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

Thursday, July 19, 2012

ESPN: What we talk about when we talk about WAR

WAR is not the Holy Grail of statistics.

It’s a conversation starter, not a conversation ender. It’s important, it’s useful and most of all it’s fun. It’s the best tool for comparing players across positions or across eras. It’s a great tool for evaluating a player’s contribution during a season. Maybe it doesn’t settle debates; but it helps us get closer to answers.

Anyway, this piece isn’t a straight primer on WAR. You can get that from Baseball Reference, which is providing us with WAR. (FanGraphs has a slightly different of WAR, which places a greater emphasis on peripheral stats.)

WAR isn’t perfect. It shouldn’t be taken as the dogmatic, absolute answer to questions. (Player A’s WAR is 2.3 and Player B’s is 2.1, ergo Player A is better!) It can certainly be too easy of a crutch in player evaluation at times, but use it wisely.

So yeah, WAR is on ESPN now.

Tripon Posted: July 19, 2012 at 03:56 PM | 114 comment(s) Login to Bookmark
  Tags: sabermetrics, war on drugs

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. tjm1 Posted: July 19, 2012 at 05:36 PM (#4187404)
Really nice article. Schoenfield does a nice job of explaining advanced stats for the average fan - he always has. Art Martone was probably the best I've seen at this in the traditional media outlets but he hasn't been writing for years now.
   2. McCoy Posted: July 19, 2012 at 06:08 PM (#4187446)
It’s a conversation starter, not a conversation ender.

I'll never understand this argument for WAR. What's the point of WAR if it simply is a conversation starter? Average is a conversation starter, home runs are a conversation starter. WAR is supposed to be the stat that takes all of that and measures it all to come up with one number. To me this line of reasoning is like measuring a piece of wood with a tape measure and then saying that that isn't the final measurement but merely the first step in the measurement process.
   3. Robert in Manhattan Beach Posted: July 19, 2012 at 06:15 PM (#4187457)
It starts conversations. Most of them go like this: "Andrleton Simmons only has 125 PAs and has more WAR than Carlos Beltran. Man, if he would have been up all year he would be having a Darwin Barney like elite season."

   4. PreservedFish Posted: July 19, 2012 at 06:18 PM (#4187465)
#2 - I think it's to head off complaints about WAR being imperfect. I used to be pretty active in a non-stathead baseball forum, and there were many people that will throw a useful statistic in the garbage if they discover a few oddities with it. Perfect is the enemy of good, you know.

Also, it's not analagous to the tape measure because of all the adjustments (era, park, position, whatever) that go into it. Homeruns is analagous to the tape measure and it is a conversation ender. How many homeruns did Jim Thome hit last month? There's a definitive answer.
   5. Srul Itza Posted: July 19, 2012 at 06:22 PM (#4187466)
To me this line of reasoning is like measuring a piece of wood with a tape measure and then saying that that isn't the final measurement but merely the first step in the measurement process.


This is a red herring. You measure a piece of wood, you are taking a simple, exact measurement.

WAR is, of necessity, based on a number of assumptions. The fielding input alone -- especially for the era before the current PBP metrics -- is subject to a lot of assumptions and questions. One-Size-Fits-All ballpark adjustments are another area of dispute. Assumptions about where to place replacement value have an effect. The effect of different distributions of talent is also relevant.

When I look at WAR, I use large error bars to consider it. So does anyone else who is serious about it.
   6. McCoy Posted: July 19, 2012 at 06:24 PM (#4187472)
When I look at WAR, I use large error bars to consider it. So does anyone else who is serious about it.

So then what is the point of WAR?
   7. PreservedFish Posted: July 19, 2012 at 06:25 PM (#4187475)
It's a conversation starter! You silly goose.
   8. Srul Itza Posted: July 19, 2012 at 07:11 PM (#4187508)
So then what is the point of WAR?


It is an estimate of overall value put together by people who have been carefully studying this issue for a long time. It has as much point as the degree to which you accept that the work they have done is reasonable.

For that reason, I completely disregard Fangraphs WAR for pitchers, which is based solely on FIP or xFIP or the like.
   9. Everybody Loves Tyrus Raymond Posted: July 19, 2012 at 07:26 PM (#4187519)
To think an entire career can be summed up and neatly ranked in one number is folly.
   10. Everybody Loves Tyrus Raymond Posted: July 19, 2012 at 07:38 PM (#4187528)
And maybe I'm hard-headed or a Luddite or whatever, but I do NOT believe Brendan Ryan is having a better year than Prince Fielder. I just don't.
   11. willcarrolldoesnotsuk Posted: July 19, 2012 at 07:42 PM (#4187530)
Even if "the point" of WAR were "a single perfect number by which players can be exactly ranked", the fact that something has a "point" does not imply that it lives up to the point. So questions like "So then what is the point of WAR", in the context of "WAR is not perfect", don't really seem terribly sensible to me.
   12. McCoy Posted: July 19, 2012 at 07:54 PM (#4187538)
So questions like "So then what is the point of WAR", in the context of "WAR is not perfect", don't really seem terribly sensible to me.

