Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Baseball Newsstand > Baseball Primer Newsblog > Discussion
Baseball Primer Newsblog
— The Best News Links from the Baseball Newsstand

Wednesday, April 16, 2014

Minuteman News Center: Giandurco: This means WAR

If we don’t end WAR, WAR will end us! (trips over imported blunderbuss stand)

I have frequently argued how economic statistics do not always give an accurate appraisal of the current climate. The same happens in sports, at times. I’d like to demonstrate why a currently voguish baseball statistic is vastly misleading, in the hopes of engendering healthy skepticism in fields other than sports.

The statistic called “Wins Above Replacement,” or WAR, is considered the best all-around rating methodology for everyday players (as opposed to pitchers). It supposedly shows how many wins a player adds or subtracts from his team, compared to just an average player at his position. I would like to use several well-known position players and their WAR statistics to evaluate this rating system.

Yogi Berra is, bar none, the winningest baseball player in history. He played 17 seasons, and the Yankees won 14 pennants and 10 World Series. His manager, Casey Stengel, called him his manager on the field. His position, catcher, is critical, touching the ball on every pitch. He won three MVP awards and made 15 All-Star teams. In his magisterial biography, journalist Allan Barra (no relation) surmises that Berra may have been the greatest catcher of all time, but was at least in the top 4. He was named to the All-Century team as the American League catcher.

According to WAR, Berra is the 97th best player of all time. 97th! By comparison, Jeff Bagwell is rated 35th. Bagwell played 15 seasons, winning one pennant and no World Series. He made four All-Star games and won one MVP. But he is 62 places better than Berra, the winningest player of all time, who played a much more crucial position. Since the entire WAR concept is based on winning, how could a player who contributed to so much more winning be rated so much lower? No offense to Bagwell, who I liked, but does anyone believe he is more valuable than Berra?

...I can go on. Take Carl Yasztremski, ranked 28. While a Yankee fan, I really admired Yaz as a kid. He was great. But his Sox never won a World Series, and he had a career average line of .285, 22 HR, 90 RBI, and 89 runs scored. There is simply no comparing the two men’s achievements, yet Yaz is ranked five spots higher than DiMaggio.

I think this brief overview shows that WAR is not rational, and also that all of us need to be skeptical when “experts” throw around statistics, especially when based on new and untested metrics. Often, such stats turn out to be meaningless in the real world. For example, we have finally regained all the jobs lost since 2007, but most of the jobs created have been low-wage, meaning that income inequality has gotten worse, not better, ever since the people who talk a lot about “income inequality” took over Washington

Repoz Posted: April 16, 2014 at 08:21 AM | 106 comment(s) Login to Bookmark
  Tags: history, sabermetrics

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

Page 1 of 2 pages  1 2 > 
   1. Good cripple hitter Posted: April 16, 2014 at 09:09 AM (#4686781)
A stat WAR
Will set off the keg
"My words are war"
Should a word have two meanings?
What the #### for?
Should words serve the truth?

I stand for language
I speak the truth
I shout for history
This article's a cesspool
For all the ####
To run down in
   2. Jolly Old St. Nick Is A Jolly Old St. Crip Posted: April 16, 2014 at 09:37 AM (#4686800)
Hard to argue with the author's point, since the only objective study of winning shows that 19 out of the 20 greatest winners of all time were Yankees. And it's about time that Charlie Silvera and Tommy Byrne got some recognition.

   3. Randy Jones Posted: April 16, 2014 at 09:46 AM (#4686807)
So I assume this guy would agree that Clay Bellinger was better than Ted Williams, 2 WS rings to 0.
   4. Repoz Posted: April 16, 2014 at 09:49 AM (#4686810)
journalist Allan Barra (no relation) surmises that Berra

Words fail me.
   5. Greg K Posted: April 16, 2014 at 10:03 AM (#4686827)
Of course, the point to bring up with regards to Berra (which from a quick scan doesn't appear to be in the article at all) is that WAR under-rates catchers. Which is as we all know a debate we've had here many a time.

EDIT: Which if the author had taken the time to look at the matter would have immediately recognized.
   6. SoCalDemon Posted: April 16, 2014 at 10:16 AM (#4686837)
Also, "Berra may have been the greatest catcher of all time, but was at least in the top 4."....who in the hell is he arguing with here?? I mean, I would say Bench, but Bench, Gibson, Rodriguez, Fisk, Piazza, Berra, those are pretty clearly the best catchers of all time, right? Who's arguing with this?
   7. I Am Not a Number Posted: April 16, 2014 at 10:48 AM (#4686853)
"It supposedly shows how many wins a player adds or subtracts from his team, compared to just an average player at his position."

Great place to start. Prove you do not even superficially understand the very thing you are criticizing.
   8. dlf Posted: April 16, 2014 at 10:54 AM (#4686861)
Bench, but Bench, Gibson, Rodriguez, Fisk, Piazza, Berra, those are pretty clearly the best catchers of all time, right? Who's arguing with this?


I feel pretty certain that you've named the top four, but once you get past Berra, Piazza, Bench, and Gibson (in no particular order), there are a several that could be included with the two Pudges you named. Guys like Carter, Dickey, Cochrane, Ewing, Mackey and a few more rank right with and perhaps above Fisk and Rodriguez.
   9. BDC Posted: April 16, 2014 at 10:59 AM (#4686863)
What is "Minuteman News Center"? This seems to me like another of those school-newspaper articles we sometimes see here.
   10. tshipman Posted: April 16, 2014 at 11:05 AM (#4686868)
I think WAR's treatment of catchers is a pretty good argument against it's credibility.
   11. if nature called, ladodger34 would listen Posted: April 16, 2014 at 11:11 AM (#4686873)
What is "Minuteman News Center"? This seems to me like another of those school-newspaper articles we sometimes see here.


Just spitballing here.. but a tea party rag?
   12. The District Attorney Posted: April 16, 2014 at 11:18 AM (#4686882)
What is "Minuteman News Center"?
The front page looks like a pretty normal local Connecticut news site. New Englanders like to just throw "Minutemen" in there sometimes.

I can only consider articles like this if they include "What is it good for? Absolutely nothing!"
   13. AROM Posted: April 16, 2014 at 11:19 AM (#4686885)
I think WAR's treatment of catchers is a pretty good argument against it's credibility.


In general, the average WAR for the sum total of a team's catchers will be about the same as for their first basemen or shortstops. Individual catchers come up short in the season and career rankings because of playing time.

If you think it through it makes perfect sense. If you start from replacement level, and can sign either an elite 1B or an elite catcher, you wind up with:

1. 150 games of great 1B, 12 games from his backups
2. 130 games of great catching, 32 games from his backups.

Choosing option #1 will win you more games every time, assuming each is equally great on a per game basis.
   14. Mike Emeigh Posted: April 16, 2014 at 11:24 AM (#4686893)
I think the primary issue with an article like this is that the author starts with the assumption that a team's stars deserve almost all of the credit for winning, and that when a team doesn't win the stars deserve to be dinged for their failure to win. (This is essentially the Cabrera/Trout argument.)

In practice, good teams win - and lesser teams don't win - in large part because of their relative abilities to fill in the gaps around the stars. The Yankees didn't win just because of DiMaggio; they won because they also had Charlie Keller and Joe Gordon and Lefty Gomez and Red Ruffing and Spud Chandler and Tommy Henrich and Bill Dickey, and their lesser players like Crosetti and Selkirk and Red Rolfe were still pretty good. They also had a genius for picking up pitchers who'd contribute for three-four years, like Bump Hadley and Marius Russo and Atley Donald, and when those guys faded they'd be replaced with other guys who'd be equally good. And that was even before Casey got there and the Kansas City shuttle began.

In 1941, when both DiMaggio and Williams had their best seasons by oWAR, the Yankees had four other hitters with at least 4 oWAR; the Red Sox had 2. In 1947, when Williams lost a controversial MVP vote, the Red Sox had one other hitter with 4 oWAR; the Di-Maggio led Yankees had four hitters with at least 4 oWAR. In both seasons, the Yankees as a team had more oWAR than did the Red Sox, despite Williams's heroics.

WAR is designed to evaluate the extent to which a player contributes to wins for his team. If a player on a winning team, like DiMaggio or Berra, contributes fewer wins than does a player on a team that doesn't win as often, like Williams or Yaz, the issue isn't with the construction of the statistic; it's with the construction of the teams that surround each player.

-- MWE
   15. Pooty Lederhosen Posted: April 16, 2014 at 11:28 AM (#4686896)
Good cripple hitter, that post #1 is great. I love that song.
   16. jdennis Posted: April 16, 2014 at 11:32 AM (#4686899)
I tossed this when I got to the word "average" and you should too.
   17. Jolly Old St. Nick Is A Jolly Old St. Crip Posted: April 16, 2014 at 11:49 AM (#4686908)
In practice, good teams win - and lesser teams don't win - in large part because of their relative abilities to fill in the gaps around the stars. The Yankees didn't win just because of DiMaggio; they won because they also had Charlie Keller and Joe Gordon and Lefty Gomez and Red Ruffing and Spud Chandler and Tommy Henrich and Bill Dickey, and their lesser players like Crosetti and Selkirk and Red Rolfe were still pretty good.

Hell, the Yankees won their record five straight World Series with only one HoFer (Yogi Berra) in his prime (or near it) for the entire run. The heart of those teams consisted of three starters not within shouting distance of the HoF, three other HoFers who made significant contributions for only a year or two (Dimaggio in 1949-50, Mantle in 1952-53, and Ford in 1950 and 1953), and a bunch of near HoVGers. IIRC during that five year run, the Yankees didn't have a single leader in any individual batting or pitching category. Those leaderboards were dominated by the Red Sox and Indians.
   18. What did Billy Ripken have against ElRoy Face? Posted: April 16, 2014 at 11:54 AM (#4686912)
journalist Allan Barra (no relation) surmises that Berra

Words fail me.


I suppose next he's going to try to tell us that Gerardo Parra is from a different family too.
   19. McCoy Posted: April 16, 2014 at 12:05 PM (#4686920)
In general, the average WAR for the sum total of a team's catchers will be about the same as for their first basemen or shortstops. Individual catchers come up short in the season and career rankings because of playing time.

