WAR versus WPA/LI

[Rebbiv]

I read an interesting article today where Sabathia leads Verlander in Wins Above Replacement. As some of you know, I am partial to WPA/LI.

Top 10 MLB Pitchers WAR

1. Halladay 8.0
2. Sabathia 7.1
3. Verlander 7.0
4. Kershaw 6.8
5. Lee 6.7
6. Haren 6.2
7. Weaver 5.7
8. Hernandez 5.5
9. Wilson 5.5
10. Bumgarner 5.4

Top 10 MLB Pitchers WPA/LI

1. Verlander 6.48
2. Kershaw 4.66
3. Halladay 4.00
4. Hamels 3.90
5. Lee 3.87
6. Weaver 3.84
7. Kennedy 3.70
8. Beckett 3.62
9. Sabathia 3.60
10. Cain 3.44


Top 10 MLB Batters WAR

1. Ellsbury 8.7
2. Bautista 8.1
3. Kemp 7.6
4. Pedroia 7.6
5. Granderson 7.3
6. Votto 7.0
7. Braun 7.0
8. J. Upton 6.8
9. Kinsler 6.6
10. A. Gonzalez 6.6


Top 10 MLB Batters WPA/LI

1. Bautista 6.58
2. Braun 5.63
3. Kemp 5.50
4. M. Cabrera 5.08
5. Votto 4.65
6. Berkman 4.61
7. Fielder 4.52
8. A. Gonzalez 4.37
9. Pujols 3.83
10. Granderson 3.81


From a WAR perspective, Ellsbury would be the AL MVP with Sabathia the Cy Young. The NL would have Halladay winning both. Of course, this would only be if you used this stat to decide.

From a WPA/LI perspective, Verlander and Kershaw would be the Cy Youngs. Bautista and Braun would be the MVPs. Now, it is really close between Verlander (6.48) and Bautista (6.58).

Additionally, with WPA/LI, if pitchers could not be MVP, the Miguel Cabrera would be second in the MVP behind Bautista.

 

[kingofdetroit57]

Great post. WPA/LI looks WAY more accurate to me on both accounts.

 

[Mitchrapp]

But I certainly wouldn't vte based on those two numbers alone. There's more to a persons game than WAR or WPA/LI.

 

[tycobb420]

Mitch[/color]]But I certainly wouldn't vte based on those two numbers alone. There's more to a persons game than WAR or WPA/LI.

I think some look at only the newer statistics and even then only a cherished couple to determine things. They don't look at the whole picture. Verlander third in WAR is a joke. The fact Bautista is not ahead of everyone, except maybe Verlander, is a joke. Shows there is no perfect stat.

 

[Mitchrapp]

I'm not sure what both these stats entail. But I'm sure they don't account for his no hitter, certainly not his other almost no hitter against the Angels. The fact he goes 6 innings, 100 pitches every game this season etc. Just by watching Baseball I can determine who the CY should go to and what players are MVP worthy. I don't need a stat like WAR and WPA/LI to tell me who the best pitcher is. I know who it is.

 

[johnny2x2x]

WPA is Win probability added. For instance say the win probability for a given game for your team is 50%, a player hits a double, that probability jumps to 55% so that player is credited with 5% or .05 wins added for that play. Every AB for a pitcher is tallied for the season. This has a strength in the a walk off HR counts as more than a HR in a blowout and such. It's a good tool to evaluate a players performance, but they say it's not very predictive of future performance.

WPA/LI Here's a good description.
http://www.fangraphs.com/blogs/index.php/get-to-know-wpali/

 

[Mitchrapp]

So what about a HR in a 10-1 game that becomes 12-11 game in the end. Does that HR become more valued at that point? Because if it doesn't its severely flawed.

 

[Rebbiv]

My primary reason for posting this was to discredited WAR. It is a very misleading stat and context neutral. Meaning, a home run with a 6 run lead counts the same as a home run that breaks a tie, or is the go ahead. The article I read was discounting Verlander from being in the MVP discussion because he had a lower WAR than Sabathia. Again, I do not like that stat, and if you looked at my preffered stat, WPA/LI, then it is pretty close between the two.

WPA/LI (LI = Leverage Index). Puts more value on pressure situations.

Both stats are "counting stats". Meaning, they add like HRs, RBIs, Runs, etc.

Now, if you look at Boesch and Santiago this year. Boesch = 1.8 WAR where Santiago = 1.7 WAR. Understand that defense UZR is included in WAR, as well as position multplier (i.e. SS get more of a multilpier than OF). Now, if you look at WPA/LI, Boesch = 1.51, where Santiago = - 0.38. In both stats, average = 0.

AL WPA/LI Leaders

1. Bautista 6.58
2. Verlander 6.48
3. M. Cabrera 5.08
4. A. Gonzalez 4.37
5. Weaver 3.84
6. Granderson 3.81
7. Ellsbury 3.81
8. Hamilton 3.78
9. Beckett 3.62
10. Sabathia 3.60

This is an objective way to evaluate the season.

1987 AL WPA/LI


1. Clemens 4.93
2. Trammell 4.49
3. McGwire 4.44
4. Boggs 4.29
5. J. Key 4.10
6. D. Evans 4.00
7. Viola 4.00
8. G. Bell 3.84
9. Puckett 3.65
10. Seitzer 3.54

1987 AL WAR

1. Clemens 9.3
2. Boggs 9.2
3. Trammell 7.9
4. Higuera 7.4
5. Langston 6.5
6. Moiltor 6.3
7. Viola 6.2
8. J. Key 6.1
9. Saberhagen 5.8
10. G. Bell 5.6

George Bell won the MVP over Trammell. Clemens won the Cy Young over Jimmy Key.

