Jimmy Rollins and the 2007 MVP Award (and WPA)

[filihok]

So as to not sidetrack the Ron Cey/Steve Garvey discussion

He started every game and went to bat more times in major league history. He was the first player with 200 hits, 20 doubles [38], triples [20], HR's [30] and stolen bases [41]. He was the first NL shortstop with 380 total bases. All while playing gold glove caliber [he won] defense. He recorded seven consecutive multi-hit games from August 26 to the 31st, part of an 18 for 32 stretch, and homered in back to back games on August 28 and 29th during the Phillies four game sweep of the Mets. In the Phillies final six wins he either scored or drove in the game winning or go ahead run. The Phils won just 5 of 26 games and none after June 9 that Rollins didn't get on base.

Career highs [to that point and currently] in: G, PA-778 [ML record], AB-716 [LL], R-139 [NL], H-212, doubles-38, triples-20 [NL], HR-30, RBI-94, SB-46, 49 BB, 85 SO, BA-.296, OBP-.344, SLG-.531, OPS-.875, OPS+-119, TB-380

He was THE [I can't emphasize it enough] reason the Phillies were able to overtake the Mets [he predicted they would pre-season too]. The Phillies were equally, if not more, historically hapless than the Mets. So him taking a team with the Phillies history on his back was/is a pretty big accomplishment. Even bigger when we look back, realizing that year was the year the NL East title streak started, culminating with 2 NL pennants and a WS.

Dug out my old Sports Illustrated I saved with him on the cover for some info. Good times.

In the Phillies final six wins he either scored or drove in the game winning or go ahead run. The Phils won just 5 of 26 games and none after June 9 that Rollins didn't get on base.

This is the part I'll focus on.

I don't know if you're aware of a stat called WPA (Win Probability Added). If not, here's a link.

Basically, WPA looks at a team's likelihood of winning a game after each plate appearance. If the likelihood increases the player is credited, if the likelihood decreases the player is debited.

For example, a leadoff single increases a team's Win Expectancy by about 3.5%. The player leading off would get WPA of .035.

Likewise, if he made an out the visiting team would be about 3.5% less likely to win, and he would have .035 subtracted from his WPA.

A home team trailing by 1 run in the bottom of the 9th with a man on first and 2 outs has about a 10% chance of winning. A walkoff home run increases the chance to 100% so a batter would be credited with .90 of WPA.

In 2007 Jimmy Rollins had a cumulative WPA of 2.69. What does that mean? To put it in context.
He was 25th in the majors and 3rd on his team behind Utley (3.40) and Howard (3.08).


His hits were not all that timely.

[Francoeurstein]

I still think he is one of the least-deserving MVP winner of all time.

[Brooklyn]

Bad choice. David wright as the presumptive MVP that year until the mets collapsed, and I think the writers felt they needed to give it to a Phillie that year since they overtook the mets. But as the op infers, he may not have even been the best choice on his own team. I think the writers also liked the 20/20/20/20 thing, as it got a lot of press, but that s more of a statistical anomaly than an MVP gauge

I still think wright was the MVP that year, but I understand the writers reluctance to give it t a Met after the collapse, and if I remember correctly Wright didn't light it up during the stretch run. If the needed to give it to a Phillie, Utley was a better choice.

That said, I'm not a fan of WPA as anything more than a novelty stat; it certainly shouldn't determine MVP. It is highly dependent on the game situation - a player at bat losing a close game has much more room for greater WPA then a player at bat winning a close game. I would think that over a course of a season that a good player on a bad team would have a higher WPA then a good player on a good team, although not sure if that is true

I know they try to solve for this by dividing WPA by LI (leveraged index). I don't know enough about LI, but this seems like a better version

[GiambiJuice]

Wright got robbed. It's not his fault his teammates collapsed all around him. All Wright did over the final month of the season is hit .352/.432/.602. His numbers after the All Star break were .364/.465/.596.

The Mets famously blew a 7 game lead with 17 games left to play. Over the final 17 games of the 2007 season, Wright hit .397/.451/.575. He did everything he could possibly be asked to do in order to will his team into the postseason. He would have easily won the award if not for his teammates collapsing.

It was a travesty that Rollins won.

[Bothrops Atrox]

Agreed that Wright (and many others) were much better choices. FTR, WPA is an offense-only stat. As far as I know, they have yet to even attempt to show wim probability for defensive plays made, for obvious reasons. Not that it would have helped Rollins enough to deserve the MVP.

[GiambiJuice]

Let's not forget, it was a double robbery for Rollins that year. He also stole the Gold Glove from Troy Tulowitzki.