Why do we need WAR if we have 10 million other imperfect and yet easier to calculate stats? Why do we need some single final number if we are going to take that number and then break it down?
   13. haggard Posted: July 19, 2012 at 08:12 PM (#4187561)
It's not really accurate to say, as the article does, that WAR attempts to calculate a player's total contribution to his team. It attempts to estimate what the player's contribution would be for a fictional team in a fictional setting. That's not exactly the same thing.
   14. willcarrolldoesnotsuk Posted: July 19, 2012 at 08:15 PM (#4187563)
Frankly, I'm not convinced that we do need it. The defensive part of it smells like bullshit to me (and worse - bullshit with significant weight). But again, regardless, questions like "why do we need it if it's not perfect" don't seem particularly sensible to me.
   15. Sunday silence Posted: July 19, 2012 at 08:20 PM (#4187567)
I'll never understand this argument for WAR. What's the point of WAR if it simply is a conversation starter? Average is a conversation starter, home runs are a conversation starter. WAR is supposed to be the stat that takes all of that and measures it all to come up with one number. To me this line of reasoning is like measuring a piece of wood with a tape measure and then saying that that isn't the final measurement but merely the first step in the measurement process.


In theory, you are right, that the goal of WAR is to measure wins and so it should reduce all the abilities of a player into one variable.

I will take a different tack than the other posters, and putting aside stuff like inexact measurements, park effects, etc; I would argue that there are enuf independent variables to make a single variable misleading/incorrect.

Take offense, scoring runs seems to be at least two if not three seperate variables: getting on base and moving runners along. Stealing bases might be somewhat related to the latter. While you can put express a double in terms of runs, and you can express a walk in terms of runs; I hesistate to regard them as both being the same variable.

The same reasoning can be applied to defense, although traditionally we look at pitchers in terms of ERA, in theory we should be able to break them down the same way.

So to make an extreme example, you might find 9 guys that all have exceptional fielding range but lack the ability to get on base, and they might lose 9 games out of ten to a similar group, with similar war but had a better distribution of OBP, power, fielding etc.

Essentially I am saying that creating a single variable out of a number of independent variables is going to lead to inexactitudes.

Another entirely different pt about WAR might be this: Most runs, other than HRs, are created by each player contributing some portion to the run, a walk here, a fielder's choice there, a fly ball there = one run. The ability to bunch these sort of outcomes together should lead to more runs as opposed to if they were scattered randomly throughout the game. It is possible that WAR does not take that into account, perhaps a player is really good at moving up runners on a fielders choice and actually helps the team; another player might have the same number of ground ball outs but werent produced at the right time. Pehaps it is a real skill, perhaps this is an illusion but in any event I dont think WAR is measuring it.

There's also another issue about leverage in the game. SB in my opinion have been historically devalued from what I've read. I forget what Pete Palmer had them at, about 0.2 IIRC. But I am not sure modern SABR methods are valuing them proper for situations when the SB impacts the chance of winning considerably more. Bunting might be another one for examination in this regard.

So you've got independent variables, you've got the problem of randomness vs. planned stuff, and you've got leverage to think about as well. Not sure one variable can handle all this.
   16. McCoy Posted: July 19, 2012 at 08:47 PM (#4187586)
Originally Palmer had SB low but by the time he wrote The Hidden Game he had bumped them up to .33 runs because people had convinced him that SB happen in high leverage moments.
   17. Ron J Posted: July 19, 2012 at 09:06 PM (#4187597)
#12 I can rarely figure out if you're being intentionally obtuse or not but:

WAR is a good place to start but it's standard error for a single season is not less than a win. And over a career (given we're generally using career WAR to talk about HOFers) it's probably at least 4 wins.

Thing is that's about as good as we can do. The standard error on for the offensive side is pretty close to 5 runs for a full-time player and there's no metric (never mind easy to calculate metric) that's better. There are in fact only a few in the same range.

Take OPS+ for instance. Standard error is in the 8-9 run range (again talking full-time players) and with OPS+ alone you can't easily compare players who play different positions. And you still haven't dealt with the value of their defense.

Yes, there are issues with the defensive component (mostly for infielders) currently, but the beauty of Sean's design is that it's modular. You can pop out DRS and pop in (say) PMR. Or any evaluation that you're prepared to defend.

WAR is a very good evaluation of the offensive contribution of a player and contains a well thought out adjustment for playing time and position. For historical players the evaluation of their defense is as good as you're going to get (and generally passes the smell test). For current players, oddly we have more data but currently less reliable results (and using the methods Sean worked out for historical probably won't work any better) -- because teams are changing the way they're playing hitters and are redistributing chances in a way that we don't currently have a good handle on.

But the problem with estimating the value of a player's contribution on defense doesn't go away if you adopt other methods.

Now you could start with (say) EQA, adjust for playing time and position (Clay conveniently already does this) and use another good defensive metric and the estimate of value would be within spitting distance of WAR. I encourage people to do so. And yes, you'll find a high correlation between the results. The guys who come out substantially different are worth looking at more closely to see why.
   18. Ron J Posted: July 19, 2012 at 09:19 PM (#4187608)
#16 And chopped back the weights in later edition because while it's true that SBs do occur more frequently in high weight situations that didn't seem to translate into wins and losses.