Does WAR still only measure a catcher's ability to throw out runners along with WP and PB?
   20. A Fatty Cow That Need Two Seats Posted: April 16, 2014 at 12:13 PM (#4686925)
The author is at least considering this an opportunity for dialogue and education on twitter
   21. Moeball Posted: April 16, 2014 at 12:24 PM (#4686941)
Conceptually, the idea of WAR at least makes sense. If your star player goes down, what you have left to replace him with usually isn’t a league-average player, unless you are a wealthy team with an exceptionally deep bench. Usually, for most teams, what’s left to replace him with is a marginal replacement player. Very definition of the term.

I still prefer WAA to WAR because I don’t care if my players are a little better than that crummy replacement player – are they helping me win games or not, relative to the league? A player with a WAR of 1.0 but a WAA of -1.0 is not helping me get to the postseason, he’s pulling the team down.

I recognize there are flaws in some of the weightings and we’ve gone on forever here about issues with the fielding calculations. One thing I liked about Michael Humphreys book Wizardry was that he includes the formulas in there that show you how he came up with his numbers. You may not agree with his conclusions but at least you can see where they came from.

What I would like to see is this – if you don’t like the TZ numbers or the DRA numbers or whatever, please show where you think the calculations are going astray. We’re all actually trying to get to the same place, where we can make some sort of realistic evaluation of players defensively as well as offensively. In this day and age where we have play by play data we ought to be able to do a better job of taking all the factors into account that can impact fielding ratings – is the player handling more or fewer plays because of the type of pitching staff? Is it because of the influence of the park itself? Is it because the manager uses shifts a lot? All of these things can skew the raw numbers but we ought to be able to sort out the noise from the useful data. Quite frankly, isn’t this the sort of thing Nate Silver should be devoting his time to? That would actually be useful.
   22. if nature called, ladodger34 would listen Posted: April 16, 2014 at 12:24 PM (#4686942)
The author is at least considering this an opportunity for dialogue and education on twitter


I especially liked his "takedown" of the bartender. Such an enlightened comment.
   23. if nature called, ladodger34 would listen Posted: April 16, 2014 at 12:25 PM (#4686943)
Moe,

You just put about 1000x more thought into your post than the fellow who wrote that article did.
   24. Publius Publicola Posted: April 16, 2014 at 12:46 PM (#4686962)
Those twitter threads are hilarious. Mike Petriello is blowtorching him and he just keeps digging in deeper.

And surprise, surprise. He's an Obama-hater too.
   25. JE (Jason) Posted: April 16, 2014 at 12:48 PM (#4686965)
The author might feel differently about WAR if it ranked Yankee players, including Luis Sojo, above everyone else.
   26. Hal Chase School of Professionalism Posted: April 16, 2014 at 12:53 PM (#4686967)
Wow, we're taking it easy on on ol' Vince compared to the people commenting on his page.

By now he's turtling and waiting for the linesmen to jump in and make the pounding stop.
   27. Rants Mulliniks Posted: April 16, 2014 at 12:59 PM (#4686974)
I think the primary issue with an article like this is that the author starts with the assumption that a team's stars deserve almost all of the credit for winning, and that when a team doesn't win the stars deserve to be dinged for their failure to win.


Why do people find it so difficult to understand that MLB is not the NBA? In the NBA, you could put Lebron James or Kevin Durant on a team like Utah (.296 W%) and the would immediately be a playoff team. That relationship never has and never will exist in baseball.
   28. Squash Posted: April 16, 2014 at 01:11 PM (#4686996)
In general, the average WAR for the sum total of a team's catchers will be about the same as for their first basemen or shortstops. Individual catchers come up short in the season and career rankings because of playing time.

If you think it through it makes perfect sense. If you start from replacement level, and can sign either an elite 1B or an elite catcher, you wind up with:

1. 150 games of great 1B, 12 games from his backups
2. 130 games of great catching, 32 games from his backups.

Choosing option #1 will win you more games every time, assuming each is equally great on a per game basis.


I agree with this. If a guy isn't playing it doesn't matter how good he is, he isn't helping you win. A catcher isn't "more great" when he plays simply because he plays less.

To keep everybody on equal footing as to how they're rated (i.e. catchers vs. the rest of the world), the solution is to go to some sort of WAR/games played stat. Basketball has already done this with Win Shares per Game, which is becoming a popular stat.

EDIT: In a way if you think about it you could make the same argument for pitchers but in reverse. If Clayton Kershaw could he would pitch every day and put up 40 WAR a season. But there are physical limitations on his job so his playing time is necessarily restricted. Hence he puts up 8 or so with a very high per-game value and a very low amount of games played.
   29. Jeltzandini Posted: April 16, 2014 at 01:19 PM (#4687003)
"It supposedly shows how many wins a player adds or subtracts from his team, compared to just an average player at his position."

Great place to start. Prove you do not even superficially understand the very thing you are criticizing.


Hey, it's right there in the name. Wins Above Raverage.
   30. Mike Emeigh Posted: April 16, 2014 at 01:34 PM (#4687022)
Why do people find it so difficult to understand that MLB is not the NBA?


Because MLB has tried to emulate the NBA in its marketing. MLB wants stars to market - and if they play in big markets on glamor teams, so much the better.

-- MWE
   31. DanG Posted: April 16, 2014 at 01:45 PM (#4687038)
In general, the average WAR for the sum total of a team's catchers will be about the same as for their first basemen or shortstops.
I'm probably showing my ignorance here, but should WAR be this way? Isn't it possible that different positions have different replacement levels? If your 5 WAR catcher gets hurt isn't it going to be harder to find a good replacement than if your 5 WAR firstbaseman gets injured?
   32. cercopithecus aethiops Posted: April 16, 2014 at 01:54 PM (#4687049)
His manager, Casey Stengel, called him his manager on the field.


That's not what Casey used to call Yogi in 1949.

In the NBA, you could put Lebron James or Kevin Durant on a team like Utah (.296 W%) and the would immediately be a playoff team. That relationship never has and never will exist in baseball.


But that just means that Mike Trout isn't as good at baseball as Lebron James is at basketball, right?
   33. Jose Can Still Seabiscuit Posted: April 16, 2014 at 02:08 PM (#4687055)
I'm probably showing my ignorance here, but should WAR be this way? Isn't it possible that different positions have different replacement levels? If your 5 WAR catcher gets hurt isn't it going to be harder to find a good replacement than if your 5 WAR firstbaseman gets injured?


Isn't that what the positional adjustments for WAR already do? I'm sure there is a compelling case to be made that the adjustments are not 100% accurate which is why it's always a good idea to use the numbers as a starting point rather than an ending point.
   34. Ron J2 Posted: April 16, 2014 at 02:15 PM (#4687060)
#31 There is a different (lower offensively) for catchers.

The logic behind WAR is that baseball talent follows a normal distribution (and that only the very right edge ever makes it to the pro level).

But if you want to perform a sanity check you can look at all of the catchers in organized baseball (MLEs for the guys in the minors). You'll find that the assumptions made in setting replacement level at any given position basically check out.
   35. DanG Posted: April 16, 2014 at 02:24 PM (#4687069)
#33-34:

So you're saying there is already a "catcher bonus" built into WAR?
   36. Ron J2 Posted: April 16, 2014 at 02:36 PM (#4687083)
#35 Yeah. Go to baseball-reference. Look at the player value section for any position player and look at the Rpos column. The difference between a full time 1B and a full time catcher is more than 20 runs.
   37. Karl from NY Posted: April 16, 2014 at 02:47 PM (#4687097)
But that just means that Mike Trout isn't as good at baseball as Lebron James is at basketball, right?

No, it means the sports are different in how they distribute opportunities among the players. Trout can't ever be more than 1/9 of his team's offense, while James could be close to 100% if he and his teammates wanted. Trout could be better at his sport by producing per opportunity at say 3 standard deviations better than average while James produces at 2, but produce less win value because his sport gives him far fewer opportunities.
   38. TDF, situational idiot Posted: April 16, 2014 at 03:10 PM (#4687116)
I think the primary issue with an article like this is that the author starts with the assumption that a team's stars deserve almost all of the credit for winning, and that when a team doesn't win the stars deserve to be dinged for their failure to win.

Why do people find it so difficult to understand that MLB is not the NBA? In the NBA, you could put Lebron James or Kevin Durant on a team like Utah (.296 W%) and the would immediately be a playoff team. That relationship never has and never will exist in baseball.
Fans and sportswriters also do this with football, which is why Marino, Tarkenton, and Moon are never considered among the best QBs.
In general, the average WAR for the sum total of a team's catchers will be about the same as for their first basemen or shortstops. Individual catchers come up short in the season and career rankings because of playing time.

If you think it through it makes perfect sense. If you start from replacement level, and can sign either an elite 1B or an elite catcher, you wind up with:

1. 150 games of great 1B, 12 games from his backups
2. 130 games of great catching, 32 games from his backups.

Choosing option #1 will win you more games every time, assuming each is equally great on a per game basis.


I agree with this. If a guy isn't playing it doesn't matter how good he is, he isn't helping you win. A catcher isn't "more great" when he plays simply because he plays less.

To keep everybody on equal footing as to how they're rated (i.e. catchers vs. the rest of the world), the solution is to go to some sort of WAR/games played stat. Basketball has already done this with Win Shares per Game, which is becoming a popular stat.
It seems to me that more than in-season durability it's career durability that hurts catchers; as WAR is a counting stat, it's going to be reflected there too.

The oft-injured Barry Larkin had 9057 PA, 17th ever among players with >50% of games at SS (he'll drop to 18th after Rollins has another 100 PA). On the other hand, that's more than all but 3 C in history.

