Bloops: NyTimes.com – The Mid-season All-Stars by WAR
Posted by Sean Forman on July 9, 2010
Keeping Score: In All-Star Calculations, Some Fall Short - Bats Blog - NYTimes.com.
This week's piece in the New York Times.
I have to say it was a bit strange to see the dueling outfields as David DeJesus, Brett Gardner, and Shin-Soo Choo vs. Angel Pagan, Matt Holliday, and Andres Torres.
July 9th, 2010 at 5:29 pm
I guess it's all a matter of what you want the All Star Game to be. Should it be a showcase of the 35 most valuable players in each league (and if so, is WAR really THE way to judge value?), or should it be the 35 players who fans, managers, and players think are the most entertaining/exciting and will put on the best show? Seems like more money would be generated in the latter case, so I'm thinking that's probably it.
July 9th, 2010 at 5:35 pm
Wait a minute! Justin Morneau has a higher Total Zone Rating than any other infielder and all but a handful of outfielders. Most of his traditional fielding stats look similar to his career norms. It's true that he has participated in 3-6-X double plays at a career-high rate so far in 2010. Is his TZR way up largely, or entirely, because of this?
July 9th, 2010 at 11:47 pm
WAR currently gives Morneau waaaaay too much credit over Cabrera. I have no problem with Morneau being the first-half MVP but my vote goes to Cabrera, much as I would like to please my fellow Yanks fans and give that hypothetical vote to Cano.
July 10th, 2010 at 12:08 am
Malcolm-
Part of the problem is how the point of the game has been co-opted/corrupted by Selig's "Now it counts" nonsense. Previously, I believed the game was for the fans and that fan balloting, at least for starters, was what mattered. It was a chance for fans to see the guys they wanted to. Personally, I voted for the guys having the best first halves, but if people wanted to see over-the-hill stars, so be it. As long as we recognized that that was what the game was, and didn't assign more value to it (either the game itself or use it as an indicator of a player's talent levels, such as in HoF conversations), that would have worked. But, instead, we use it to determine home field advantage and it gets factored into HoF votes and now you have Omar Infante possibly deciding the World Series AND probably getting a single HoF vote someday and making me want to jump of a bridge.
July 10th, 2010 at 1:55 am
Omar Infante will never receive a single HOF vote.
July 10th, 2010 at 1:56 am
I can't believe he's only 28. It feels like he has been around forever.
July 10th, 2010 at 2:13 am
If you have a formula that spits out Brett Gardner as an All-Star selection over Carl Crawford, then it is time to re-think that formula.
July 10th, 2010 at 2:23 am
These advanced defensive numbers seem very wrong (or more precisely, unreliable) by the eyeball test, though there isn't really a thorough way of testing them.
July 10th, 2010 at 6:51 am
Part of the problem is that All-Star selections for the outfield are just listed as "Outfield" rather than LF, CF, or RF. So basically there was no Center Fielder selected among the starters. The National League has two right fielders and a left fielder in Either, Heyward, and Braun and the American League has two Left fielders and one right fielders in Crawford, Hamilton, and Ichiro.
Center Fielders are among the most underrated players in baseball because they're often lumped together with other outfielders as if each outfield position has the same defensive demands.
Defense is rarely taken into consideration when voting/selecting outfielders so as a result you usually have 6 of the best hitting outfielders rather than 6 of best outfielders.
July 10th, 2010 at 6:57 am
@7,@9
Yeah, WAR has a different replacement level for CF's so comparing Gardner to Crawford is comparing apples to oranges.
DeJesus and Choo over Crawford is a much more reasonable call:
DeJesus .328/.398/.463
Choo .286/.390/.475
Crawford .320/.378/.516
They are all neck and neck, though. As of this morning they all have 3.3 WAR.
July 11th, 2010 at 1:15 am
#10
I understand your stat comparison with DeJesus and Choo, but you only include the BA, OBP, and SLG. Here is my argument for Crawford:
Crawford: 67 runs scored
Choo: 48 runs scored
DeJesus: 45 runs scored
Crawford: 48 RBI
Choo: 43 RBI
DeJesus: 35 RBI
I believe that run production is an important statistic in evaluating hitters. While Crawford may have a slightly lower OBP than the other two players, he is a better run-scoring threat with his base-stealing ability (30 SB....Choo has 12 and DeJesus 3), speed on the basepaths, and power (higher SLG % than the other two guys). In addition to runs scored, Crawford has also driven in more runs than both Choo and DeJesus.
A statistical analysis of the 3 players gives Crawford the clear edge. If defensive ability is included, then the gap becomes even wider.
July 11th, 2010 at 1:28 am
Pete-
RBI and Rs tell us more about a player's teammates than it does him. Crawford scores those runs only in part because of what he does. Without talented teams, he gets stranded. And Crawford drives in runs only in part because of what he does. With the bases empty, there is no one to drive in.
BA/OBP/SLUG, while not ideal, are quick and dirty ways at getting at what a hitter is doing individually, apart from his teammates.
Do you really think Crawford would have as many RBI/Rs if he played on KC or Cleveland?
July 11th, 2010 at 8:51 am
BSK - I agree with the premise of your position. A player who hits a lot of singles and does not steal bases relies on his teammates to drive him in. However, a player who (A) has some HR power; (B) is perrenially among the league leaders in triples; (C) is perrenially among the league's SB leaders; and (D) has speed on the basepaths; greatly increases his own chances of scoring. Take a look at Rickey Henderson's career numbers. His prolific Runs totals were a result of his own speed and the power that he developed during his career.
