It's Never Too Early To Talk MVP

by Sky Kalkman on Aug 4, 2009 12:56 PM EDT in Awards & Honors Comment 67 comments

Warning: After the jump I'm going to present information which could be interpreted as stating, "Franklin Gutierrez is the American League MVP." Without proper preparation, that information could cause physical and emotional damage, so whatever you need to do to prevent such damage, do it now.

MVP. It's amazing how three little letters can open up such a huge can of worms. We'll have a debate about defining what MVP means and should mean down the road, but today, I want to present some data to match one popular definition that doesn't usually get attacked from an objective angle. That definition is:

The MVP is the player who did the most to help his team try to win baseball games, with production weighted by game state.

An expansion of that definition and the jaw-dropping data which supports the claim I warned you about up front are after the jump.

Some notes on that definition:

For hitting, we're using Win Probability Added, which takes into account the score, inning, runners, and number of outs. A bases-loaded walk in a tie game in the bottom of the ninth is more valuable than a grand slam in a blowout.
Defense matters. Fangraphs' UZR and position adjustments are used. Unfortunately, we can't yet adjust fielding for game state, but I'm not sure that wouldn't just be introducing more noise than usefulnsess.
For pitchers, I'm using FIP-based WAR, not WPA. Why? WPA has too much noise for pitchers, including team defense, luck, and ballpark effects. Adding that noise hurts more than adding in clutchiness helps.
Team context doesn't matter. If a player racks up .5 WPA in a loss, that counts just as much as .5 WPA in a win. And a 10 win season on a bad team counts just as much as a 10 win season on a playoff team. This definition treats the MVP as mostly an individual award.

In other words, we're pretty much using WAR from Fangraphs, but substituting WPA for hitters in place of batting runs above average (scaled by using 10 runs per win.) If you don't like that definition of MVP, no problem, but you won't find the following data useful. Please feel free to define MVP your own way and present the data to match.

National League MVP

Rank	Player	Team	WAR*
1	Tim Lincecum	Giants	6.6
2	Chase Utley	Phillies	6.1
3	Albert Pujols	Cardinals	5.5
4	Prince Fielder	Brewers	5.2
5	Raul Ibanez	Phillies	4.9
6	Hanley Ramirez	Marlins	4.7
7	Dan Haren	Diamondbacks	4.7
8	Pablo Sandoval	Giants	4.7
9	Matt Kemp	Dodgers	4.6
10	Javier Vazquez	Braves	4.6
11	Joey Votto	Reds	4.5

What surprises me up front is not just that Albert Pujols trails Tim Lincecum, but that he also trails Chase Utley. As you'll see below, part of the reason is that UZR sees Pujols merely as average in the field this year, which goes against recent history. (Then again, Utley barely rates better than average, himself.)

Whatever you think of Raul Ibanez's huge season, you have to give him credit for it. Although, maybe someone will try to give Ryan Howard credit for it somehow.

Regarding Matt Kemp, when was the last time an eight-hole hitter was an MVP candidate, and it wasn't for his glove?

American League MVP

Rank	Player	Team	WAR*
1	Zack Greinke	Royals	6.4
2	Franklin Gutierrez	Mariners	5.8
3	Justin Verlander	Tigers	5.3
4	Ben Zobrist	Rays	5.3
5	Roy Halladay	Blue Jays	5.2
6	Chone Figgins	Angels	4.7
7	Josh Beckett	Red Sox	4.6
8	Cliff Lee	- - -	4.6
9	Felix Hernandez	Mariners	4.3
10	Joe Mauer	Twins	4.2
11	Jon Lester	Red Sox	4.1
12	Brandon Inge	Tigers	4.0

There you go. If you buy this definition of MVP, you either need to vote for a pitcher or agree that Franklin Gutierrez is the AL MVP so far. Crazy, huh? Not crazier than FGut's 2.3 wins due to clutch hitting.

What I find fascinating about this list is there are zero traditional big boppers on here. FGut, Figgins, and Inge are glove-first position players having solid years at the plate. Zobrist is hitting.298/.409/.573, but well, he's Ben Zobrist. And Mauer, well, maybe this is finally his year, but people have made an effort to avoid voting for him in the past. The AL MVP really should be Zack Greinke's to lose, and many starting pitchers should finish in the top ten. Cliff Lee's numbers include his time in NL.