Very good base stealing teams don't tend to beat their pythags and vice-versa. The weights Palmer used in the early edition made his estimates less accurate (or would have if Palmer hadn't included the "slope correctors" a year by year and league by league adjustment)

And #15 that's part of the reason you get a standard error of 5 runs. Tom Ruane wrote a great article on evaluation of the running game using the base/out situation and found that a huge part of Rickey Henderson's edge in base stealing over Tim Raines disappears when you consider this.

Anything you think isn't properly handled by the metric is fodder for discussion (in exactly the same way that we all know that OPS+ undervalues OBP)
   19. cardsfanboy Posted: July 19, 2012 at 09:35 PM (#4187614)
I liked the article. The last paragraph about Jack Morris hits on one of my complaints about war.
One, it's my opinion that WAR shortchanges pitchers with durability. I'm not necessarily saying that applies to Jack Morris. But take, for example, Roy Halladay. When figuring his WAR, you're calculating his value over a replacement-level pitcher, maybe a pitcher with a 5.00 ERA or whatever that level is these days. But a Triple-A pitcher isn't going to give you seven or eight innings every start like Halladay does, and certainly not 233 innings like Halladay gave the Phillies in 2011. But WAR, the way I understand it, sort of assumes that's the case.


I agree, it short changes durability in all cases, but with pitchers moreso. I like war for position players(except first base and catcher) dislike it for starting pitchers, plus first baseman, catchers, utility and dh , and hate it for relievers.

So then what is the point of WAR?


It's a very good tool for comparing dissimilar numbers from players across era/position. But just like ops+, the components are probably more important for the conversation than the sum. I've never been a fan of comparing raw war without at least listing plate appearances in the number. I honestly think looking at seasonal war with less than 400 plate appearances is pure folly.


Yes, there are issues with the defensive component (mostly for infielders) currently, but the beauty of Sean's design is that it's modular. You can pop out DRS and pop in (say) PMR. Or any evaluation that you're prepared to defend.


It would be nice if it was modular enough that we could just put in our own defensive components(without doing all the lifting). I prefer to think of defense as excellent, good, average, below average, poor and should be moved off the spot. And would love it if I could just pop in that estimate as a replacement with a drag and drop interface :)



   20. Mefisto Posted: July 19, 2012 at 09:39 PM (#4187617)
To me, the issue with WAR is fairly simple. Granted that it has potential or actual flaws. What's important to me, whether I'm a GM or some guy having an internet debate, is whether using WAR is more accurate than not using it. If I don't try to make comprehensive evaluations of a player, including his defensive contribution, I'm certain to make serious mistakes. I'm convinced the mistakes with WAR are going to be less than the mistakes in failing to use it.

That doesn't mean anyone should stop trying to improve it. And it doesn't mean we should all start to treat WAR as ending a conversation. What it should do is put the burden of coming forward with evidence on those who argue that it's wrong in a particular case.
   21. SoSH U at work Posted: July 19, 2012 at 09:43 PM (#4187620)
One, it's my opinion that WAR shortchanges pitchers with durability. I'm not necessarily saying that applies to Jack Morris. But take, for example, Roy Halladay. When figuring his WAR, you're calculating his value over a replacement-level pitcher, maybe a pitcher with a 5.00 ERA or whatever that level is these days. But a Triple-A pitcher isn't going to give you seven or eight innings every start like Halladay does, and certainly not 233 innings like Halladay gave the Phillies in 2011. But WAR, the way I understand it, sort of assumes that's the case.


If that is indeed the case, wouldn't Morris be a perfect example of who it shortchanges?

   22. PreservedFish Posted: July 19, 2012 at 09:46 PM (#4187624)
I'm convinced the mistakes with WAR are going to be less than the mistakes in failing to use it.


Funny, I think it's just about useless for a general manager. The snark in #3 points out the weirdness that WAR gives you in small samples, and GMs are basically only evaluating what a guy has done for the last 2-3 years. WAR is best used exactly the way that it is used: as a starting point for little debates on websites like this one, mostly about career-length comparisons or other large sample topics.

If I'm a GM I would much much rather just have the raw offensive stats and apply my own mental defensive adjustment.
   23. cardsfanboy Posted: July 19, 2012 at 10:18 PM (#4187654)
If that is indeed the case, wouldn't Morris be a perfect example of who it shortchanges?


Agreed. I understand that 2.0 war is considered average and a guy who is pitching 95-105 era+ over 200-240 innings is going to generate in the 2.0-3.0 war range. But from a team point of view, he's much more valuable than just an average player. As I said, I do not like war for pitchers so I would rarely use it or to be honest I would never use it except as a sort/filter tool in Pi.

On the other side of the argument though, is that war has never really been about showing a players value to a franchise(even if Fangraphs tries to do that) it's about showing the players performance relative to a baseline. In that point of view, a league average pitcher over a standard season should be worth around 2.0-3.0 war, and to be honest, that is what you get with war. (except the insanity of fangraphs war for pitchers)

   24. Tom Nawrocki Posted: July 19, 2012 at 10:21 PM (#4187659)

Funny, I think it's just about useless for a general manager.