Using WPA/LI, you can make the case that Trammell should have won over Bell without being subjective. However, if we are wanting Verlander to win the MVP this year, than Clemens would have a case in 1987.

 

[Mitchrapp]

Trammell should have won it on a subjective level. Did they have WAR or WPA/LI back in '87? I don't remember if thy did.

 

[Rebbiv]

Mitch[/color]]So what about a HR in a 10-1 game that becomes 12-11 game in the end. Does that HR become more valued at that point? Because if it doesn't its severely flawed.

It is the conditions as to when the event occurred, not the end result. You could also add that a HR down 4-6, put the team ahead 7-6, but the team ultimately lost 7-12. A player hitting cannot control his team's pitching from that point on. But, at that moment in time, did he succeed or fail.

 

[jared]

Now my head hurts. I don't know if I have room for anymore new stats. I used to know my kids ss #'s, but I think that was replaced with UZR/150. If you make me forget my anniversary, you're a dead man Reb.

 

[Rebbiv]

Mitch[/color]]Trammell should have won it on a subjective level. Did they have WAR or WPA/LI back in '87? I don't remember if thy did.

No, they did not exist. However, you can go back an review data from pretty much all seasons/decades.

 

[Mitchrapp]

Mitch[/color]]So what about a HR in a 10-1 game that becomes 12-11 game in the end. Does that HR become more valued at that point? Because if it doesn't its severely flawed.

It is the conditions as to when the event occurred, not the end result. You could also add that a HR down 4-6, put the team ahead 7-6, but the team ultimately lost 7-12. A player hitting cannot control his team's pitching from that point on. But, at that moment in time, did he succeed or fail.

Also, did he hit that HR against a team like Texas or a team chasing you like Cleveland where it could result in loss of a game in the standings? I want Detroit pitchers and hitters to do better against the ALC teams that are chasing you.

 

[tycobb420]

Mitch[/color]]

It is the conditions as to when the event occurred, not the end result. You could also add that a HR down 4-6, put the team ahead 7-6, but the team ultimately lost 7-12. A player hitting cannot control his team's pitching from that point on. But, at that moment in time, did he succeed or fail.

Also, did he hit that HR against a team like Texas or a team chasing you like Cleveland where it could result in loss of a game in the standings? I want Detroit pitchers and hitters to do better against the ALC teams that are chasing you.

Just to echo...I remember an interview with Greg Maddox. He had a monster lead and started laying grapefruits up for hitters. His plan was to confuse them when he faced them again by changing up his pitching strategy.

 

[Rebbiv]

http://www.realclearsports.com/lists/dominant_pitching_seasons/intro.html

Here is the article.

WPA/LI Leaders since 1969:

1. Martinez (00) 8.5
2. Maddux (95) 7.2
3. Clemens (97) 7.2
4. Carlton (72) 7.1
5. Tudor (85) 6.9
6. Gooden (85) 6.9
7. Martinez (99) 6.7
8. Scott (86) 6.6
9. Guidry (78) 6.5
10. Seaver (71) 6.4
11. Blue (71) 6.3
12. Maddux (94) 6.3
13. Martinez (97) 6.3
14. Maddux (98) 6.3
15. Santana (04) 6.3
16. Maddux (97) 6.2
17. Verlander (11) 6.2

Keep in mind, the season isn't over. So, Velander's WPA/LI can go up even further. Bold is for those that won the CY. Only Maddux' 98 did someone not on this list win the CY, and it was Glavine.

Bob Gibson's 1968 season WPA/LI was 7.3.

 

[kingofdetroit57]

You don't hear much of him anymore, but Martinez was ridiculously dominant. He's honestly one of the absolute best to ever pitch.

 

[@Stakesarehigh]

You don't hear much of him anymore, but Martinez was ridiculously dominant. He's honestly one of the absolute best to ever pitch.

His 2000 season is the best of my time....and to do that in the height of the steroid era....amazing.

people reference to gibson's 68 often but to me given the hitting stats you almost have to think Martinez's season could be the GOAT of the modern era.

 

[jmakula87]

You don't hear much of him anymore, but Martinez was ridiculously dominant. He's honestly one of the absolute best to ever pitch.

His 2000 season is the best of my time....and to do that in the height of the steroid era....amazing.

people reference to gibson's 68 often but to me given the hitting stats you almost have to think Martinez's season could be the GOAT of the modern era.

Yeah, that 2000 season was absolutely ridiculous! Every time I look at it, I still find myself surprised at how good of a year it was.

 

[tycobb420]

His 2000 season is the best of my time....and to do that in the height of the steroid era....amazing.

people reference to gibson's 68 often but to me given the hitting stats you almost have to think Martinez's season could be the GOAT of the modern era.

Yeah, that 2000 season was absolutely ridiculous! Every time I look at it, I still find myself surprised at how good of a year it was.

Some folks claim 1999 was better. In fact, someone (don't remember who) on MLB Net did so just this week.

 

[jmakula87]

Yeah, that 2000 season was absolutely ridiculous! Every time I look at it, I still find myself surprised at how good of a year it was.

Some folks claim 1999 was better. In fact, someone (don't remember who) on MLB Net did so just this week.

He did have a few more Ks and Ws in that 99 season, but his 2000 season is just incredible. It would be tough for me to say his 99 season was better, but I would have liked to hear that guy's argument.

I don't get the MLB Network here at college It has been a rough transition haha.

http://espn.go.com/mlb/player/stats/_/id/2717/pedro-martinez