* * *

How about a data dump to finish up? Here are the top 58 players in MLB so far this year -- everybody to hit the 3.0 WAR* level through Sunday's games. Most of the data is self-explanatory and taken from Fangraphs. Again, WAR* is WAR but with WPA (times ten, which is the "hitting" column) instead of batting runs. Clutch is WPA*10 - batting runs.

In addition to some interesting names on the list (Asdrubal Cabrera, Nyjer Morgan, Pedro Feliz, Juan Rivera, and Dallas Braden) there are some interesting names not on the list (Dustin Pedroia, Alex Rodriguez, Adrian Gonzalez, Justin Morneau, Matt Cain, and Johan Santana.)

Rank	Player	Team	WAR*	Hitting	(clutch)	Field	Pos	Rep
1	Tim Lincecum	Giants	6.6
2	Zack Greinke	Royals	6.4
3	Chase Utley	Phillies	6.1	41	6	3	2	15
4	Franklin Gutierrez	Mariners	5.8	29	23	15	2	12
5	Albert Pujols	Cardinals	5.5	47	3	0	-8	15
6	Justin Verlander	Tigers	5.3
7	Ben Zobrist	Rays	5.3	28	-2	13	0	12
8	Prince Fielder	Brewers	5.2	50	14	-5	-8	15
9	Roy Halladay	Blue Jays	5.2
10	Raul Ibanez	Phillies	4.9	39	14	1	-4	12
11	Chone Figgins	Angels	4.7	23	7	8	1	16
12	Hanley Ramirez	Marlins	4.7	28	-3	2	5	14
13	Dan Haren	Diamondbacks	4.7
14	Pablo Sandoval	Giants	4.7	35	15	-1	0	14
15	Matt Kemp	Dodgers	4.6	22	1	8	2	14
16	Josh Beckett	Red Sox	4.6
17	Cliff Lee	- - -	4.6
18	Javier Vazquez	Braves	4.6
19	Joey Votto	Reds	4.5	44	20	-4	-6	10
20	Felix Hernandez	Mariners	4.3
21	Casey Blake	Dodgers	4.2	26	17	1	1	13
22	Ryan Zimmerman	Nationals	4.2	12	-2	13	2	15
23	Joe Mauer	Twins	4.2	26	-4		3	12
24	Jon Lester	Red Sox	4.1
25	Brandon Inge	Tigers	4.0	16	6	10	2	14
26	Ryan Howard	Phillies	4.0	29	13	4	-8	15
27	Miguel Tejada	Astros	4.0	26	15	-6	5	15
28	Ubaldo Jimenez	Rockies	4.0
29	Marco Scutaro	Blue Jays	3.8	12	-3	6	5	16
30	Bobby Abreu	Angels	3.8	29	7	-1	-5	14
31	Joel Pineiro	Cardinals	3.8
32	Aaron Hill	Blue Jays	3.8	17	4	3	1	16
33	Kevin Youkilis	Red Sox	3.8	27	-4	2	-4	13
34	Asdrubal Cabrera	Indians	3.7	24	17	-3	3	12
35	Mark Reynolds	Diamondbacks	3.7	27	3	-4	-1	15
36	Johnny Damon	Yankees	3.7	35	18	-8	-5	14
37	Nyjer Morgan	- - -	3.6	1	-4	23	-2	15
38	Derek Jeter	Yankees	3.6	11	-9	5	4	16
39	Justin Upton	Diamondbacks	3.6	20	-3	7	-5	14
40	Randy Winn	Giants	3.5	13	19	12	-4	14
41	CC Sabathia	Yankees	3.5
42	David Wright	Mets	3.4	24	2	-6	2	15
43	Mark Teixeira	Yankees	3.4	27	1	-1	-8	16
44	Josh Johnson	Marlins	3.4
45	Gavin Floyd	White Sox	3.4
46	Victor Martinez	- - -	3.4	19	5	0	0	15
47	Evan Longoria	Rays	3.3	7	-11	11	1	14
48	Chris Carpenter	Cardinals	3.3
49	Adam Wainwright	Cardinals	3.3
50	Pedro Feliz	Phillies	3.3	12	16	5	2	14
51	Juan Rivera	Angels	3.2	17	1	9	-5	12
52	Ichiro Suzuki	Mariners	3.2	16	-4	6	-5	15
53	Scott Rolen	- - -	3.2	15	5	3	1	13
54	Ryan Braun	Brewers	3.2	24	-4	-3	-5	15
55	Shane Victorino	Phillies	3.1	23	7	-9	2	15
56	Matt Holliday	- - -	3.1	17	-7	3	-5	15
57	Curtis Granderson	Tigers	3.0	12	2	1	2	15
58	Dallas Braden	Athletics	3.0

What's that, you're up for a little schadenfreude? Ok, but Royals fans, look away...