No one has ever explained to me why it would be useful for any GM to have a single number to attach to the value of a player's season. It's useful for Hall of Fame debates, and MVP debates, and things like that, but in the practical world, I can't imagine why anyone would use it.
   25. tshipman Posted: July 19, 2012 at 10:32 PM (#4187664)
Man, that is the worst Raymond Carver parody I've ever read.
   26. PreservedFish Posted: July 19, 2012 at 10:51 PM (#4187680)
No one has ever explained to me why it would be useful for any GM to have a single number to attach to the value of a player's season. It's useful for Hall of Fame debates, and MVP debates, and things like that, but in the practical world, I can't imagine why anyone would use it.


If it was deadly accurate, then it would be a great tool. But it isn't.
   27. cercopithecus aethiops Posted: July 19, 2012 at 10:57 PM (#4187686)
When figuring his WAR, you're calculating his value over a replacement-level pitcher, maybe a pitcher with a 5.00 ERA or whatever that level is these days. But a Triple-A pitcher isn't going to give you seven or eight innings every start like Halladay does, and certainly not 233 innings like Halladay gave the Phillies in 2011. But WAR, the way I understand it, sort of assumes that's the case.


But of course WAR makes no assumption about how many innings a replacement level pitcher will give you in a start or a season. Rather, it only assumes that in the absence of Mr. Halladay, the Phillies would have had to find 233 innings somewhere, and those innings likely would have been at a certain average performance level had they been thrown by some collection of freely available pitchers.
   28. tshipman Posted: July 19, 2012 at 11:03 PM (#4187693)
No one has ever explained to me why it would be useful for any GM to have a single number to attach to the value of a player's season. It's useful for Hall of Fame debates, and MVP debates, and things like that, but in the practical world, I can't imagine why anyone would use it.


A single number allows for a different concept of value. If WAR were 100% accurate, you could very easily evaluate existing MLB players according to a consistent rubric: we value x player at x dollars. This allows you to target players that potentially other GMs are not targeting. For instance, 2nd basemen might be consistently underpaid in your analysis, allowing you to target them as an upgrade to your team at a lower salary than other positions available.
   29. Ron J Posted: July 19, 2012 at 11:09 PM (#4187696)
#27 Some of whom will actually pitch well (Aaron Small 2005 for instance), some of whom will put up a negative WAR. On balance, well there's been a reasonable amount of research on the matter (in particular by Keith Woolner) and the place WAR sets replacement level makes theoretical sense as long as baseball talent is normally distributed (and we see only the extreme right of the curve in professional baseball)

There is some sense in looking at what I call VOTW (Value Over Train Wreck). Guys like Morris keep you from finding out what a replacement level pitcher having a bad season looks like.
   30. Matt Welch Posted: July 19, 2012 at 11:34 PM (#4187706)
I understand that 2.0 war is considered average and a guy who is pitching 95-105 era+ over 200-240 innings is going to generate in the 2.0-3.0 war range.

There have been 358 such seasons (95-105/200-240) since 1961. WAR breakdown like this:
5+ -- 1 (John Smoltz, with 5.1 in '91)
4-5 -- 2 (Elmer Dessens' 4.1 in '01, and Francisco Barrios' 4.0 in '77)
3-4 -- 40
2-3 -- 148
1-2 -- 130
0-1 -- 34
Less than zero -- 3 (Tom Phoebus' -0.1 in '69, Brian Kingman's -0.3 in '80, and Mike LaCoss' -0.4 in '86)
   31. Textbook Editor Posted: July 20, 2012 at 12:02 AM (#4187717)

Man, that is the worst Raymond Carver parody I've ever read.


I think my favorite is the New Yorker cartoon with a man and woman walking next to each other down a rainy street, looking miserable. Above them is the heading WHAT WE TALK ABOUT WHEN WE TALK ABOUT THE WEATHER, and the man says "I hate you," and the woman says "I hate you more."
   32. cardsfanboy Posted: July 20, 2012 at 12:13 AM (#4187721)
I think my favorite is the New Yorker cartoon with a man and woman walking next to each other down a rainy street, looking miserable. Above them is the heading WHAT WE TALK ABOUT WHEN WE TALK ABOUT THE WEATHER, and the man says "I hate you," and the woman says "I hate you more."


The New Yorker stole a strip from a Ziggy? :)
   33. valuearbitrageur Posted: July 20, 2012 at 12:39 AM (#4187729)
No one has ever explained to me why it would be useful for any GM to have a single number to attach to the value of a player's season. It's useful for Hall of Fame debates, and MVP debates, and things like that, but in the practical world, I can't imagine why anyone would use it.


Every GM thinks in terms of WAR, every single one. They mentally frame every player acquisition in how many wins they expect to gain, either this season or in future years, or how many they can buy with the trade chips or salary they free up. They are always cognizant of the replacement player value, what they have in inventory in their minor league system or can acquire very cheaply from others.