Number of players with =>7000 PA, 50% of G at each position (another 28 players had => 7000 PA, but didn't play enough at any one position):

C 17
1B 53
2B 37
SS 47
3B 42
OF 153

So while WAR may make a 130 game C equal to a 150 game 1B in any one year, over the long-haul that may not be enough. IOW, a season-level WAR adjustment of 20 runs for a C over a 1B may make sense, but at a career-level the difference may be much higher per year.
   39. cercopithecus aethiops Posted: April 16, 2014 at 03:16 PM (#4687126)
37: crank up your sarcasm detector, please.
   40. Mirabelli Dictu (Chris McClinch) Posted: April 16, 2014 at 03:18 PM (#4687127)
No, it means the sports are different in how they distribute opportunities among the players. Trout can't ever be more than 1/9 of his team's offense, while James could be close to 100% if he and his teammates wanted. Trout could be better at his sport by producing per opportunity at say 3 standard deviations better than average while James produces at 2, but produce less win value because his sport gives him far fewer opportunities.


The question was obviously tongue-in-cheek, but this is also why LeBron and Durant are ludicrously underpaid by the standards of professional sports. No player in any other sport has the impact on odds of winning a championship that an MVP-caliber basketball player does, and yet they're making less money than Vernon Wells made last season.
   41. McCoy Posted: April 16, 2014 at 03:43 PM (#4687154)
Trout can't ever be more than 1/9 of his team's offense,

That isn't true. Trout can never be more than 1/9 of a lineup and at most he can get around 1/9th the team's PA but he most certainly can be worth more than 1/9 the team's offense.
   42. cercopithecus aethiops Posted: April 16, 2014 at 03:44 PM (#4687157)
Hell, they're making less money than Vernon Wells is making THIS season.
   43. Karl from NY Posted: April 16, 2014 at 04:50 PM (#4687244)
39: Put some actual sarcasm in your posts. :) It looked like an entirely reasonable question.
   44. cercopithecus aethiops Posted: April 16, 2014 at 05:05 PM (#4687261)
Well, at least one person realized I wasn't being literal, and he posted to that effect an hour and a half before you did, so there.
   45. cardsfanboy Posted: April 16, 2014 at 05:41 PM (#4687298)
In general, the average WAR for the sum total of a team's catchers will be about the same as for their first basemen or shortstops. Individual catchers come up short in the season and career rankings because of playing time.

If you think it through it makes perfect sense. If you start from replacement level, and can sign either an elite 1B or an elite catcher, you wind up with:

1. 150 games of great 1B, 12 games from his backups
2. 130 games of great catching, 32 games from his backups.

Choosing option #1 will win you more games every time, assuming each is equally great on a per game basis.


The article was more or less crap, and your point is very solid, but it doesn't really detract from the criticism about catcher and war. The biggest point is that the catcher inherently figure into more play than any other player on the diamond, including the pitcher, but his defensive value is still evaluated at near the same level as a shortstop. The argument is that catchers rpos isn't high enough or maybe their replacement level isn't low enough, or whatever, but it's still undervaluing them.

Johnny Bench, probably the greatest catcher of all time by the numbers in his ten year peak, he posted 59.9 war over 6216 plate appearances for a war per 600 pa of 5.78. Mike Schmidt arguably the greatest third baseman of all time and a contemporary of bench over his 10 year peak posted a war of 78.9 over 6409 pa for a war/600 of 7.38... and it doesn't matter who you look at, the top 5 best non-catcher position players of all time will pretty handidly beat the consensus best catchers of all time on a per plate/game rate of war. (Mike Piazza is probably, by war the best rate catcher of all time, and still gets around 6.2 war/600 in his 10/11 year peak.)

I think War is for the most part great, (and yes I prefer bb-ref over fangraphs by a huge margin) but I think it's weakness's become apparent when talking about pitching (both starters and relievers post 1980 especially) and catching (with minor quibbles on first base also) and the catching part is absolutely 100% about undervaluing the defensive aspect of the catcher...something that probably can never be quantified for past players.

But generally speaking in an average game you would expect the starting pitcher to be responsible for around 30-35% of the team's win, the defense on the diamond for around 10% and pitcher handling for around 5%(or something like that...percentages to be determined.---with the offense side of the equation being spread among all the batters in descending order from the batting order, regardless of position they play) One of the problems with looking at it the way I spelled it out here, is that this would mean the rpos for a catcher would be around 20-30 over the course of a 160 game season, and that would probably get a lot of people bent out of shape.

The best way to view catcher by war is just assume not all war is created equally, don't compare catchers war to anyone else other than catchers.

Does WAR still only measure a catcher's ability to throw out runners along with WP and PB?

Don't forget rPos... that is probably the biggest "defensive" advantage of catchers. Example Yadier Molina last year got an 8 rpos for roughly 136 games as a catcher, Asdrubal Cabrera played 136 games at shortstop and got 6 rpos. (don't know where the decimals are at, and Cabrera got 21 more pa than Yadier...but gives you a rough idea)


I think the primary issue with an article like this is that the author starts with the assumption that a team's stars deserve almost all of the credit for winning, and that when a team doesn't win the stars deserve to be dinged for their failure to win. (This is essentially the Cabrera/Trout argument.)


I would argue that the primary issue with this article, is that the writer thinks that the narrative 100% should fit the facts, and doesn't realize that in the context of journalism in regards to a sports story, the narrative is more important than the facts. He thinks Joe Dimaggio is the greatest living ballplayer, because it was repeated enough so it must be true. And these new fangled stats that undermine this are wrong as he can point to one or two people of the time that say it was wrong. Never mind that Mays and Mantle as greater than Dimmaggio was widely considered true long before the numbers revolution happened. (I am 44 years old, and have followed baseball since 1977, and I have never ever heard one single person consider Dimaggio as better than Mantle, Mays, Musial or Williams...about the only thing that has really changed in the past 20 years, is that more people are coming down on the Mays side of greatest centerfielder of all time than the Cobb side....but Dimaggio was never ever in the discussion)

   46. Sunday silence Posted: April 16, 2014 at 05:48 PM (#4687306)
can we get a little more elaboration on what is meant/defined by replacement level?

let's say for instance, "replacement level CF." do they look to set an actual number or percentage for both his off and his def? Like they artificially set replacement level at say 80% of what an MLB CF would bat; and then also they set it at say 75% of what he would be able to field?

or do they try to do it realistically in terms of what is currently available in the high minors? So in fact it could turn out that replacement level for CF might be a 90% of MLB offensive, but only 65% of his defense? I.e. these relative off/def factors could be out of whack with one another, because at this moment are a lot of decent hitters but few that can hit at the minor league level?

another way do it would be to take realistically what is available like the second method I suggest, but then make use an equation to divide it equally between off/def. Because it really doesnt matter which skill they are better it, both skills are related to one another so we take an average.

So even if there were decent hitters in AAAA, they would set it at both 87% of both off and def?

LIke what I am most concerned about is, if you were to take say the top 5 replacement CF in all of AAAA; would they a) more likely to be MLB avg at one skill (like hitting) and sub average at another; or would be they be more likely to be both 80% of MLB avg at both hitting and defense? I think it makes a difference in how you set it up.

thanks.
   47. Walt Davis Posted: April 16, 2014 at 05:53 PM (#4687309)
Nobody even noticed the bit about the catcher touching the ball on every pitch? Damn those Yankee pitchers were good although their walk rates were a bit extreme.

Anyway, if this is his takedown on WAR I can only imagine how brilliant his insight on economic statistics must be.

WAR of course has no way of accounting for Josh Gibson but Berra is "just" 6th in career WAR, 8th (t) in WAR7 and JAWS puts him 6th overall. The career ranking is partly due to being 1-3 seasons worth of PAs behind Carter and the two Pudges and he's all of .1 WAR behind Piazza. Berra is rated with the most career WAR and tied for WAR7 (Cochrane) for a C at the time of his retirement.

And surely that was the first time that Allen Barra has ever been cited as if he's an unimpeachable expert on the subject.

The point I tend to make about WAR and Cs (at least for these sorts of arguments) is that WAR is a measure of value not of "greatness" per se. Cs can't produce as much value for the durability reasons cited. If we ignore ARod and Yount (too much time away from SS), the #6 career SS is Appling, a bit ahead of Vaughan and Jeter (a good comp for Berra in many ways). The #6 3B is among Brooks, Beltre, Santo with each admittedly probably a bit below Berra in "greatness" -- but #5 3B Chipper/Brett/Boggs seem around Berra's level.

#6 CF is between Griffey and DiMaggio (a big gap to #4) who might both be a bit above Berra (maybe not) but #7 is Snider who I'd probably put below Berra.

In terms of "greatness", Berra as Jeter, Santo, Griffey seems about right to me. At the time of his retirement, Santo had the 2nd most career WAR for a 3B (unless Brooks was already a bit ahead but still playing). Appling/Vaughan were the best of the liveball era until Ozzie/Ripken (unless you want to count Banks/Yount), DiMaggio was the best of the liveball era until Mantle/Mays. And, in terms of greatness, I don't have a particular problem moving Berra ahead of the two Pudges.
   48. Sunday silence Posted: April 16, 2014 at 06:05 PM (#4687320)

The best way to view catcher by war is just assume not all war is created equally, don't compare catchers war to anyone else other than catchers.


It's a very good post, but I thought you were going to tell us whether you indeed feel that catcher WAR is undervaluing catchers. You brought up this point in the first para. so I thought you were going to go back to it. Perhaps you are agnostic on that issue?

And I would be asking in terms of rate, i.e. say WAR per 600 PA, just as in the example you give w/ Bench/Schmidt. It seems to me that there is a certain reality that when you look at how managers set up lineups that if in fact 3B were inherently more valuable then catchers (because apparently C dont have as much to contribute defensively, as silly as that sounds) then managers would punt the defense and just find catchers who could hit the hell out of the ball, and not be able to throw out baserunners.

WHich seems counter intuitive to me. It seems that the best teams and managers already have a very good idea of just how much def and off. they need at each position and make good selections accordingly. If they are indeed operating ideally, or say 60% of teams do have the most ideal candidate at each position, then each player should be able to contribute the same WAR per rate. A SS might have more def WAR to give, a 3B more off, but in the end each position should be ideally chosen to maximize the overall wAR.

Or so it seems to me, with the upshoot being that all positional players should be about equal value in WAR and hence Catcher WAR must undervalued.

But not sure how you come out on that.
   49. cardsfanboy Posted: April 16, 2014 at 06:06 PM (#4687321)
I'm probably showing my ignorance here, but should WAR be this way? Isn't it possible that different positions have different replacement levels? If your 5 WAR catcher gets hurt isn't it going to be harder to find a good replacement than if your 5 WAR firstbaseman gets injured?


