (and coaches, apparently) on a single player's plus minus.

It's no different than the algorithms that went into weighting the different stats for PER, WS, WP, etc, except without the somewhat arbitrary values assigned to those stats, so if you lend any credence at all to those catch-alls, you should have no issue with this one.

And no surprise that once again, you see solid contributors like Amir, Lowry, Fields, Patterson rise to the top.
DanH is online now   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 01:28 PM   #11 (permalink)
whatever

Senior Member
 
Join Date: Dec 2007
Location: toronto
Posts: 9,280
Representing:
Default

Quote:
Drawing on advanced statistical modeling techniques (and the analytical wizardry of RPM developer Jeremias Engelmann, formerly of the Phoenix Suns), the metric isolates the unique plus-minus impact of each NBA player by adjusting for the effects of each teammate, opposing player and coach.

The RPM model sifts through more than 230,000 possessions each NBA season to tease apart the "real" plus-minus effects attributable to each player, employing techniques similar to those used by scientific researchers when they need to model the effects of numerous variables at the same time.

RPM estimates how many points each player adds or subtracts, on average, to his team's net scoring margin for each 100 possessions played. The RPM model also yields separate ratings for the player's impact on both ends of the court: offensive RPM (ORPM) and defensive RPM (DRPM).
based on this description, this was months if not years in the making. Like any advanced statistical modelling, the proof will be in how well it matches reality. Of course, unlike most phenomenons, the one modelled by this can't really do that. With weather and economy and even global warming, all you have to do is wait and see how it goes. Since there will never be any way to rank nba players with certainty, I guess people will love or hate this model based on how much it matches their own perception.

Which is why I said many raptor fans will hate it ...
moremilk is offline   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 01:33 PM   #12 (permalink)
whatever

Senior Member
 
Join Date: Dec 2007
Location: toronto
Posts: 9,280
Representing:
Default

The real issue though is how other GMs will treat this. It makes less difference if it's good or bad, what matters is how much faith will GMs put on it. Because you can use it to your advantage in trades. If it more or less matches the other advanced stats DanH mentioned, it will make no difference.

Still, we can assume that teams like Dallas, Houston, Philly or Boston etc will value Amir and lowry a lot more than they will demar or vasquez. So trade discussions with those teams should be structured accordingly. We may be able to oversell Amir to one of these teams, while we'll have to beg them to accept derozan in a trade etc.
moremilk is offline   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 01:38 PM   #13 (permalink)
payin Dwyane Wade

Senior Member
 
JoeyJoJo Shabbadu's Avatar
 
Join Date: Jul 2010
Posts: 4,445
Representing:
Default

Can someone post the table/list. Hate ESPN insider stuff.
JoeyJoJo Shabbadu is offline   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 01:45 PM   #14 (permalink)
sticking to his guns... is stuck to his guns...

Senior Member

 
DanH's Avatar
 
Join Date: Jul 2010
Location: Montreal, QC
Posts: 4,126
Representing:
Default

Raptor player rankings:

Player | ORPM | DRPM | RPM

Amir | 1.96 | 2.88 | 4.84
Lowry | 3.59 | -0.10 | 3.49
Novak | 1.59 | -0.33 | 1.26
Patterson | -0.05 | 1.14 | 1.09
Fields | -0.07 | 1.11 | 1.04
Hayes | -1.31 | 1.84 | 0.53
GV | 1.37 | -1.02 | 0.35
DD | 0.35 | -0.05 | 0.30
Salmons | -1.36 | 0.19 | -1.17
Ross | 0.69 | -2.05 | -1.36
JV | -2.09 | -0.06 | -2.15
de Colo | -1.86 | -0.78 | -2.64
TH | -3.16 | 0.29 | -2.87
DanH is online now   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 01:56 PM   #15 (permalink)
LX
with pink peppercorns

In the Paint


 
LX's Avatar
 
Join Date: Dec 2007
Location: Toronto
Posts: 27,655
Representing:
Default

So Demar drops down the list, not due to his defense, but because of his offensive output. Can that be explained, or do we just figure that he isn't a solid contributor?

In a certain sense it works in that he is below a lot of Toronto guys ahead of him in terms of team play, particularly if you don't separate out a pretty big change over the last half of the season as to how he has played. But does he not also drop down overall relative to other stars due to being on a team that stresses team play to such a degree? In the end there is maybe stuff to interpret here, which i have no problem with - it's how the rankings will be used by people that have no interest in interpreting anything, but rather see this as a final "objective" interpretation of cold hard numbers.
LX is online now   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 02:08 PM   #16 (permalink)
sticking to his guns... is stuck to his guns...

Senior Member

 
DanH's Avatar
 
Join Date: Jul 2010
Location: Montreal, QC
Posts: 4,126
Representing:
Default

I think it speaks a lot to DD's use - throughout most of the game I would think he has a positive impact on the offense, but in the 4th he absolutely kills it due to bad playcalling and excessive iso's (not necessarily all on him, obviously). He does still have an overall positive impact offensively, which is good to see, and he's improved drastically defensively this year compared to similar stats in previous years. Anything positive at all is "above average", so this looks pretty good on him, I think.

Players like Fields and Amir who do so many little things will always look great on these team-based stats, but you still need guys like DD, Lowry, etc to make the wheels turn, and if they can do so while still posting solid (or superstar, in Lowry's case) numbers in team-oriented stats, that's great news.