But that doesn't mean they measure WAR the same way as BBRef or Fangraphs. They might think Roy Halliday is worth 10 wins shoring up their rotation, or Prince Fielders home runs will add a dozen wins. Most GMs use whatever internal framework they have mentally created to estimate hw many wins a player is worth and that guides their decisions.
   34. Walt Davis Posted: July 20, 2012 at 12:59 AM (#4187735)
Sometimes WAR is a conversation ender: Paul Konerko is an HoFer; No way, he's only got 25-26 WAR and even bad HoFers get to around 50. If WAR can't measure a career to that level of accuracy then it's worthless.

What is the purpose of WAR? Darwin Barney sucks, he can't hit. Actually I think he adds enough in defense and baserunning that he's pretty much an average 2B. You're ####### insane. No, really. Look, there's no way he can possibly make up for being such a crappy hitter even if he's Maz with the glove. You don't have to Maz with the glove. You're ####### insane -- show me how he could possibly be average. OK, here's WAR which combines hitting, defense, baserunning and durability and even if you don't quite believe those high defensive numbers, it shows how that can all add up to average.

Maybe you weren't around for the heady early days of what Don Malcolm used to derisively call neo-sabermetrics but lots of us (including me at least some of the time) were totally focused on offensive statistics. The statement "I'm sorry, there's no way his defense can make up for his bat" were exceptionally common -- and it's the principle that McCoy seems intent on clinging to. Those conversations go nowhere because all anybody could say in response was "he looks really good to me and he's won some gold gloves." As someone above says, WAR is an attempt by very serious people doing the most careful work they can to develop the best estimates of these varying components we can have.

I don't know why any baseball fan wouldn't find it intersting to have an estimate of a player's contribution on offense, defense, the bases, etc.

I agree it's pretty much useless for GMs. First, you need to establish how well it predicts the future (relative to other info the GM has). But also I'm not sure it's clear that offense, defense and baserunning follow the same development/aging curves. Even assuming we know the true talent at a given point, is the guy who's 1 win below average with the bat and 1 win above with the glove going to age the same/better/worse than the guy who's average with the bat and the glove and the guy who's +1 with the bat and -1 with the glove. At a minimum, a GM is going to have to keep it at the component level unless somebody can show that, when it comes to prediction, a WAR is a WAR is a WAR.

To the extent it's useful to GMs it is in highlighting players worth further investigation but that's just a convesa ... err ... the first step in the process.
   35. The Anthony Kennedy of BBTF (Scott) Posted: July 20, 2012 at 01:28 AM (#4187750)
Wins is a currency we can all understand.


This is my primary complaint with the article. Given the recent rejiggering of DRS, I'm pretty pleased by this development.
   36. McCoy Posted: July 20, 2012 at 01:29 AM (#4187751)
It's quite possible to be an "average" second baseman and still "suck".

I'm not clinging to whatever principle you are describing.
   37. cardsfanboy Posted: July 20, 2012 at 01:42 AM (#4187754)
First off, I loved your post in 34 Walt, it threw me for a loop for a second due to style, but once I grasped what you were doing, fantastic post.

It's quite possible to be an "average" second baseman and still "suck".

I'm not clinging to whatever principle you are describing.


This is one of my beef's with War, it's built on the framework that all the positions are equal. In theory, if you had a perfectly average player across the board(offense/defense) and war was 100% accurate, and that player played every inning of every game and got exactly the correct number of plate appearances as an average player, he would be worth 2 war, no matter what position he plays.

I find that to go counter to what I see front offices do or what I see on the field. Teams routinely punt positions, and as mentioned, second base seems to be the favorite position to punt on. Teams don't care to invest or develop second baseman, it's the position where they put a guy who they hope isn't going to hurt them. Of course this philosophy makes the actual good second baseman much more valuable relative to their performance. But that doesn't really mean they are that good, it means that teams are putting replacement level players at second base frequently and it's happening enough that the average second basemans performance as recorded by war, is probably closer to actual replacement level performance.



   38. Everybody Loves Tyrus Raymond Posted: July 20, 2012 at 01:46 AM (#4187758)
You're ####### insane -- show me how he could possibly be average. OK, here's WAR which combines hitting, defense, baserunning and durability and even if you don't quite believe those high defensive numbers, it shows how that can all add up to average.


If you show someone skeptical about Barney even being average a metric that says he's one of the top 6 players in the league isn't the skeptical guy really going to tune you out at that point? WAR works in this capacity if it has him somewhere around average. Instead, it says he's fantastic. Even if he doesn't suck (and, fwiw, I don't think he sucks), he's closer to sucking than being what WAR says he is.
   39. cardsfanboy Posted: July 20, 2012 at 01:53 AM (#4187759)
My post made me think of something, which I'm sure someone has pointed out before but I must have missed. One drawback with all of these numbers is that they include replacement level performances in their calculations. Doesn't that hurt their accuracy to some degree? Wouldn't that mean that the average that they are creating is actually below true average?
   40. Walt Davis Posted: July 20, 2012 at 04:12 AM (#4187775)
This is one of my beef's with War, it's built on the framework that all the positions are equal.

Kinda, in the sense that an "average" 2B should be able to move to 1B and you'd gain enough in defense to offset the offensive hit. But really that's just an assumption that the observed, empirical difference in offense between positions equals the defensive difference. That's hard to prove but does have at least some backing from studies Tango's done and they actually measure it a bit more precisely than that. But if the extra offense were worth the hit, you'd see more Daniel Murphys at 2B.