In response to the first question you asked...it's debatable. I've argued that it shouldn't and the rpos adjustment supposedly for catchers does assign a little more value to a catcher on a per game basis than other positions, but it's probably not enough still.

In response to the second question, War is figured based upon average and a replacement level adjustment is made, rpos is the only thing that differentiate between positions, and it does (from my understanding, you would have to ask Arom for sure) assign a lower replacement level to catchers than other players.
War is basically a math construct, it compares the players batting relative to league average batting(no positional adjustment) to come up with rbat(batting runs above average) and rbase(baserunning above average) and rdp(double play avoidance above average) and rfield(fielding runs above average) all of these numbers are above average for the league, not position adjusted. Add those together along with a rpos(positional adjustment that accounts scarcity of position) and rrep (average number of runs an average player is over replacement) and you get roughly the number of runs better than replacement which is used to figure war...


In regards to the third part. In theory, no. If war is set up correctly then a 5 war catcher should be replaced equally as easy as a 5 war first baseman (ignoring the practicalities of an individual teams bench construction and the true availability of a replacement player.) The debate is over whether or not war is set up correctly to recognize what a true 5 war catcher looks like versus a true 5 war first baseman.

But...
In general, the average WAR for the sum total of a team's catchers will be about the same as for their first basemen or shortstops.


The problem with this concept isn't about the replacement level, it's about the average value on a per game basis. Nobody denies that a pitcher figures into more of the game result than other players, but it seems hard to convince people that catchers have more say in the result than other players. (although that is changing)
In a 160 game season, if you sum all the war for innings pitched by a team it would dwarf any individual position, in my opinion, I think that it should also be true-to a lesser extent- for catchers.
   50. cardsfanboy Posted: April 16, 2014 at 06:07 PM (#4687322)
It's a very good post, but I thought you were going to tell us whether you indeed feel that catcher WAR is undervaluing catchers. You brought up this point in the first para. so I thought you were going to go back to it. Perhaps you are agnostic on that issue?


Far from agnostic on the issue, but I was typing too long of a post and wanted to break it down.
   51. Sunday silence Posted: April 16, 2014 at 06:17 PM (#4687327)
#6 CF is between Griffey and DiMaggio (a big gap to #4) who might both be a bit above Berra (maybe not) but #7 is Snider who I'd probably put below Berra.



dont you have to question a system that comes up with Snider as no. 7? Have you ever compared the raw fielding numbers of him and Richie Ashburn? It is a vast discepancy. I dont claim to know how to adjust defensive stats for contours of Ebbets field, and I suppose it's possible that Carl Furillo was stealing a lot of flyballs from Snider. But even if you cant quantify it accurately, if you begin to make allowances for what is reasonable variability in fielding range it is hard to see how Snider is better than Ashburn.

There are any number of great defensive CFs that seem to be shortchanged by whatever methods we are using to evaluate players before modern SABR eras. Zack Wheat got to a lot balls. Speaker got to hell of a lot of balls. Dom DiMaggio was very good, but I dont think his career was that long.
   52. Sunday silence Posted: April 16, 2014 at 06:21 PM (#4687331)
Nobody denies that a pitcher figures into more of the game result than other players...


Just to be even more contrarian than usual; is this actually provable? I know I tend to believe it, but I am questioning just about everything at this point.

How would one go about actually proving this, as a starting point? Do you look at the variability in pitchers ERA+ and then say that it must be this large because the team fielding cant be this different? Or do you look at something like wins/losses and do the same thing?
   53. cardsfanboy Posted: April 16, 2014 at 06:25 PM (#4687336)
can we get a little more elaboration on what is meant/defined by replacement level?


Again Arom or Walt are probably the best to elaborate, replacement level is a generic concept that basically states if you fill a team up with replacement level players at all positions, your team would have roughly a .294 winning percentage. (47-115 record) This is a math construct more than it is an actual replacement player.

let's say for instance, "replacement level CF." do they look to set an actual number or percentage for both his off and his def? Like they artificially set replacement level at say 80% of what an MLB CF would bat; and then also they set it at say 75% of what he would be able to field?

or do they try to do it realistically in terms of what is currently available in the high minors? So in fact it could turn out that replacement level for CF might be a 90% of MLB offensive, but only 65% of his defense? I.e. these relative off/def factors could be out of whack with one another, because at this moment are a lot of decent hitters but few that can hit at the minor league level?

another way do it would be to take realistically what is available like the second method I suggest, but then make use an equation to divide it equally between off/def. Because it really doesnt matter which skill they are better it, both skills are related to one another so we take an average.

So even if there were decent hitters in AAAA, they would set it at both 87% of both off and def?

LIke what I am most concerned about is, if you were to take say the top 5 replacement CF in all of AAAA; would they a) more likely to be MLB avg at one skill (like hitting) and sub average at another; or would be they be more likely to be both 80% of MLB avg at both hitting and defense? I think it makes a difference in how you set it up.


War's construct is based upon average and then adjusting it to replacement level. It's easy to figure out what an average player does. (any number I put in here is for illustrative purposes) You can look at league splits and see what an average player does at each position. So war basically figures out what the offensive value of each position, and looks at the difference per plate appearance. So let's say that the average first baseman in baseball has an rBat of 70 runs per 600 pa...and the average shortstop in baseball has 50 runs per pa, then you would create an rpos difference between the positions of 20 runs(say +10 for the shortstop, and -10 for the first baseman).

Base running and fielding is figured out relative to league average runners and fielders. After you add those numbers up, you are going to have what the player compares to average....to apply the replacement level you just add in the replacement factor. (which is a constant based upon plate appearances...I think)


   54. cardsfanboy Posted: April 16, 2014 at 06:40 PM (#4687350)
dont you have to question a system that comes up with Snider as no. 7? Have you ever compared the raw fielding numbers of him and Richie Ashburn? It is a vast discepancy. I dont claim to know how to adjust defensive stats for contours of Ebbets field, and I suppose it's possible that Carl Furillo was stealing a lot of flyballs from Snider. But even if you cant quantify it accurately, if you begin to make allowances for what is reasonable variability in fielding range it is hard to see how Snider is better than Ashburn.

There are any number of great defensive CFs that seem to be shortchanged by whatever methods we are using to evaluate players before modern SABR eras. Zack Wheat got to a lot balls. Speaker got to hell of a lot of balls. Dom DiMaggio was very good, but I dont think his career was that long.


First off Duke Snider was a significantly better hitter than Ashburn. (140 ops+ vs 111 or 390 rbat vs 197) the fact that Ashburn makes the cumulative total war close is a testament to his great defense.

But we've had this discussion before, and looking at plays made is not a good way to look at defense, you have to compensate for a lot of things before you begin to look for plays made. If you have two teams playing a game, with the home team losing(to guarantee 27 outs) and neither team has a strikeout, error, double play or a homerun hit, yet one team allows 20 hits and another team allows 5 hits, they would still have the exact same number of plays made.

All the advance systems out there adjust for 1. balls in play 2. strikeouts 3. homeruns 4. park 5. handedness of the pitching staff 6. double plays etc...before even looking at the number of plays a team makes (and yes they start at the team level ahead of the individual player, at least the good ones do)
   55. cardsfanboy Posted: April 16, 2014 at 07:01 PM (#4687362)
Nobody denies that a pitcher figures into more of the game result than other players...


Just to be even more contrarian than usual; is this actually provable? I know I tend to believe it, but I am questioning just about everything at this point.

How would one go about actually proving this, as a starting point? Do you look at the variability in pitchers ERA+ and then say that it must be this large because the team fielding cant be this different? Or do you look at something like wins/losses and do the same thing?


Not sure how to model it, but based upon some things we do know about studying the past.
1. Pitchers are as much responsible for strikeouts as the batter. (you can debate the percentage of the responsibility as much as you want, but history does show both the batter and the pitcher have a strong correlation with the results, indicating both have a strong say in this particular result)
2. Pitchers are as much responsible for walks as the batter. (see above)
3. Pitchers are partially responsible for type of balls hit off of them (flyball/linedrive/groundball, infield popups)
4. Pitchers bat(assume an NL team)
5. Pitchers are as important as preventing the steal as a catcher.

So we have pitchers batting which offensively makes them potentially equally valuable as any individual hitter (adjusted for batting order) so the real difference is how much defensive value do the individual positions have.

You have pitchers who have a large say in what happens to balls put in play or not put in play, an average shortstop might be involved in 5 plays per game, corner outfielder maybe 1, yet a pitcher on average strikes out 6+ batters a game. Add in that they have a limited control on what the ball does after it leaves the bat and it's hard not to assume that on a per game basis pitchers influence the game more than any other position.

   56. tshipman Posted: April 16, 2014 at 08:33 PM (#4687411)
In general, the average WAR for the sum total of a team's catchers will be about the same as for their first basemen or shortstops.


Stopping you here. This is a flaw in WAR. Catchers have a greater defensive impact than first basemen (and shortstops). If the average WAR/PA for all catchers is equal to the average WAR/PA for all 1B, that is a flaw in WAR.

Individual catchers come up short in the season and career rankings because of playing time.

If you think it through it makes perfect sense. If you start from replacement level, and can sign either an elite 1B or an elite catcher, you wind up with:

1. 150 games of great 1B, 12 games from his backups
2. 130 games of great catching, 32 games from his backups.

Choosing option #1 will win you more games every time, assuming each is equally great on a per game basis.


That doesn't actually make sense, because it's not how baseball has operated. WAR wants me to believe that Lou Whitaker and Johnny Bench had equally valuable careers. I loved the hell out of Sweet Lou's game, but that assessment of value is crazy.
   57. cercopithecus aethiops Posted: April 16, 2014 at 08:47 PM (#4687424)
Catchers have a greater defensive impact than first basemen (and shortstops). If the average WAR/PA for all catchers is equal to the average WAR/PA for all 1B, that is a flaw in WAR.


Sorry, but why exactly? After all, a whole lot of catcher PAs get sucked up by guys who have virtually no offensive value. Sounds like you want to adjust for position twice.

WAR wants me to believe that Lou Whitaker and Johnny Bench had equally valuable careers.