This should probably be moved to the History section.

[OmartgSS]

Holliday with 336 votes came in 2nd behind Rollins 353........Oh yeah, Coors Field, Holliday hasn't hit a lick in St Louis.

Managers and coaches never have and probably never will vote a rookie SS a GG. I do hope that Rollins is keeping Tulo's GG dusted. Same goes for the RoY Award that Braun stole from Tulowitzki in 2007.

WAR 2007
Tulowitzki 6.5
Braun 1.8

[GiambiJuice]

Holliday with 336 votes came in 2nd behind Rollins 353........Oh yeah, Coors Field, Holliday hasn't hit a lick in St Louis.

Managers and coaches never have and probably never will vote a rookie SS a GG. I do hope that Rollins is keeping Tulo's GG dusted. Same goes for the RoY Award that Braun stole from Tulowitzki in 2007.

WAR 2007
Tulowitzki 6.5
Braun 1.8

To be fair, Braun had a historic season at the plate for a rookie. Everyone knew he was a butcher at 3rd base but defense was and is difficult to quantify. No one was looking at WAR in 2007. I'm not sure it even existed yet in its current form.

I agree that Holliday would have been a very good candidate. I wouldn't have had a problem with him winning the MVP over Wright.

[filihok]

Agreed that Wright (and many others) were much better choices. FTR, WPA is an offense-only stat. As far as I know, they have yet to even attempt to show wim probability for defensive plays made, for obvious reasons.

One could replace batting wins with WPA...

If one did do that for Jimmy Rollins' 2007 season one would find that instead of 23.5 runs (2.29 wins) Rollins would be credited with 2.69 wins (WPA), slightly increasing his overall WAR to about 7.3. Still well behind Wright...and Utley

[filihok]

Oh yeah, Coors Field, Holliday hasn't hit a lick in St Louis.

Holliday 2006-2008: .400 OBP/.579 SLG - 43% better than league average (once adjusted for home field)
Holliday 2010-2012: .385 OBP/.517 SLG - 48% better than league average (once adjusted for home field)

Looks like hitting in Coors certainly helped Holliday's raw numbers, almost exactly to the degree that we'd expect (his production relative to league average is roughly the same).

[GiambiJuice]

Holliday 2006-2008: .400 OBP/.579 SLG - 43% better than league average (once adjusted for home field)
Holliday 2010-2012: .385 OBP/.517 SLG - 48% better than league average (once adjusted for home field)

Looks like hitting in Coors certainly helped Holliday's raw numbers, almost exactly to the degree that we'd expect (his production relative to league average is roughly the same).

For what it's worth, his splits in 2007 were quite extreme. I can understand why the voters would have considered him a product of Coors Field.

Home: .376/.435/.722 25 HR, 82 RBI
Away: .301/.374/.485 11 HR, 55 RBI

I remember starting this thread during that 2007 season...

[OmartgSS]

I've always thought this really helped Rollins win the 2007 MVP.

2007: "The team to beat" and MVP

In January, Rollins stated:

"The Mets had a chance to win the World Series last year. Last year is over. I think we are the team to beat in the NL East, finally. But, that's only on paper.

This put him in the voters eye early. I truly believe that Rollins wouldn't have won the MVP in a tight race without the prediction, no matter what the stats say. Curtis Granderson(WAR 7.2) had a season that compares with Rollins with the four 20's and finished a distant 10th in the AL MVP voting. Of course, Granderson didn't beef up his BBWAA stats with a prediction.

[OmartgSS]

Holliday 2006-2008: .400 OBP/.579 SLG - 43% better than league average (once adjusted for home field)
Holliday 2010-2012: .385 OBP/.517 SLG - 48% better than league average (once adjusted for home field)

Looks like hitting in Coors certainly helped Holliday's raw numbers, almost exactly to the degree that we'd expect (his production relative to league average is roughly the same).

I was being sarcastic with the Coors Field statement. Holliday became the first slugger to leave the Rockies in his prime so he is a good gauge to see how he does away from Coors Field.

[GiambiJuice]

I was being sarcastic with the Coors Field statement. Holliday became the first slugger to leave the Rockies in his prime so he is a good gauge to see how he does away from Coors Field.

Andres Galarraga also didn't skip a beat when he went to the Braves for two years after his Rockies stint.

[filihok]

For what it's worth, his splits in 2007 were quite extreme. I can understand why the voters would have considered him a product of Coors Field.

Because the average writer, and voter, doesn't understand the proper use of splits.