I'm doing a quick and simple wins analysis with these plus-minus numbers (I hate ESPN's WARP) to see how it shakes out compared to stuff like WS and WP.
DanH is online now   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 02:13 PM   #17 (permalink)
doesn't care where LeBron plays. Or Melo.

OY!!
 
niggles's Avatar
 
Join Date: Dec 2012
Location: Loch Ness
Posts: 2,741
Representing:
Default

Quote:
Originally Posted by JoeyJoJo Shabbadu View Post
Can someone post the table/list. Hate ESPN insider stuff.
Or like anyone else, you could just click on the link that starts jeff's post and read the chart.
niggles is offline   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 02:19 PM   #18 (permalink)
doesn't care where LeBron plays. Or Melo.

OY!!
 
niggles's Avatar
 
Join Date: Dec 2012
Location: Loch Ness
Posts: 2,741
Representing:
Default

Quote:
Originally Posted by DanH View Post
I think it speaks a lot to DD's use - throughout most of the game I would think he has a positive impact on the offense, but in the 4th he absolutely kills it due to bad playcalling and excessive iso's (not necessarily all on him, obviously). He does still have an overall positive impact offensively, which is good to see, and he's improved drastically defensively this year compared to similar stats in previous years. Anything positive at all is "above average", so this looks pretty good on him, I think.

Players like Fields and Amir who do so many little things will always look great on these team-based stats, but you still need guys like DD, Lowry, etc to make the wheels turn, and if they can do so while still posting solid (or superstar, in Lowry's case) numbers in team-oriented stats, that's great news.

I'm doing a quick and simple wins analysis with these plus-minus numbers (I hate ESPN's WARP) to see how it shakes out compared to stuff like WS and WP.
I like the fact that the end result adds both offensive and defensive stats to give a final figure. Easier to identify whose a 2 way player. Look at the defensive stats for LeBron. Teams just have to make him work on the defensive end- results could be interesting.
niggles is offline   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 02:35 PM   #19 (permalink)
sticking to his guns... is stuck to his guns...

Senior Member

 
DanH's Avatar
 
Join Date: Jul 2010
Location: Montreal, QC
Posts: 4,126
Representing:
Default Wins

The average would be 0.100, as is usual with win stats.

Player | RPM Wins/48 | RPM Wins to date | Equivalent team record

Amir | .136 | 6.1 | 56-26
Lowry | .126 | 7.1 | 52-30
Novak |.110 | 1.2 | 45-37
Patterson | .108 | 2.3 | 44-38
Fields | .108 | 0.7 | 44-38
Hayes | .104 | 1.2 | 43-39
GV | .103 | 2.6 | 42-40
DD | .102 | 6.1 | 42-40
Salmons | .091 | 2.3 | 37-45
Ross | .090 | 3.8 | 37-45
JV | .083 | 3.7 | 34-48
de Colo | .080 | 0.2 | 33-49
TH | .078 | 1.5 | 32-50

As a frame of reference, I put the equivalent team wins for a team made up entirely of that calibre of player as well, since it is easier than going by individual wins.

In terms of summing wins and predicting actual success, this method seems to come up a bit short. Doesn't mean much, as it is not really a wins model. Either way though, all players (including the Gay era) summed up gives a total of 41 wins right now, when the team sits at 45. TOR does have a .496 SOS (by ESPN), or a -0.4 by Bball-ref. The ESPN SOS doesn't change much, but the Bball-ref bumps it to 42 wins. Still, undercutting it a bit. Mind you, a few wins either way falls well within the error expected. I'd like to take a quick look league wide at a few other teams and see if there's a pattern.
DanH is online now   Boss Key Wife Key Reply With Quote
Old 04-07-2014, 03:29 PM   #20 (permalink)
sticking to his guns... is stuck to his guns...

Senior Member

 
DanH's Avatar
 
Join Date: Jul 2010
Location: Montreal, QC
Posts: 4,126
Representing:
Default

Oh, wow, looking at it now, I'm really artificially damping down the player ratings here. The individual plus-minus is actually isolated from their teammate effects - so in theory, Amir, for example is a plus-4.84 when playing with 4 average teammates. So his impact on 1/5 of the game (his personal win rate) would be 5 times what I calculated. I thought looking at my earlier attempt that the ratings were too flat across the team. Let me try that again.

Player | RPM Wins/48 | RPM Wins to date

Amir | .198 | 8.7
Lowry | .185 | 10.3
Novak |.144 | 1.6
Patterson | .141 | 2.9
Fields | .139 | 0.9
Hayes | .122 | 1.4
GV | .113 | 2.8
DD | .111 | 6.7
Salmons | .056 | 1.4
Ross | .054 | 2.3
JV | .030 | 1.3
de Colo | .022 | 0.1
TH | .016 | 0.3

That results in the team summing to 43 wins, 44.3 when adjusting for SOS per Bball-ref. Also seems more valid. Points to a lot of improvement needed from Ross and JV (unsurprising, them being sophomores and all) - although that is also the most likely source of improvement, so really bodes well for the future. Certainly makes TH look bad (so yay! I'm officially biased towards this stat).

I'm impressed with Hayes and Novak in this stat. Seems they are quietly a couple of the more effective (effectively used?) players on the team.
DanH is online now   Boss Key Wife Key Reply With Quote
Reply


Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On


All times are GMT -4. The time now is 11:32 AM.


Powered by vBulletin® Version 3.7.1
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Copyright RaptorsForum.com 2005-2011

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24