But, yes, I don't think you can necessarily swap positions willy-nilly as WAR implies ... and I doubt the WAR creators really think that either. In that sense, it gets back more to the "value" sort of question -- whether a 2B would produce the same value at 1B is kinda moot until he actually moves to 1B; while at 2B, here's how much value he's producing.

I find that to go counter to what I see front offices do or what I see on the field. Teams routinely punt positions, and as mentioned, second base seems to be the favorite position to punt on.

I don't agree with that. There are ebbs and flows but the average 2B historically hits about the same as the average CF and average 3B which is right around league average. That doesn't sound like punting to me. And that makes sense -- those are the positions for guys who can't handle SS defensively but hit better than the typical SS (LH CF excepted in a way). It's selection bias in that we rarely see the 2B who can't hit better than a SS (at least not for very long).

We like to dump on conventional wisdom but baseball history is perfectly consistent with the notion that there are plenty of guys out there who make up for their bat with their glove. I think there's some evidence that SS (and doubly so C) are positions that teams are NOT willing to punt defense for offense but I don't see any evidence that they feel that way about 2B or 3B which suggests they're going with the balance of the two. Of course any given team may be constrained as to who they have available.

The inconsistent part is that average 1B/LF/RF get paid a lot better than average 2B. This may be the flipside -- there's evidence teams are willing to punt defense for offense at those positions and being a really good hitter seems to be rarer than being a good defender. In that sense it is probably easier to find a guy who can defend 2B and put up a 85-90 OPS+ (Barney) than a guy who can hit for a 115 OPS+.

Teams don't care to invest or develop second baseman,

I'm not so sure that's really true. I'm not sure this is anything more than "everybody starts as a SS." If your defensive ability prior to being drafted is already at the point where you're playing 2B then you need to rake because you are probably ending up in a corner. It probably is fair to say that 2B are failed SS (in the same sense that relievers used to be failed starters) except that they're not totally failed SS in that the ones that make the majors hit better than SS. But, sure, in essence Darwin Barney is an offensively average, defensively below-average SS playing 2B.

One drawback with all of these numbers is that they include replacement level performances in their calculations. Doesn't that hurt their accuracy to some degree? Wouldn't that mean that the average that they are creating is actually below true average?

I'm not sure what you're asking. What I think you're trying to get at is that there's a difference between "league average" and "average starter" at each position because league average is a mix of starter, bench and emergency replacement performance. That's true and saying that Barney is an average 2B is not the same as saying he's an average starting 2B. We, including I, should be more careful about that.

But I think people forget how few players are "full-time" players -- whether due to injury or platooning or suckitude or whatever. In 2011, only 145 players reached the 502 PA mark (which would include only a handful of Cs); only 68 made it to 600 and only 41 made it to 650. Bench/replacement guys get a lot of playing time and that's what Rrep is all about -- the more you keep the replacement guy off the field, the more value you have even if you're just Darwin Barney.

And of course b-r now conveniently gives you WAA which means you don't have to pay any attention to replacement level if you don't want to although you do still have to believe that the measure of average defense at a position is reasonably accurate. (And you have to decide how you're going to reward playing time.)

If you show someone skeptical about Barney even being average a metric that says he's one of the top 6 players in the league isn't the skeptical guy really going to tune you out at that point?

Different question really. I'm talking the "concept" of WAR, this is a question of the "quality." Are Barney's defensive numbers believable is not the same question as to whether a measure which combines offensive, defensive and baserunning contributions is a useful way to value a player's total contribution (i.e. what is WAR for).

But regarding your question -- maybe he will still tune you out ... even as you point out that if you took away every single run above average defense, Barney still grades out as 3 runs above-average on the season so far. Zero out his baserunning and positional adjustment and he's just 3 runs below average -- that's now just 3 runs below the average hitter with no positional adjustment. (Fangraphs puts him at 5 runs below offensively and 4 runs above overall.)
   41. tjm1 Posted: July 20, 2012 at 04:57 AM (#4187776)
I'm not so sure that's really true. I'm not sure this is anything more than "everybody starts as a SS." If your defensive ability prior to being drafted is already at the point where you're playing 2B then you need to rake because you are probably ending up in a corner. It probably is fair to say that 2B are failed SS (in the same sense that relievers used to be failed starters) except that they're not totally failed SS in that the ones that make the majors hit better than SS. But, sure, in essence Darwin Barney is an offensively average, defensively below-average SS playing 2B.


I think this is mostly true, but is the defensive positional adjustment between SS and 2B really as large as the amount by which Barney is rated to be above average at 2B? If not, then at least one of three things is true: (1) the numbers are off; (2) he's an above average defensive SS playing 2B because his team has another SS (3) his defensive skill set is uniquely suited to 2B. There will be guys like that - great hands and quick feet, especially on the DP pivot, but a poor throwing arm. Mazeroski was one of those guys.