Whitaker needed 1300 more PA to produce his value, and 21% of Bench's career was played at positions other than catcher.
   58. tshipman Posted: April 16, 2014 at 08:57 PM (#4687429)
Sorry, but why exactly? After all, a whole lot of catcher PAs get sucked up by guys who have virtually no offensive value. Sounds like you want to adjust for position twice.


By definition, offensive environment should be neutral--the replacement level for 1B offense should be much higher than replacement level for C offense. If that is not correct, then either 1B replacement level offense is too low or C replacement level offense is too high.
   59. cardsfanboy Posted: April 16, 2014 at 09:07 PM (#4687439)
Sorry, but why exactly? After all, a whole lot of catcher PAs get sucked up by guys who have virtually no offensive value. Sounds like you want to adjust for position twice.


Isn't that evidence that there is probably something defensively that is being missed?

The ultimate point is though that catchers probably do have a larger affect on the defensive side of the equation that is traditionally considered by advanced stats.
   60. I Helped Patrick McGoohan Escape Posted: April 16, 2014 at 10:30 PM (#4687488)
#1 wins the thread by a country mile.
   61. BDC Posted: April 17, 2014 at 09:36 AM (#4687594)
catchers probably do have a larger affect on the defensive side of the equation

That could be, but the general principle that total WAR at catcher roughly equals total WAR at other positions could also be true.

After all, you have to have a catcher, or as Casey Stengel once remarked, you are going to have a lot of passed balls. If the pool of catchers shows less defensive variation than the received belief would hold, then the value of a great catcher is going to consist mostly of oWAR. It may be very hard to learn to play catcher, and it may be a position that is central to a team's defense, but it's still only one position.

IOW if Lou Whitaker is far ahead of a replacement 2B offensively and defensively, and Johnny Bench is very far ahead of a replacement catcher offensively but only "somewhat" ahead defensively, WAR would rightly see them as about equally valuable to a ballclub.

The problem is in defining the "somewhat." You could see that Johnny Bench was an astonishing athlete, and you could see that Pat Corrales was a nondescript plugger. And obviously Bench was a hilariously better hitter. But how much differential impact on a ballgame did it have defensively to have Corrales behind the plate instead of Bench?
   62. Ron J2 Posted: April 17, 2014 at 09:44 AM (#4687605)
#46 Different assumptions are possible. When Bill James came up with the concept he set the level so that a team of replacement level players and pitchers would go 52-110. Keith Woolner's actually done the math as to where replacement level "should" be assuming a normal distribution of talent and it came out a little lower than that/

Right now IIRC WAR assumes a team of replacement level players would play ~ .280 ball. Doesn't care how you get there.
   63. Karl from NY Posted: April 17, 2014 at 12:19 PM (#4687792)
Right now IIRC WAR assumes a team of replacement level players would play ~ .280 ball.
The question from #46 is, from where does WAR get that assumption? Is there a calculation that actual replacement players for this year/league/whatever would play at ~ .280, or is it a historical average of some sort, or is it an eyeball guess like Win Shares' replacement level?
   64. Ron J2 Posted: April 17, 2014 at 12:41 PM (#4687823)
#63 I think you can get there a couple of ways. Keith Woolner's research suggests that a complete replacement level team would in fact play a little worse that 52-110 (this assumes a normal distribution of baseball talent). And if you look at expansion teams, you'll see that they generally had a few players who obviously were better than replacement level. Suggesting that James' starting point was a little high. James initially just threw out the 52-110 for replacement level because that's what most expansion teams played at (remember that while he didn't call it replacement level, he first used the concept in 1982. He didn't do any serious thinking about the matter until linear weights came along)



I don't know what arguments were ultimately persuasive. I do know that BBRef and Fangraphs now use a fairly similar (if not actually identical) definition of replacement level. And I see from another post here that my memory of where it's set was ~a game per season low.

   65. DJS and the Infinite Sadness Posted: April 17, 2014 at 01:37 PM (#4687899)
I think my favorite part of this is the author thinks his article is so groundbreaking that it should inspire a second-look at WAR from the sabermetric community.

"@NateSilver538 NS, I know you'll hate this - but I think it may spur a long-needed review of the value of #WAR http://bit.ly/1nqiRi6 "
   66. JE (Jason) Posted: April 17, 2014 at 03:22 PM (#4688004)
I think my favorite part of this is the author thinks his article is so groundbreaking that it should inspire a second-look at WAR from the sabermetric community.

He also thought it would spur a groundbreaking conversation with WFAN's Joe [Benigno] and Evan [Roberts].
   67. CrosbyBird Posted: April 17, 2014 at 08:00 PM (#4688186)
I think WAR's treatment of catchers is a pretty good argument against it's credibility.

As a Mets fan, I'm more inclined to think that it rates catchers properly. I've watched a bunch of a seasons where one of the team's stars doesn't start 25-30 times. Also, non-starting catchers are pretty heavily selected for defense, and the general offensive level of catchers is relatively low, so the replacement level catcher might well be closer to average rather than further away. Catchers probably aren't as valuable as other players relative to their replacements (which is what WAR measures).

Criticizing WAR for "underrating" catchers because the greatest catchers in baseball history rank relatively low seems misplaced. WAR isn't trying to identify the best player on an absolute scale. If for some reason, it suddenly became a lot easier to find high-quality replacement-level CF, Mike Trout wouldn't be any worse a player, but he'd have fewer wins over replacement.
   68. zenbitz Posted: April 17, 2014 at 09:08 PM (#4688230)
I think CB has it right. People are confusing strict value (estimated number of runs/wins added) with scarcity and cost to replace.
   69. Karl from NY Posted: April 18, 2014 at 10:37 AM (#4688416)
The catcher thing becomes pretty obvious if you compare to other sports that have more highly specialized roles. Think kicker in football. The best are maybe 5% more accurate than a replacement, which per game on say 3 FG attempts comes to 0.45 points over replacement. Of course the best QBs/WRs/RBs contribute way more than 0.45 points per game. The kicker just can't create as much value because of his limited role and so will have much less football WAR. The limited role of catchers works the same way. It's not a flaw in WAR. Catchers really do create less value by playing 130 games instead of 150.

Something like the HOF should adjust for this and set a lower WAR bar for catchers. WAR shouldn't in itself.

Trying to improve catcher WAR by giving additional credit for their defense is trying to combine two wrongs to make a right. There's no reason to believe that a catcher's pitch framing is equal in value to the playing time lost to their limits of physical endurance. Those are completely separate concepts.
   70. cardsfanboy Posted: April 18, 2014 at 05:43 PM (#4688754)
Trying to improve catcher WAR by giving additional credit for their defense is trying to combine two wrongs to make a right. There's no reason to believe that a catcher's pitch framing is equal in value to the playing time lost to their limits of physical endurance. Those are completely separate concepts.


Have to disagree. If a catcher provides more value on a per game/inning played basis than a position player (and I think that is pretty obvious) then War should reflect that. War is set up so that roughly all position players are worth the same on a per game basis* and pitchers are worth significantly more on a per game basis. I don't see what is wrong with believing that the catchers active ability to call the game, pitch framining etc.. have a larger impact on the game than passive defenders.

Most of these "all encompassing stats" make one basic assumption, and that is MLB has already selected positions players value as being roughly equal all across the 8 non-pitcher positions. I'm not sure I agree with that assessment.
   71. Publius Publicola Posted: April 18, 2014 at 06:10 PM (#4688773)
Now that plate blocking is no longer a necessary skill for a catcher, it will be interesting to see how the players selected to play it evolve. We might see more speed at catcher, for instance. Or smaller players like Biggio will tend to stick instead of being moved elsewhere.
   72. DanG Posted: April 18, 2014 at 06:20 PM (#4688779)

Most of these "all encompassing stats" make one basic assumption, and that is MLB has already selected positions players value as being roughly equal all across the 8 non-pitcher positions. I'm not sure I agree with that assessment.
Indeed, the treatment of Biggio and others indicates the assumption is wrong. The rigors of the position tend to make teams avoid wasting very talented offensive players by making them catchers. Perhaps a study of players who established themselves as MLB catchers and were then switched away from there would prove enlightening. I'm thinking of guys like Downing, Surhoff, Sweeney, Chance, Inge, Cross, Daly, etc.
   73. tshipman Posted: April 18, 2014 at 06:26 PM (#4688786)
The catcher thing becomes pretty obvious if you compare to other sports that have more highly specialized roles. Think kicker in football. The best are maybe 5% more accurate than a replacement, which per game on say 3 FG attempts comes to 0.45 points over replacement. Of course the best QBs/WRs/RBs contribute way more than 0.45 points per game. The kicker just can't create as much value because of his limited role and so will have much less football WAR. The limited role of catchers works the same way. It's not a flaw in WAR. Catchers really do create less value by playing 130 games instead of 150.


You are using almost exactly the wrong analogy here. First basemen are more like placekickers than catchers are. Catchers are involved in more plays than any other position on the field. I'm not saying catchers are equal in value to quarterbacks. It's more like comparing left tackles to wide receivers (catchers being the LTs).
   74. cardsfanboy Posted: April 18, 2014 at 06:49 PM (#4688799)
You are using almost exactly the wrong analogy here. First basemen are more like placekickers than catchers are. Catchers are involved in more plays than any other position on the field. I'm not saying catchers are equal in value to quarterbacks. It's more like comparing left tackles to wide receivers (catchers being the LTs).