As for the RBI totals, Crawford has been in the #2 hole the entire season. It is not like he has a Rickey Henderson or Vince Coleman hitting in front of him. Jason Bartlett began the season as the leadoff hitter, but his production has greatly dropped this year. Maddon has tried a couple of other guys in the leadoff spot with mixed results, including rookie John Jaso, whose production has leveled off after a hot start; B.J. Upton, who has a very low BA; and Ben Zobrist. Ben has been the most successful of the 4.
I will analyze some numbers to determine how many of Crawford's runs/RBI (and Choo and DeJesus) are a function of the batting order around them. It is likely that both of us may be right in our arguments. Crawford might score more often because he has Longoria hitting behind him (but also a sub-.200 Pena), but I think that he has created his own run-scoring opportunities and hit better with RISP.
No hard feelings. I love these baseball debates, and if the numbers show I'm wrong, I will be the first to admit it.
July 11th, 2010 at 9:54 am
BSK, Valid points on the problem with Runs/RBI.
Pete,
I think what you're leaving out of the equation is that Dejesus and Choo are both Right Fielders which is a more defensively demanding position than Left Field. And on top of that Dejesus and Choo are both very good defensive right fielders.
As far as just offensive players, Crawford is the best of the three: Crawford 138ops+, Choo 135ops+, Dejesus 134ops+. Plus, Crawford is 30/38 in stolen bases. But I think they all come out to about 3.3 WAR because of the defense Dejesus & Shoo display in Right.
July 11th, 2010 at 10:14 am
Pete-
Just to be clear, I'm not saying necessarily that Crawford is not as good as DeJesus or Choo... only that RBI/Rs isn't the way to really show it. While you make some good points that there are ways individual players' abilities impact RBI/Rs, it still only tells PART of the story. RBI/Rs are also good at telling us what happened, but not necessarily what WILL happen. They're not particularly predictive.
Now, you pointed to some interesting aspects of Crawford's offensive package that point towards his value. So THOSE are the numbers we should look at, not RBI/Rs, which only in part are derivatives of those numbers. Your second case is a far more compelling one than the RBI/Rs one. RBI/Rs without context are meaningless. You provided context, but that is where we should start rather than where we should end.
July 11th, 2010 at 3:43 pm
I haven't analyzed Crawford's offensive opportunities yet (my wife wanted to go to the outlet mall), but I noticed that Crawford scored a run in the 1st inning on a botched rundown play (perhaps the throw was rushed because of Crawford's quickness?). He also popped a 2-run HR to add to his run and RBI total (it also gave the Rays a brief 4-3 lead).
In response to the LF/RF comparison, I was only comparing Crawford to Choo and DeJesus because the previous poster brought those 2 players into the discussion. Is RF defensively 'more demanding'? Yes, I will give you that, but I don't believe that the gap between the two positions is that wide. It's not like the difference between SS and 1B.
Let us look at the defensive prowess of DeJesus and Crawford. I think both are equally as good. Crawford is #2 among active LF in the 'Total Zone Runs as LF' category. He trails only Garrett Anderson, who is about 10 years older. Look for Crawford to soon take the lead in the active category. In terms of Career leaders at LF, Crawford is #6 all-time (trailing only Barry Bonds, Yaz, Willie Wilson, G. Anderson, and Luis Gonzalez).
David DeJesus is also an excellent outfielder, and he and Crawford are in the same category of excellence. If you look at the 'Total Zone Runs' stats for active OF's, Crawford is #8 with 65 and DeJesus is #10 with 62. All-time numbers? Crawford is #41 all-time among OF's in that statistical category, while DeJesus is #47 all-time.
Also, Crawford's range is head and shoulder above most OF's. He led AL LF's in putouts in 2005, 2006, 2007, (not 2008-he missed a lot of time due to injury) and 2009. He is currently 2nd to Podsednik in the AL this year.
Based on those stats, it's my position that DeJesus does NOT have a clear edge defensively. I did not analyze Choo because he does not have the body of work that Crawford and DeJesus have compiled. Since both DeJesus and Crawford are top of the line defensive OF's, we must look at their offensive numbers to determine who is having the better season. Based on every that I have submitted (and will submit once I break down Crawford's offensive stats), I believe that Crawford is clearly more deserving of an All-Star appearance than the other 2 guys.
Here's some video evidence of Crawford's speed creating a run and some defensive videos. Take a look at him scoring earlier today after being caught in a rundown, and take a look at the July 5 catch against Scutaro.
http://mlb.mlb.com/video/play.jsp?content_id=9877843
July 11th, 2010 at 10:53 pm
Quickly:
1. the difference between RF and LF is marginal. RF requires a better arm, but I think Crawford has a solid arm, i.e., runners wouldn't be going 1st to 3rd on him all the time were he in RF. And if the league has bigger left fields than right fields (no idea if this is true for the current/recent AL), then LFs do have more responsibility.
2. No question that Crawford is a very good LF, but how many guys play strictly LF for even 130 games a year? If Crawford is one of just a few, it's easier for him to lead the league in PO repeatedly.
3. History has nothing to do with whether Crawford has actually been better defensively than DeJesus this year. Of course, Total Zone is also far from perfect, and even less so over 3 months of the season.
July 12th, 2010 at 5:24 am
I think the thing that's making things seem a little odd is that Dejesus is having a career year. So to see him rated as equal to Carl Crawford might seem a little strange. But if you look closer, he's hitting .327/.397/.463 with an ops+ 134 and he's playing a very good defense right field. He's 7th in the league in B.A. and 6th in on base percentage.
Who knows if he will be able to maintain these number because he has a career 108ops+.