As for your comment about the salaries of corner outfielders and first basemen - I think there are several issues again. One is that the most extreme outliers in value are going to be the guys who are great hitters, regardless of position. Is it the average or the median which is higher at the corner positions? Another is that defense tends to fall off at a younger age than hitting. The best fielders are usually guys under club contractual control. Also, players then move to easier positions if they can hit, but can't field a tough position any more, meaning that the 1B/LF/RF types tend to be older, and more likely into higher salary post-free agency years.

There may be a market inefficiency here, but I don't think so, really. You need a decent hitting LF to compete, and some of the better ones are guys who are 30 now, and when they were 25 were good-hitting, good fielding CFs under contractual control at low salaries. Compare e.g. Johnny Damon as a Royal with Johnny Damon as a Yankee. As a Royal he was a bit better overall player, but at a much lower salary and at a tougher defensive position.

I think it's extremely hard to work through all this information and figure out if some positions on the field genuinely have better players than others. The one thing I will say is that it's extremely rare to see a team try to move a good hitter from any position other than SS to 2B and to have that be anything other than a disaster. Biggio worked (but from C, also a hard position). The Bill Madlock experiment was successful, but short-lived, and Madlock really wasn't a good fielder at 2B, but was acceptable enough that getting him, Evans and McCovey in the lineup at the same time made it worthwhile. There are probably one or two others, but I can't think of any off the top of my head.
   42. Jose Bautista Bobblehead Day Posted: July 20, 2012 at 05:39 AM (#4187781)
Ron J in #17 pretty much nails it. The non-defensive assumptions made to estimate value in WAR have some error, but in the aggregate, they yield a pretty good estimate of player value in a single number. I don't think very many value comparisons or rankings would look obviously wrong without the defensive runs component. If we want to quantify this, we can do as Ron suggested and generate our own alternative player value estimators, then measure correlations and check for outliers.

Also, another point Ron makes in #17 deserves further consideration:

"Yes, there are issues with the defensive component (mostly for infielders) currently, but the beauty of Sean's design is that it's modular. You can pop out DRS and pop in (say) PMR. Or any evaluation that you're prepared to defend."

The main obstacle toward people accepting the validity of BB-Ref WAR comes from the fact that the defensive metric it uses sometimes spits out wonky results and they get baked into the pie for overall player value*. Happens all the time here. It would be great if Sean could figure out a way to automate what Ron suggests in the quote above, and allow people to generate different WAR estimates based on a range of defensive metrics. For instance, when considering the range of value estimates from available defensive metrics, I feel pretty confident that Darwin Barney has accrued a value of 2-4 WAR this season. Such uncertainty is not ideal, but it is preferable to 1) abandoning any meaningful attempt to quantify and incorporate defensive value in player value estimators or 2) disregarding estimates of WAR because individual point estimates of defensive value seem wrong.

*Just because Barney or Brett Lawrie have very high DRS estimates, though, it does not mean that an 87+ OPS 2B who doesn't steal bases can never be the sixth-best player in the league.
   43. McCoy Posted: July 20, 2012 at 08:18 AM (#4187796)
Barney is now at a 4 WAR. His Rbat went from -4 to -2 and his Rfield rose from 23 to 24 in the last 3 games or so.
   44. tjm1 Posted: July 20, 2012 at 08:33 AM (#4187802)
I think the real question with Barney is how he improved his defense so much in one season, if the stats are actually reliable. Last year he rated out as a solid, but unspectacular 2B. This year, he rates as one of the greatest defensive 2B's of all time. About 90 games into the season, he's already got a DWAR right in the neighborhood of Mazeroski's best season (2.9 vs 3.3 for Maz). I haven't seen enough Cubs games to have a comment on this - is this guy really one of the all-time great 2Bs?
   45. BDC Posted: July 20, 2012 at 08:47 AM (#4187808)
is this guy really one of the all-time great 2Bs?

One possibility is that he isn't, but that he's actually achieving great things this year. For instance, neither Brady Anderson nor Davey Johnson was one of the all-time great home run hitters, and Darin Erstad was not one of the all-time great average hitters, but they had great seasons in those categories.

The difference between "is great" and "is doing a great thing at the moment" is sometimes blurred in these discussions. (The converse was the thread where TFA announced that Tim Lincecum sucks; well, he's having a sucky year, but I don't think "he sucks" is a reasonable account of it.) If we need a few years of data to be able to say something valid about a fielder, too, all the more reason why a solid guy might show as extremely great over just four months.

I don't know if this is truly the case with Barney, but it's a possibility.
   46. Joey B. is counting the days to Trea Turner Posted: July 20, 2012 at 08:57 AM (#4187814)
And maybe I'm hard-headed or a Luddite or whatever, but I do NOT believe Brendan Ryan is having a better year than Prince Fielder. I just don't.

Don't feel bad; nobody really believes that Ben Zobrist is actually the second best player in baseball either.
   47. Der-K: downgraded to lurker Posted: July 20, 2012 at 09:55 AM (#4187848)
but I do NOT believe Brendan Ryan is having a better year than Prince Fielder. I just don't.