Player	Team	WAR*	Hit	(clutch)	Field	Pos	Rep
Vernon Wells	Blue Jays	-2.1	-18	-10	-20	1	15
Jose Guillen	Royals	-1.6	-4	1	-18	-4	10
Delmon Young	Twins	-1.6	-11	-1	-9	-3	8
Josh Geer	Padres	-1.5
Yuniesky Betancourt	- - -	-1.5	-19	2	-10	4	10
Fernando Tatis	Mets	-1.4	-20	-16	1	-4	7
Ben Francisco	- - -	-1.4	-12	-15	-13	-2	12
Brian Giles	Padres	-1.3	-10	4	-8	-3	9
Mike Jacobs	Royals	-1.3	-15	-9	0	-9	11
Austin Kearns	Nationals	-1.2	-17	-11	1	-3	7
Ty Wigginton	Orioles	-1.2	-19	-15	0	-2	9
Alexi Casilla	Twins	-1.2	-11	3	-7	1	6

0 recs | Comment 67 comments | Share

Story-email Email Printer Print

Comments

Silver lining for the Royals

Betancourt essentially replaced two terrible players at shortstop: Tony Pena, Jr. and Luis Hernandez, who combined for -1.1 WAR…which means that with Betancourt’s -1.5 WAR, not only did the Royals find a way to add suckitude to their team, they also opened up a roster spot. It takes some creative thinking to find a way to add that kind of value!

by RoyalsFanInBillings on Aug 4, 2009 1:23 PM EDT reply actions 1 recs

phillies

5 phillies have topped the 3 win mark so far this year.

kinda incredible.

by jamiethekiller on Aug 4, 2009 1:41 PM EDT reply actions 0 recs

Sky and his WAR, I need to break out my work that shows what voters really care about

HRs, RBIs, Hits and making the playoffs.

Jeff Zimmerman - Protecting the world from RBI's and Wins from my mom's guest house.

by Jeff Zimmerman (TucsonRoyal) on Aug 4, 2009 1:43 PM EDT reply actions 0 recs

I don't care what voters care about. I don't care about their votes.

(That’s what I keep telling myself at least.)

Predicting who other people are going to vote for just seems like a silly exercise. Well, silly that we need to do it. Not silly that it can be modeled using hilarious stats.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Aug 4, 2009 1:54 PM EDT up reply actions 0 recs

Sky - You are probably one of those people that think Rickey H. actually belongs in the Hall of Fame

Jeff Zimmerman - Protecting the world from RBI's and Wins from my mom's guest house.

by Jeff Zimmerman (TucsonRoyal) on Aug 4, 2009 2:33 PM EDT up reply actions 0 recs

Depends which Rickey H. you're talking about.

Rickey Halbrin was awful.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Aug 4, 2009 2:35 PM EDT up reply actions 0 recs

I still don't like FIP for pitchers

It ignores timing with runners on base, which, whether repeatable or not, does actually happen and provides value to the club. I would like to see something like defense adjusted RA WAR; although I’m not sure how you would do that.

F*** Billy Beane... actually, I kinda like Holliday

by vivaelpujols on Aug 4, 2009 1:46 PM EDT reply actions 0 recs

Isn't that still context independent

F*** Billy Beane... actually, I kinda like Holliday

by vivaelpujols on Aug 4, 2009 8:59 PM EDT up reply actions 0 recs

Wouldn't need to be.

Given whatever game situations you wanted to take into account, you adjust the expected run/win value of the batted ball accordingly.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Aug 4, 2009 9:47 PM EDT up reply actions 0 recs

Wouldn't something like WPA/LI work?

F*** Billy Beane... actually, I kinda like Holliday

by vivaelpujols on Aug 4, 2009 9:00 PM EDT up reply actions 0 recs

For pitchers?

You’ve still got the issue of all the value going to the pitcher and none to the fielder.