If he wanted to use a football analogy, then comparing a catcher to the center would have been more accurate, if the point was to show how relatively little importance the catcher has, even though he is involved in every play. Mind you, I wouldn't use that analogy, unless you change it where it's the center who was responsible for audibles and he was the offensive coordinator. It's still up to the quarterback/pitcher to execute, but for most teams, historically, it's the catcher who calls the game, who directs the defense(less so nowadays than in the past of course), who is responsible for keeping the pitcher "centered" and who is equally responsible for stopping the running game.
   75. DanG Posted: April 18, 2014 at 06:50 PM (#4688801)
Then there's the question of WAR's valuation of deadball catchers. How could Bill Bergen exist if WAR is valuing him correctly? This list shows the many deadball catchers who were bad hitters. All caught at least 800 games:

Rk              Player OPSWAR/pos   PA From   To    G
1          Bill Bergen   21   
-13.5 3228 1901 1911  947
2    Malachi Kittridge   56     2.8 4454 1890 1906 1216
3        Bill Killefer   63     3.5 3401 1909 1921 1035
4       Billy Sullivan   63     6.0 3981 1899 1916 1147
5          Otto Miller   67     0.1 3050 1910 1922  926
6        Oscar Stanage   69     6.0 3845 1906 1925 1096
7            Red Dooin   72     3.8 4271 1902 1916 1290
8           Lou Criger   72    10.7 3619 1896 1912 1012
9          Jack Warner   73    10.6 3827 1895 1908 1074
10      Eddie Ainsmith   76     5.7 3416 1910 1924 1080
11      Frank Bowerman   77     6.9 3659 1895 1909 1037
13        Jack OConnor   79     8.2 5788 1887 1910 1452
14        Bill Rariden   81     8.7 3311 1909 1920  982
15       George Gibson   81    15.1 4189 1905 1918 1213 

In the 90+ years since then, these are the catchers with the lowest OPS+, minimum 800 G at C and 3200 PA:

Rk           Player OPSWAR/pos   PA From   To    G
1      Mike Matheny   65    
-0.3 4287 1994 2006 1305
2    Kirt Manwaring   69     5.2 3338 1987 1999 1008
3       Luke Sewell   70     3.8 6044 1921 1942 1630
4    Bruce Benedict   71     7.3 3295 1978 1989  982
5        Mike Tresh   71     3.1 3640 1938 1949 1027
6       Joe Girardi   72     5.7 4535 1989 2003 1277
7     John Flaherty   74     1.6 3640 1992 2005 1047
8      Buck Rodgers   74     3.4 3351 1961 1969  932
9         Jim Hegan   74     3.7 5320 1941 1960 1666
10   Rollie Hemsley   74     3.7 5511 1928 1947 1593 
   76. BDC Posted: April 18, 2014 at 09:00 PM (#4688905)
I suspect that Bill Bergen was just wildly overrated by his managers, like some primeval Jeff Mathis. His teams were generally terrible, and there's every reason to think he helped them be terrible.
   77. Pasta-diving Jeter (jmac66) Posted: April 18, 2014 at 09:04 PM (#4688908)
I suspect that Bill Bergen was just wildly overrated by his managers, like some primeval Jeff Mathis.Frenchy

FFIFA
   78. Kiko Sakata Posted: April 18, 2014 at 09:48 PM (#4688954)
If a catcher provides more value on a per game/inning played basis than a position player (and I think that is pretty obvious) then War should reflect that. War is set up so that roughly all position players are worth the same on a per game basis* and pitchers are worth significantly more on a per game basis. I don't see what is wrong with believing that the catchers active ability to call the game, pitch framining etc.. have a larger impact on the game than passive defenders.


But the question isn't how much absolute value there is in pitch framing and pitch calling, but how much marginal value there is in being very good at those things versus the ability of replacement-level catchers. If Johnny Bench wasn't any better at pitch framing and pitch calling than Barry Foote - to name a marginal catcher from that era whose name I remember (because I think I had a half-dozen of his baseball cards), then WAR is valuing Bench just fine. If Jose Molina is as great at pitch framing as the studies suggest, then yes, WAR is specifically under-valuing Jose Molina, but it's doing so at the expense of over-valuing whatever catchers are lousy at pitch framing and not getting dinged for it by WAR.
   79. cardsfanboy Posted: April 18, 2014 at 10:00 PM (#4688963)
But the question isn't how much absolute value there is in pitch framing and pitch calling, but how much marginal value there is in being very good at those things versus the ability of replacement-level catchers. If Johnny Bench wasn't any better at pitch framing and pitch calling than Barry Foote - to name a marginal catcher from that era whose name I remember (because I think I had a half-dozen of his baseball cards), then WAR is valuing Bench just fine. If Jose Molina is as great at pitch framing as the studies suggest, then yes, WAR is specifically under-valuing Jose Molina, but it's doing so at the expense of over-valuing whatever catchers are lousy at pitch framing and not getting dinged for it by WAR.


I understand that part, but again I don't agree with it. It's fairly simple question, why is a pitcher worth more on a per inning/game basis than other position players, and why can't a catcher also be worth more for the same reason?

Using your argument, every player on the team should have the exact same per game value, including the pitcher, which would mean that the best starting pitcher in baseball would be worth maybe 1 war over the course of a season.

I don't see it that way. Again, roughly speaking game value is divided into offense and defense(for sake of argument we'll go with 50/50)
Now the offense side of the equation has 9 batters, each batter is worth 1/9 of that 50% (not true value, but potential value and of course adjusting for plate appearances) Meanwhile on the defensive side of the equation, the pitcher is getting roughly 35 % of the value, and the defense is getting roughly 15% of the value split up between the nine players based upon chances.... Isn't it arguable, if not likely, that the catcher's handling of the pitcher should suck up some of that pitcher value and maybe even some of that defensive value?


   80. BDC Posted: April 18, 2014 at 10:33 PM (#4688977)
Makes sense, fanboy, but if the defensive differences among catchers are small while the defensive differences among pitchers are large, it makes little sense to credit the existence of catching with a large amount of defensive value. All teams find a complement of catchers at all times. The value that they get out of them may reside largely in their offense and their durability. (I'm actually not arguing that it must as much as trying to understand the argument by stating it.)

In a sense, catching might be SO important that a catcher absolutely must clear a bar to catch at all, guaranteeing that they're all in a weird sense defensively above average and very few stand out to any degree. Yet at SS or CF you can be pretty bad (Jeter, Bernie) if you win those runs back with the bat, so the defensive variation can be greater. A "Jeter of Catchers" would be intolerable. You don't want every fifth pitch to go past a diving catcher :)
   81. Kiko Sakata Posted: April 18, 2014 at 10:42 PM (#4688982)
Using your argument, every player on the team should have the exact same per game value, including the pitcher, which would mean that the best starting pitcher in baseball would be worth maybe 1 war over the course of a season.


Not at all. If the Cubs have to scrounge around and bring up Casey Coleman to pitch, the gap between Casey Coleman and a real major-league pitcher at the act of actually pitching is HUGE. If the Cubs have to scrounge around and bring back Koyie Hill to catch, Koyie Hill is perfectly cromulent at calling a game and framing pitches: all that the Cubs lose by having Koyie Hill as their catcher is the fact that he can't hit major-league pitching worth a damn. And the difference between what actual major-league catchers hit and what Koyie Hill would hit is already being accounted for in WAR.

I think you're confusing WAR with Win Shares. WAR doesn't make any explicit assumptions about the pitching/fielding breakdown. WAR begins with a WAA framework - wins above average - and then tacks on a replacement adjustment. All fielding is measured relative to average at the position. The average value of pitch framing and pitch calling by major-league catchers is zero, by definition. So, in the aggregate, WAR is not under-valuing catchers for leaving out skills that an average catcher is average at. Individual catchers may be under- or over-rated by WAR if they are especially good or bad at the things that WAR isn't measuring, and I'm fine with bringing those into the conversation on a case by case basis, but for every catcher who's above average at pitch calling and pitch framing, there's one who's below average. Major-league catchers don't all come from Lake Woebegone.
   82. Sunday silence Posted: April 19, 2014 at 02:06 AM (#4689021)
looking at plays made is not a good way to look at defense, you have to compensate for a lot of things before you begin to look for plays made. If you have two teams playing a game, with the home team losing(to guarantee 27 outs) and neither team has a strikeout, error, double play or a homerun hit, yet one team allows 20 hits and another team allows 5 hits, they would still have the exact same number of plays made.


Correct, but the problem is that your argument is entirely based upon one team allowing lots of hits/balls to the OF OVER THE COURSE OF SEASON. When in fact, we know that this not true, and has never been true in MLB for any two teams for any two year. It just doesnt happen that way. I forgot how to find defensive efficiency but I'm guessing if you look up def eff. for the Dogers and Phillies of 50s you are not going to find much more than 1% difference. Same with pitching ERA+ and by extrapolation: balls hit to the OF.

WHat if instead of saying one team allowed 20 hits and another 5 we say this:

One team allowed 5.1 hits and the other allowed 5.0 hits per game on a seasonal basis; one team allowed 650 balls to the OF and the other allowed 620 balls to the OF.

And one guy catches 75 more balls than the other guy, and most of those are in the gap. What would you say then?

Your argument is on the same level as Bill James saying he doesnt see how Evans or Clemente or whomever could possibly worth 25 runs on defense because, if you multiply by 8 that's 200 runs and that doesnt seem possible.

of course it's not possible, because NO TEAM CAN FIELD 8 stellar defensive players! How stoopid is that? He makes up a completely non sensical hypothetical that has never happened in the entire history of baseball (OK maybe the 1941 Brooklyn Dodgers had 8 great players or the 1911 Athletics) and that's his reasoning that no OF can be +25 runs on def...
   83. Sunday silence Posted: April 19, 2014 at 02:11 AM (#4689022)
Yet at SS or CF you can be pretty bad (Jeter, Bernie) if you win those runs back with the bat, so the defensive variation can be greater. A "Jeter of Catchers" would be intolerable.[/quote

Can you explain why a good hitting catcher couldnt in theory make up for his bad defense? I fail to see how you can be so sure of this. What if in fact it turns out that Piazza was essentially the same thing as "the Jeter of catchers?" I dont know if he was or wasnt, but yo seem perfectly convinced that he could not be. How can you be so sure?
   84. Sunday silence Posted: April 19, 2014 at 02:28 AM (#4689023)
I wouldn't use that analogy, unless you change it where it's the center who was responsible for audibles and he was the offensive coordinator.


in fact in modern football that is exactly what the center does. He makes what are called "line calls" he adjusts the blocking assignments based upon variations in the defensive front 7. This job usually (not not always) falls on the center because he is right in the middle of the off line.