For batters, WPA/LI gives you the actual win value of the play, but takes away the measurement of clutchness. So it’s a good alternative to BRAA, but might not be what Sky is looking for.

by Dan Turkenkopf on Aug 4, 2009 9:05 PM EDT up reply actions 0 recs

The whole /LI part is what you want, but like Dan said, you still have the park/defense issues.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Aug 4, 2009 9:48 PM EDT up reply actions 0 recs

So FIP/LI?

F*** Billy Beane... actually, I kinda like Holliday

by vivaelpujols on Aug 4, 2009 10:26 PM EDT up reply actions 0 recs

Hmm, yeah, it might be that simple.

tRA/LI?

Actually, I’m liking that a lot.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Aug 4, 2009 10:49 PM EDT up reply actions 0 recs

Same

Get David Appelman on the phone.

F*** Billy Beane... actually, I kinda like Holliday

by vivaelpujols on Aug 4, 2009 10:55 PM EDT up reply actions 0 recs

You could pronounce it like "trolley."

Or like “tra-li la-li la.” Sing when you say “tRA/LI.”

And That Happened: Sports with a Hint of Sarcasm

by philkid3 on Aug 5, 2009 1:48 AM EDT up reply actions 0 recs

I don't know if that's going to work

I think you’re just throwing together two unrelated measures at that point.

tRA doesn’t really describe the win value based on the events, so dividing by the LI of the plate appearance doesn’t really correct for the situation.

Perhaps if you apportioned WPA credit based on the hit chances defined by tRA, then you’d be in better shape.

And have an interesting team-based alternative to DER.

by Dan Turkenkopf on Aug 5, 2009 7:53 AM EDT up reply actions 0 recs

I'm thinking less of WIN LI and more of RE24 LI

You’re right that we’d need to create a new framework for the weights of various events (HR, K, BB, BIP for FIP, more for tRA) based on RE24 game states.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Aug 5, 2009 8:11 AM EDT up reply actions 0 recs

I've got some concerns about the base/out leverage

Which I’ll expand upon when I get home from work.

I’m not sure the numbers match up.

by Dan Turkenkopf on Aug 5, 2009 11:31 AM EDT via mobile up reply actions 1 recs

Doesn't pass the smell test

Randy Winn ahead of Mark Teixeira? I’m sorry, that is just ridiculous.

by fletchdm on Aug 4, 2009 2:17 PM EDT reply actions 0 recs

Shocked

I rescind my criticism after reading your criteria a bit more closely. That doesn’t mean I’m not utterly shocked and slightly incredulous. I mean, Randy Winn has a .691 OPS for crying out loud…compared with Morneau’s .979 and Teixeira’s .926.

by fletchdm on Aug 4, 2009 2:31 PM EDT up reply actions 0 recs

Two explanations.

Winn has 17 run advantage on defense over Tex. Similar over Morneau.
Winn’s hitting has been worth 19 runs more than you’d expect from a typical .691 OPS, because of clutchiness: he’s hit when it’s mattered. Tex hasn’t been clutch or unclutch. Morneau’s been 11 runs unclutch.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Aug 4, 2009 2:41 PM EDT up reply actions 1 recs

Defense

Haven’t we established that UZR varies so much from year-to-year that looking at a 4 month sample size is essentially worthless? I’d be interested in re-running this MVP test using three year UZR averages, or something of the like…

I like the idea of adding Win Probability Added into an MVP formula.

by fletchdm on Aug 4, 2009 4:50 PM EDT up reply actions 0 recs

Yes.

I’m not arguing that Winn IS 17 runs better, just that defense is one place he’s getting a big edge in the numbers used.

Looking longer term… While Tex and Morneau have various degrees of good defensive reputations, Tex owns a career +2 UZR/150 and Morneau’s at +3. Winn’s been great in RF, with a career +15 UZR/150. Average in CF (which is as valuable as +10 in a corner spot), but only +2 in LF, which is weird. If we call Winn +10, Tex/Morneau +2, and include the +5 for position, that’s a 13 run edge, not far off from the +17 used above.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Aug 4, 2009 5:21 PM EDT up reply actions 0 recs

Whether or not Winn's UZR is sustainable/reflects true talent/what have you, it still happened and still provided value to the team.

by acblue on Aug 4, 2009 6:34 PM EDT up reply actions 0 recs

Well, not exactly.

UZR’s not like AVG, where it actually happened. UZR’s a good estimate of what happened, but there’s some error beyond statistical variation. It could be that Winn’s actually performed at +5 runs, but UZR is off and says +12.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Aug 4, 2009 8:50 PM EDT up reply actions 0 recs

My understanding was always that the margin of error for actual recorded events was much smaller than that.

by acblue on Aug 5, 2009 1:15 AM EDT up reply actions 0 recs

Sky's talking about this

As cwyers put it:

Observed Performance = True Talent + Random Error + Bias

I take Sky to be saying there is perhaps bias in the way that UZR measures defense. And if that is the case, I agree with him.

by Tommy Bennett on Aug 5, 2009 1:19 AM EDT up reply actions 0 recs

For UZR?

I think the margin of error is huge. The zones are large and determined by visual guesswork on the part of people watching, and the batted-ball velocities are S/M/H (again determined by guesswork) – imagine if pitched balls were measured as S/M/H. Can’t wait for Hit F/X. :>

by BobbyMac on Aug 5, 2009 4:55 AM EDT up reply actions 0 recs

Well supposedly most of that stuff somewhat evens out over a full season

And MGL applies a lot of adjustments (park factors, batter handedness, pitcher tendencies, etc.) to help that process.

F*** Billy Beane... actually, I kinda like Holliday

by vivaelpujols on Aug 5, 2009 5:00 AM EDT up reply actions 0 recs

the end product...

is only as good as the data coming in. He does the best he can with what’s available.

by BobbyMac on Aug 5, 2009 5:02 AM EDT up reply actions 0 recs

regression

I think for this reason, you either need to take a multi-year sample, or regress the current season. Gutierrez has gotten only about 300 balls so far, and while his past shows he’s an elite RF, the +14.5 runs in CF this year don’t really mean what they seem to mean (yet).

by BobbyMac on Aug 5, 2009 4:59 AM EDT up reply actions 0 recs

Another problem with UZR is that it is meant to be used to find how much a player is above or below league average for a single season, not season to season.

Say you have a shortstop that makes the exact number of good and bad defensive plays in two different seasons. He is the exact same defender. In the second year, the top 4 shortstops all get injured. These 4 are all replaced with below average defensive shortstops. The original shortstop’s UZR value will then go up in the second year because the average league ability went down even though he fielded the same.

Jeff Zimmerman - Protecting the world from RBI's and Wins from my mom's guest house.

by Jeff Zimmerman (TucsonRoyal) on Aug 5, 2009 11:51 AM EDT up reply actions 0 recs

good call...

I’ve been sort of wondering why so many 1b/LF types rank out so high, knowing that part of it was “compared to peers”, and thinking that a lot of “non-fielders” get put at these two positions… but being too lazy to dig into how much effect that was really having. :>

by BobbyMac on Aug 5, 2009 2:37 PM EDT up reply actions 0 recs

What if

You put together a lot of other metrics and try to spit out on the other side your MVP.

In the fan post I made yesterday, I include WAR (did not include pitchers) but hings like Runs Created, HR, BAvg, RBI, OPS, WPA, UZR, etc. (I think I look at 18 different measures) and just use rankings…

Of course this system would not account for huge gaps in difference between HR for a guy with 35 and leading and a guy with 28 and in 2nd place, but it gives a measure of relation.

by backtocali on Aug 4, 2009 2:22 PM EDT reply actions 0 recs

But why use all of those? Where's the logic?

Is OPS just as important as UZR? Isn’t HR already included in OPS?

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Aug 4, 2009 2:34 PM EDT up reply actions 0 recs

email...

When I get a free few moments I will email you a spreadsheet.

Im no math whiz that is for sure, its just a way I try to figure out who the best player is…

The method no doubt has flaws, but it seems to me to be an accurate measure of an mvp type of year.

by backtocali on Aug 4, 2009 3:12 PM EDT up reply actions 0 recs

Flaws but it's accurate?

Let me throw this warning out there, which I mentioned in the recent Elias Free Agent rankings post:

It’s really tough to throw together a hodge podge of random stats and have the rankings that pop out be awful. Good hitters tend to be better at everything and the more stats you include, the more likely you are not to reward something good at least a little bit. But a few players do fall through the cracks, like Ichiro as a Type B free agent possibly. It’s really not hard to fix those cracks and once the small differences between players become more important, such as for MVPs or putting together an entire team as an MLB executive, those small errors are magnified into a big whiff.