Of course this thing is never seen on tv coverage and lots of fans have no idea that it exists. he probably makes more audibles in that sense then the QB since he is making line calls on pretty much every play.
   85. Fancy Pants Handles lap changes with class Posted: April 19, 2014 at 08:46 AM (#4689034)
I understand that part, but again I don't agree with it. It's fairly simple question, why is a pitcher worth more on a per inning/game basis than other position players, and why can't a catcher also be worth more for the same reason?

Using your argument, every player on the team should have the exact same per game value, including the pitcher, which would mean that the best starting pitcher in baseball would be worth maybe 1 war over the course of a season.

To add to Kiko's response a little. The issue is, we know exactly how much impact a pitcher has on the run scoring of the other team. If your #1 with a sub-3 ERA goes down, and you have to bring in a guy with an ERA over 5, it is obvious how much that hurts you.

If catcher's were as important as pitchers, due to pitch calling, and handling the ball every play and whatever, we would expect that those runs saved/cost would show up somewhere in a meaningful way. But whatever effect there is, when you examine things like CERA and the like, is basically less than statistical noise. There is simply no endemic of catcher's going down, and team RA going up.

Or to put it another way, say the League for American Athlete's Arms lobbies congress to outlaw human pitching, and MLB has to put in pitching machines instead of pitchers for every game. Those machines all throw the exact same pitch all the time. Those machines would still be as "involved" in play as today's pitchers are, but obviously their value would be zero.

It's not the number of opportunities that matters, it's what you do with them. A fielder who fields 999 of 1000 popups, when his peers would have grabbed 998 if tthm, isn't more valuable, then the player who in the same playing time, fields 70 of 100 line drives, when his peers are getting 50.
   86. cardsfanboy Posted: April 19, 2014 at 09:40 AM (#4689044)
Makes sense, fanboy, but if the defensive differences among catchers are small while the defensive differences among pitchers are large, it makes little sense to credit the existence of catching with a large amount of defensive value. All teams find a complement of catchers at all times. The value that they get out of them may reside largely in their offense and their durability. (I'm actually not arguing that it must as much as trying to understand the argument by stating it.)


And nobody is actually saying it's huge or anything, I'm just saying it's more than likely there.

Let's say you have a team of 9 players, and 9 players only, and they play every inning at the one position they occupy and every player is perfectly average overall. War has it set up so that your 7 position players would be worth roughly 2.0 War. Your catcher is worthy 2.1 war and your pitcher would be worth around 14 war. (War does have built into it, a very slight catchers bonus beyond what the raw numbers would say, already...at least according to Sean or Arom from last time we had this conversation---unless I misunderstood what they were saying)

The argument I'm making is that it's reasonable to say that in reality your position players are probably worth 2.0 war, but that the pitcher should probably be 13.5 and the catcher 2.6 (or something like that)

Just look at the studies on pitch framing, even though there is (massive) flaws in their system, they are coming up with numbers that a great defensive catcher is responsible for up to 20 runs a year. My point on is to point out, that there is tangible evidence that the catcher does affect the pitches/count and that is still ignoring the "second manager" on the field job description of the catcher.


To add to Kiko's response a little. The issue is, we know exactly how much impact a pitcher has on the run scoring of the other team. If your #1 with a sub-3 ERA goes down, and you have to bring in a guy with an ERA over 5, it is obvious how much that hurts you.


We absolutely do not know this, we have a fairly good idea, but considering that the two systems of war massively disagree on this particular point, I think it's fair to say that we do not have exact knowledge on the impact a pitcher has on run scoring.

If catcher's were as important as pitchers, due to pitch calling, and handling the ball every play and whatever, we would expect that those runs saved/cost would show up somewhere in a meaningful way. But whatever effect there is, when you examine things like CERA and the like, is basically less than statistical noise. There is simply no endemic of catcher's going down, and team RA going up.


I have absolutely never, ever said or implied that catchers were as important as pitchers. I have consistently and constantly stated, that we do not have an idea of how much of an impact that they have on the game, but that it is reasonable to think they have more of an impact on the defensive side of the game than other players.

As to where it would show up, again the pitch framing studies have already shown that it does show up, we just didn't have the necessary data to discover it yet.

   87. cardsfanboy Posted: April 19, 2014 at 10:07 AM (#4689047)
Correct, but the problem is that your argument is entirely based upon one team allowing lots of hits/balls to the OF OVER THE COURSE OF SEASON. When in fact, we know that this not true, and has never been true in MLB for any two teams for any two year. It just doesnt happen that way. I forgot how to find defensive efficiency but I'm guessing if you look up def eff. for the Dogers and Phillies of 50s you are not going to find much more than 1% difference. Same with pitching ERA+ and by extrapolation: balls hit to the OF.


The argument is designed to illustrate a point more obviously. Range factor has been a joke stat, that Bill James himself wish he could take back. It is utterly useless stat. You cannot, and never ever should point to defense and range factor.

For years prior to play by play data, you have to model defense by starting at the team level. Every single advanced defensive system that is in use, does this, because they quickly realized how poor of a decision it was to model it at the individual level in the past. (notably range factor was created)

And one guy catches 75 more balls than the other guy, and most of those are in the gap. What would you say then?

Today we have pbp data and it would capture that, and I generally trust it's rankings, although I'm not 100% on board with it's math. (I understand why they do it the way they do it, just not sure it feels accurate to effectively count every positive play as if it is worth two) if you are talking about in the past, we don't have that data, so we can only extrapolate with the data we have into the past to build a model. These models have to hedge their bets because we know they aren't 100% accurate, so there is of course going to be regression in the models.

I would have no problem with you arguing for the regression is too strong on Ashburn or whoever your other pet outfielders happen to be, but arguing it because of range factor is ridiculous. As it stands, by the advanced system that bb-ref uses for war, Ashburns defense over Snider is responsible for 93 runs difference between the two players. Until you accept that you can't argue based upon individual stats withou examining the larger data, I'm not sure your argument has a point other than to prop up your guy.

of course it's not possible, because NO TEAM CAN FIELD 8 stellar defensive players! How stoopid is that? He makes up a completely non sensical hypothetical that has never happened in the entire history of baseball (OK maybe the 1941 Brooklyn Dodgers had 8 great players or the 1911 Athletics) and that's his reasoning that no OF can be +25 runs on def...


Why not? Theoretically speaking why couldn't a team field 8 stellar defensive players, and if they did, could they save that many runs? I don't know if I agree with it's not possible for an outfielder to save 25 runs or not, never really did the study or seen one saying that.
   88. Fancy Pants Handles lap changes with class Posted: April 19, 2014 at 10:24 AM (#4689051)
We absolutely do not know this, we have a fairly good idea, but considering that the two systems of war massively disagree on this particular point, I think it's fair to say that we do not have exact knowledge on the impact a pitcher has on run scoring.

That was poorly phrased, and I should not have used the word exactly. The point is, that there is an undeniable, obvious and measurable difference between pitchers, and their effect on opponent's ability to score runs.

I have absolutely never, ever said or implied that catchers were as important as pitchers. I have consistently and constantly stated, that we do not have an idea of how much of an impact that they have on the game

The fact that we don't know how big of an impact it has, is an extremely strong indicator that it is negligible.
   89. BDC Posted: April 19, 2014 at 11:00 AM (#4689055)
Silence, my hypothesis that Piazza wasn't the Jeter of catchers is related to Fancy Pants' point that you simply don't see much variation in RA by catcher. I honestly don't know how great a catcher-defense component of RA might be compared to that for SS and will leave it to brighter minds to figure. But that's how it might work. Any catcher who'd have a markedly negative effect stops catching - Joe Torre perhaps, though even there it's hard to quantify if Torre really was all that terrible. He was perhaps moved largely because he was a great hitter and could stay healthy and provide lots of oWAR elsewhere.
   90. cardsfanboy Posted: April 19, 2014 at 11:33 AM (#4689064)
The fact that we don't know how big of an impact it has, is an extremely strong indicator that it is negligible.


That was again, what was said about pitch framing. Something that is currently not measurable, doesn't mean it's negligible.

Again, War is going with the hypothesis(not proven by any means) that all position players are equal. So it's numbers are based upon that conclusion. I do not agree with that conclusion. I do not for the life of me think baseball has evolved to the point of equilibrium that each position on average is equally valuable to the team.

I think there is plenty of room for study on this issue, and don't think it's remotely close to being confirmed.

Silence, my hypothesis that Piazza wasn't the Jeter of catchers is related to Fancy Pants' point that you simply don't see much variation in RA by catcher. I honestly don't know how great a catcher-defense component of RA might be compared to that for SS and will leave it to brighter minds to figure. But that's how it might work. Any catcher who'd have a markedly negative effect stops catching - Joe Torre perhaps, though even there it's hard to quantify if Torre really was all that terrible. He was perhaps moved largely because he was a great hitter and could stay healthy and provide lots of oWAR elsewhere.


The problem of measuring catchers impact is because there is just no large data to go by. You can find examples to fit your hypothesis, but the data is too jumbled up with so many other variables that there is not much to show for it. If you look at a pitcher and try to figure out how much his catcher figured into his growth as a pitcher, there just isn't that much data.


I think that what we will see over time with the pitch FX data is new theories on catchers defense, and some of that will even help develop methods to make backwards looking defensive adjustments.
I fully expect to see some indication that catchers improve in pitch handling with age. I wouldn't be surprised at all also to see some pitchers show a tendency with certain style catchers to use more of their pitch repertoire and that usage of more pitches leads to weaker hit balls in play. Especially in crucial situations (in english... I think a catcher who has a good reputation of blocking balls in the dirt, means that with a man on base, the pitcher is more comfortable with throwing his big breaking ball, than with a catcher without that reputation---or a catcher with a great arm, might mean a pitcher is willing to throw fewer fastballs with a threat to steal on first)



   91. zenbitz Posted: April 19, 2014 at 02:10 PM (#4689125)
I think the above arguments about pitch framing taking WAR from pitchers and giving it to catchers is probably right. Of course, the magnitude is hard to estimate.

And the Football Center analogy is apt as well. You have to have a Center or no one can snap the ball. But Centers aren't very highly paid relative to OL. Being an RL MLB C (defensively) is clearly a very difficult task - the pool of acceptable or even marginal catchers is much smaller even than the pool of SSs. But that doesn't mean their relative value (WAR) is the same.