So sure, play around with random stats, combine the numbers, and have fun. Just don’t present it as anything meaningful unless it actually is. What’s the logic behind the system? Why is it better than something else? Why should we care about it?

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Aug 4, 2009 3:34 PM EDT up reply actions 0 recs

Why not just use the Elias Free Agent Rankings?

Jeff Zimmerman - Protecting the world from RBI's and Wins from my mom's guest house.

by Jeff Zimmerman (TucsonRoyal) on Aug 4, 2009 3:17 PM EDT reply actions 0 recs

Because then Ryan Howard would win

F*** Billy Beane... actually, I kinda like Holliday

by vivaelpujols on Aug 4, 2009 9:01 PM EDT up reply actions 0 recs

Matt Kemp only 9th in the NL?

Really? And I have a problem with any list that has Figgins ahead of Mauer, I’m sorry.

"Hey we got a lot in common here... I'm gonna rape you"

by MerryGoByeBye on Aug 4, 2009 3:24 PM EDT reply actions 0 recs

Playing time.

Mauer missing 27 games (or whatever it was) probably has something to do with that.

---
http://www.beyondtheboxscore.com
http://www.rightfieldbleachers.com

by Jack Moore on Aug 4, 2009 4:11 PM EDT up reply actions 0 recs

and his presence in the top 10

does have something to do w/ defense, as his is outstanding.

by chuckb on Aug 4, 2009 8:58 PM EDT up reply actions 0 recs

Hm, interesting. Though, i think of MVP as “best player” without weighting for the game state, so give me Ben Zobrist and Chase Utley or give me death!

Fire Jim Leyland

And now at Beyond the Boxscore!

by Mike Rogers on Aug 4, 2009 3:54 PM EDT reply actions 0 recs

My issue with WPA

Yes, a bases-loaded walk in a tie game in the bottom of the ninth is more valuable than a grand slam in a blowout. But the situation the batter comes up in is a matter of luck. Isn’t the ballpark a player plays in a matter of luck? We neutralize for that. We don’t use things like RBI and R to neutralize for the effect teammates have. WPA undermines that. If a player hits a walk-off HR in the 9th, should it really count more than a HR in the 1st? Without the HR in the first, the HR in the 9th suddenly loses most of its WPA value. I guess my point is, either you decide MVP should include luck-based context or you decide it shouldn’t.

by yellomellojello on Aug 4, 2009 6:35 PM EDT reply actions 0 recs

Maybe you prefer using RE24, which takes into account outs and runners on base, but not score or inning?

It’s still somewhat independent, but rewards hitters for hits with runners in scoring position versus treating all singles the same.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Aug 5, 2009 10:48 AM EDT up reply actions 0 recs

I guess it comes down to not being sure about how to include luck-based factors.

On the one hand, I understand rewarding players for performing well in high leverage situations, but I can’t reconcile that with the fact that being in those situations is a matter of luck. I think I would just prefer to ignore “clutch” completely, but I can see the argument for the other side.

by yellomellojello on Aug 5, 2009 6:58 PM EDT up reply actions 0 recs

Don't tell anyone, but I agree with you.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Aug 5, 2009 11:11 PM EDT up reply actions 0 recs

Your criteria after the jump could have been wOBA+UZR+most ridiculous facial hair and I would have bought into it fully.

Gutz is dangerously close to unseating Mike Cameron on my all-time-favorite-player list.

by acblue on Aug 4, 2009 6:35 PM EDT reply actions 0 recs

False Premise

For hitting, we’re using Win Probability Added, which takes into account the score, inning, runners, and number of outs. A bases-loaded walk in a tie game in the bottom of the ninth is more valuable than a grand slam in a blowout.

…

Team context doesn’t matter.

I’m sorry, but aren’t those statements in conflict? Doesn’t “team context” include the rest of the lineup getting on base so the player in question can have an opportunity for a clutch at-bat? Aren’t players like Mauer, who are surrounded by largely mediocre lineups with poor on-base potential hurt by this, while players like Figgins benefit from increased likelihood that they will come to bat with runners on?

I’ll readily admit I’m not a cutting-edge stat guy, and f I’m missing something I’d love to be educated.

Carriage Return

by CarterHayes on Aug 4, 2009 6:39 PM EDT reply actions 0 recs

I think (think) that what Sky meant to imply with that statement

is that it doesn’t whether the player’s team is good or bad (which is certainly a factor in MVP voting) or in what game situation the player racked up WPA; win probability added is win probability added, in other words.

For example, a player that performs well down the stretch and has memorable hits/defensive plays in close games against a rival might be more likely to garner MVP consideration than someone who provides similar overall value but does not necessarily turn in such memorable performances. Or that a player that has a bonkers game in what ends up being a blowout is not valued any less highly than a player that turns has a similarly impressive game in a close contest.

But if I am wrong, of course feel free to correct me.

by acblue on Aug 4, 2009 6:50 PM EDT up reply actions 0 recs

Clarification

I guess what I’m driving at is the idea that WPA is completely team-independent strikes me as poorly supported. Using the Twins as an example again, at least a third of their lineup is terrible on any give day, but they’ve suffered from some particularly putrid “production” from Alexi Casilla, Brendan Harris, and the rest of the motley crew who they’ve slotted in the No. 2 spot in the lineup (well beyond what we should expect from a professional lineup). If those guys aren’t getting on or moving runners into scoring position, isn’t that going to have a negative impact on the middle of that order (Mauer, Morneau, Kubel, and Cuddyer) to rack up WPA?

Carriage Return

by CarterHayes on Aug 4, 2009 7:22 PM EDT up reply actions 0 recs

Exactly what I was about to add to my previous comment

Players on worse teams are punished by WPA. By the nature of Win Probability Added, if your team loses a lot, you will have fewer opportunities to accumulate WPA.

by yellomellojello on Aug 4, 2009 7:35 PM EDT up reply actions 0 recs

Precisely.

As such, it seems just as flawed as RBI as a metric.

Carriage Return

by CarterHayes on Aug 4, 2009 7:45 PM EDT up reply actions 0 recs

If your team wins a lot, you also have fewer opportunities to accumulate WPA.

What help is playing lots of close games, which create high leverage situations, which make the same basic batting events more important. On a player level within the lineup, having players who get on base in front of you will also tend to make your at-bats higher leverage.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Aug 4, 2009 8:56 PM EDT up reply actions 0 recs

No, you're right.

That’s why I typed this:

This definition treats the MVP as mostly an individual award.

But the beginning of that paragraph pretty much negates the value of writing that. In that sense, team context matters. It doesn’t matter in terms of where the team is in the standings.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Aug 4, 2009 8:54 PM EDT up reply actions 0 recs

Another question.

To what extent is Mauer penalized by the lack of quality defensive metrics for catchers in comparison to the other positions on the diamond? Obviously we get a positional adjustment for him, but that’s it. Seems like we could do better in that area.

Carriage Return

by CarterHayes on Aug 4, 2009 7:48 PM EDT reply actions 0 recs

We tend to think he's good, but we've missed on player defense before, right? +5 runs? +10 runs? I'm guessing more like +5.

We actually have a guest post tomorrow presenting the saber world’s best guess as to catcher defensive value.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Aug 4, 2009 8:58 PM EDT up reply actions 0 recs

Excellent.

I’ll be looking forward to it.

Carriage Return

by CarterHayes on Aug 4, 2009 9:02 PM EDT up reply actions 0 recs

Mystery guest? Who is it? I can't wait.

Rally’s methodology seems to peg the right guys, but we don’t know if the numbers are as good quantitatively as they are quantitatively.

Marlin Maniac, a Florida Marlins blog

by SFiercex4 on Aug 5, 2009 10:38 AM EDT up reply actions 0 recs

Hint:

add another 8-10 runs or so, at least, for Mauer’s defense. He gets into the top 5 but he’s not going to get the 20 runs or so he needs to surpass Greinke.

by chuckb on Aug 4, 2009 9:01 PM EDT up reply actions 0 recs

Your formula needs some tweaking

First, divide all pitcher WAR by 5 because they only pitch one out of every five games. Second, get rid of Votto because no Canadian who wasn’t from BC has ever won an MVP. Third, you need to add 1.5-2.5 WAR for players with a white complexion to adjust for scrappiness. Forth minus 5 WAR if the player’s position is between two black players (The Utley rule.) Finally, catchers and former catcher get -10 WAR because Mike Pizza never won an MVP. This will give you a more accurate level of performance

Don't believe the lies Bill!!!! look at the sparkly ERA!!! Sparkly, Sparkly!!! - McCovey Chronicles

by Trenchtown on Aug 6, 2009 1:02 AM EDT reply actions 0 recs