If you just want "top 10 catchers" (say for an all-star team) then WAR is fine. If you want to know whether Johnny Bench was better then Willie McCovey, it's a bit trickier. I guess instead of an average by year you could construct an average "Catcher" career curve (and 1B) and measure value above that - then correct for Replacement.
   92. CrosbyBird Posted: April 20, 2014 at 07:44 PM (#4689873)
Any catcher who'd have a markedly negative effect stops catching - Joe Torre perhaps, though even there it's hard to quantify if Torre really was all that terrible.

I think that's exactly it. The replacement level for catching in terms of plate-blocking/pitch-framing is almost certainly pretty high.

Especially since backup catchers are strongly selected for "ability to catch" and fairly weakly selected for hitting ability.
   93. tshipman Posted: April 20, 2014 at 07:58 PM (#4689888)
I think that's exactly it. The replacement level for catching in terms of plate-blocking/pitch-framing is almost certainly pretty high.

Especially since backup catchers are strongly selected for "ability to catch" and fairly weakly selected for hitting ability.


Crosby, this is more evidence that catchers should not be equivalent in value for a single game to first basemen. If catching is so important that throughout all of history, managers have been willing to take a player who cannot hit as well, doesn't that indicate that catchers are more valuable on a per-game basis than other position players?
   94. CrosbyBird Posted: April 20, 2014 at 08:32 PM (#4689917)
Crosby, this is more evidence that catchers should not be equivalent in value for a single game to first basemen. If catching is so important that throughout all of history, managers have been willing to take a player who cannot hit as well, doesn't that indicate that catchers are more valuable on a per-game basis than other position players?

In an absolute sense, certainly. But relative to their replacements, nope. WAR doesn't translate to absolute value. It's a purely relative measure.

A catcher could be worth twice as many runs as a 1B relative to zero, but if he's half as far away from the replacement as his 1B counterpart, WAR will value the 1B more. And it wouldn't be a flaw or a problem with WAR, but the statistic measuring what it attempts to measure.

I would have a lower bar for things like HOF admission or MVP selection for great catchers (not that I'd use only WAR anyway). But for helping me decide how much to pay a player, or whether my team is going to improve more by upgrading catcher or OF, I think something like WAR is really, really valuable. Because I don't really care how good a player is in a vacuum anyway; I only care about how good he is relative to the other options available.
   95. BDC Posted: April 20, 2014 at 09:05 PM (#4689933)
WAR is an abstraction that may help compare players, or think about trades and contracts. It's not highly nuanced. For instance, if your CF or SS goes down, you can move an adjacent player over or tap a utility man (who might sometimes even play both SS and CF). But if your C goes down, unless it's a total haywire emergency, you must draw from your organization's pool of dedicated catchers. That makes C seem scarce and special, which indeed they are, which means that there's always a pool of them. It's a bit of a paradox.
   96. Karl from NY Posted: April 21, 2014 at 10:12 AM (#4690158)
If he wanted to use a football analogy, then comparing a catcher to the center would have been more accurate, if the point was to show how relatively little importance the catcher has, even though he is involved in every play.

That wasn't my point. I brought up the football analogy to illustrate that less playing time can and should cause less WAR. It's not a flaw in the WAR system if one position comes out with less WAR.

Folks here are conflating two things: that career WAR totals for top catchers undercut those for other positions, and that there may be value in catching that isn't captured in WAR. These are two completely totally absolutely separate ideas. The first does not imply the second at all. Catcher WAR could be less for many other reasons, chiefly durability and career length. Catching WAR could be going unrecognized by the formulas. But those have nothing to do with each other. It's not a treasure hunt for the missing catcher WAR, it may simply not exist.

Football kicker is the analogy I'm using here because it illustrates the point clearly: football kickers certainly have less WAR than other positions, but there's no reason to think there's value in kicking that isn't captured in football WAR.
   97. tshipman Posted: April 21, 2014 at 11:08 AM (#4690178)
In an absolute sense, certainly. But relative to their replacements, nope. WAR doesn't translate to absolute value. It's a purely relative measure.


That's not its claim. It claims to allow you to compare players across positions, across eras.

A catcher could be worth twice as many runs as a 1B relative to zero, but if he's half as far away from the replacement as his 1B counterpart, WAR will value the 1B more. And it wouldn't be a flaw or a problem with WAR, but the statistic measuring what it attempts to measure.

I would have a lower bar for things like HOF admission or MVP selection for great catchers (not that I'd use only WAR anyway). But for helping me decide how much to pay a player, or whether my team is going to improve more by upgrading catcher or OF, I think something like WAR is really, really valuable. Because I don't really care how good a player is in a vacuum anyway; I only care about how good he is relative to the other options available.


No, it is a problem. Let's say you have 20 million in your budget for next year, and you base all your decisions off WAR (you're Murry Chass's strawman GM, basically). If you have to choose between two players, let's call them Puster Bosey, a catcher, and Breddie Breedman, a first baseman. Puster Bosey has a three year weighted average of 5.5 WAR and Breddie Breedman has a three year weighted average of 5.8 WAR. They both want the same amount of money--20 million per for 6 years. Do you give Breddie your bread or do you put it out for Puster?

I would submit that baseball GMs would overwhelming choose the catcher in that situation. Would you do otherwise?

WAR is a somewhat kludgy thing designed to make most things look right. It breaks down in a few areas. No one actually thinks that the best seasons of all time were a bunch of 19th century pitchers. It treats each game as equivalent for all position players. I think that is a mistake.
   98. cardsfanboy Posted: April 21, 2014 at 12:53 PM (#4690285)
If you have to choose between two players, let's call them Puster Bosey, a catcher, and Breddie Breedman, a first baseman. Puster Bosey has a three year weighted average of 5.5 WAR and Breddie Breedman has a three year weighted average of 5.8 WAR. They both want the same amount of money--20 million per for 6 years. Do you give Breddie your bread or do you put it out for Puster?


That case is missing war per inning/game played. In that theoretical, more than likely the catcher guy plays less game, and it's always better to get the higher per value per game.


The rest of the comment I agree with, and I think even if you change your example from average war per year, to average war per 600 pa, that they would still take the catcher(for many of the reasons I stated on this thread)

Of course we also have to assume that war is more accurate to the decimal than it claims it is with these examples.
   99. Slivers of Maranville descends into chaos (SdeB) Posted: April 21, 2014 at 01:05 PM (#4690297)
No one actually thinks that the best seasons of all time were a bunch of 19th century pitchers.


Why not? Pitchers in the 19th century were like quarterbacks in football today -- they were the key player in something like 40-50% of a team's games.

Are you saying that if Verlander pitched 83 games last year and went 60-23 he wouldn't be more valuable than any other position player by a mile?
   100. Sunday silence Posted: April 21, 2014 at 01:14 PM (#4690311)
Why not? Theoretically speaking why couldn't a team field 8 stellar defensive players, and if they did, could they save that many runs? I don't know if I agree with it's not possible for an outfielder to save 25 runs or not, never really did the study or seen one saying that.


I dont think I really disagree with you, but I want to clarify this. Bill James was saying it was impossible (or nearly so) for some player e.g. Kaline or someone to have saved 25 runs in RF. And his argument was that no team was saving 400 runs on defense (8 x 25) or some argument akin to that...

And I am saying in REALITY (not theoretical as you are saying) that we dont have any data pts to show that. No team has been able to field a team of say outstanding position players at all positions. So the pt is entirely theoretical and it hardly rebuts the claim.

ANd that is the pt. I was making with your pt; your claim about teams getting 20 hits to the OF would be true, if you could find a season like that. But in REALITY that doesnt happen teams produce hitting and fielding and pitching stats that are very close to one another. And they are being paid commensurate sums to one another, so we have really good reason to think that we have fielders that are getting very similar numbers of balls it to their zone. Your argument is totally theoretical and doesnt exist in the real world of MLB.
Page 1 of 2 pages  1 2 > 

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
Adam M
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Hot Topics

NewsblogBlue Jays Acquire Mayberry Jr.
(4 - 7:45am, Sep 01)
Last: Edmundo got dem ol' Kozma blues again mama

NewsblogBob Melvin calls Athletics 'pathetic' after Angels sweep four-game set
(3 - 6:24am, Sep 01)
Last: RMc's desperate, often sordid world

NewsblogOT: The Soccer Thread August, 2014
(944 - 6:08am, Sep 01)
Last: Swedish Chef

NewsblogSherman: How Reds react to second-half swoon will be major factor in offseason
(6 - 6:06am, Sep 01)
Last: Davo Dozier

NewsblogOT: Politics, August 2014: DNC criticizes Christie’s economic record with baseball video
(6297 - 5:07am, Sep 01)
Last: BrianBrianson

NewsblogTigers' Miguel Cabrera appears to re-injure ankle, leaves game
(10 - 3:48am, Sep 01)
Last: Cooper Nielson

NewsblogBackman named PCL’s top manager
(12 - 2:57am, Sep 01)
Last: ReggieThomasLives

NewsblogOT: NBC.news: Valve isn’t making one gaming console, but multiple ‘Steam machines’
(779 - 2:09am, Sep 01)
Last: DJS and the Infinite Sadness

NewsblogJesus Montero gets heckled by Mariners cross checker during rehab stint
(58 - 2:00am, Sep 01)
Last: Weratych

NewsblogAthletics Acquire Adam Dunn
(22 - 1:37am, Sep 01)
Last: 6 - 4 - 3

NewsblogOMNICHATTER 9-1-2014
(1 - 1:13am, Sep 01)
Last: Best Regards, President of Comfort, Esq.

NewsblogOT August 2014:  Wrassle Mania I
(90 - 12:53am, Sep 01)
Last: andrewberg

NewsblogOMNICHATTER 8-31-2014
(100 - 12:29am, Sep 01)
Last: Jose Can Still Seabiscuit

NewsblogHigh School Baseball Game In Japan Takes 50 Innings, Four Days
(6 - 11:46pm, Aug 31)
Last: Gamingboy

NewsblogPhoto of the day: Bill Murray, indy league ticket-taker
(42 - 11:33pm, Aug 31)
Last: Robert in Manhattan Beach

Page rendered in 1.0291 seconds
52 querie(s) executed