I'll say this (and it's all been said better above):

The idea that there's a significant premium for the ability to play average defense at short over doing the same at first is credible.
The idea that Brendan Ryan's defense is worth quite a bit more than an average defender at short is also credible. (He may be the best glove at that spot in the world. Unlikely, but possible.)
The idea that Prince Fielder defense at first is worth quite a bit less than an average first baseman is too. (Presumably, we agree on this.)
This adds up. Quick. (You can also throw in differences from baserunning, double play avoidance, park effects, etc...)

I don't think Ryan (.190/.287/.287, 65 OPS+ in 290 PA) is having a better year than Fielder (.308/392/.504, 141 OPS+ in 403 PA) either - as I doubt Ryan is really having a +22 year on defense (other metrics have him having a strong year as well, but not that strong) ... but it is not outside the realm of possibility.

(I have a separate issue with how we define replacment level across positions, as do others - but that's not what I want to talk about.)
   48. McCoy Posted: July 20, 2012 at 09:59 AM (#4187851)
For instance, neither Brady Anderson nor Davey Johnson was one of the all-time great home run hitters, and Darin Erstad was not one of the all-time great average hitters, but they had great seasons in those categories.

But we can also understand how these guys did that. Brady was a juicer hitting juiced balls. Davey moved to the Launching Pad in a year when the NL saw a spike in HR. Darin in 2000 got to play in a high offensive environment and managed to hit lefties really well that year. Sure all of them had a good bit of flukish luck as well as environment help them out but the point is we can see the conditions that made it possible for them to do what they did. With fielding you can't really see that or really why would flukey conditions exist for fielding? There are no juiced gloves, they don't change the rules on what is an out or not, so on and so on. That isn't to say fielding should be consistent from year to year but it does make it hard to accept numbers that vary wildly from year to year.
   49. McCoy Posted: July 20, 2012 at 10:04 AM (#4187856)
Brendan Ryan in 290 PA is given 10 replacement runs and Darwin Barney with 332 PA is given 9 replacement runs. Apparently the AL is about 15% or so tougher than the NL this year.
   50. Rally Posted: July 20, 2012 at 10:09 AM (#4187858)
Brendan Ryan in 290 PA is given 10 replacement runs and Darwin Barney with 332 PA is given 9 replacement runs. Apparently the AL is about 15% or so tougher than the NL this year.


For the last decade or so, at the end of the year a full time player in the AL should have 22 replacement runs, and in the NL 18. With the AL going 142-110 in interleague once again, there's no need to revisit that.
   51. zack Posted: July 20, 2012 at 10:29 AM (#4187871)
It would be nice if it was modular enough that we could just put in our own defensive components(without doing all the lifting). I prefer to think of defense as excellent, good, average, below average, poor and should be moved off the spot. And would love it if I could just pop in that estimate as a replacement with a drag and drop interface :)


AROM used to do that on his site, "if you believe his defense is..." "then his WAR is...". I liked that.

#12 I think you're assigning Sean some credit that goes to a lot of other guys too, though he of course deserves tons of credit for making it so accessible.

I think it's likely that a site like bb-ref should not publish in-season WAR. The defensive component is too misleading. But I can't stand the thought of another conversation about fielding metric in-season variability so I'm going to stop typi
   52. tjm1 Posted: July 20, 2012 at 10:41 AM (#4187881)
There are no juiced gloves, they don't change the rules on what is an out or not, so on and so on. That isn't to say fielding should be consistent from year to year but it does make it hard to accept numbers that vary wildly from year to year.


Exactly. I suppose that Barney could be having a lot of balls hit to the exact edge of his range, instead of just beyond it, or something. I also suppose that it's possible that his intrinsic defensive ability has been beyond Mazeroski-level for years, and it's just that last year maybe he had a hamstring strain affecting his range that we didn't know about. I just doubt that he's the greatest defensive player ever at his position by a wide margin after being just a bit above average last year. Belanger is the only guy I can find who broke dWAR of 4.0 more than once, according to the BBREF system. There might be a few others - I just checked some obvious candidates.
   53. PreservedFish Posted: July 20, 2012 at 10:46 AM (#4187886)
But we can also understand how these guys did that. Brady was a juicer hitting juiced balls. Davey moved to the Launching Pad in a year when the NL saw a spike in HR. Darin in 2000 got to play in a high offensive environment and managed to hit lefties really well that year. Sure all of them had a good bit of flukish luck as well as environment help them out but the point is we can see the conditions that made it possible for them to do what they did.


Some of these excuses are pretty lame, and of course it wouldn't be difficult to come up with other fluke performers that didn't have these contextual factors.

But I probably agree with the overall point, and I certainly agree with this:

I think it's likely that a site like bb-ref should not publish in-season WAR. The defensive component is too misleading. But I can't stand the thought of another conversation about fielding metric in-season variability so I'm going to stop typi
   54. Rally Posted: July 20, 2012 at 10:58 AM (#4187893)
Belanger is the only guy I can find who broke dWAR of 4.0 more than once, according to the BBREF system.


It's really not comparable. Sean is now using John Dewan's defensive numbers, and the spread of performance in those are higher than what you'll find in the pre - 2002 numbers. The farther you go back in time, the smaller the spread in defensive ratings, because of the data available to make those ratings.

If the kind of video analysis that Dewan is doing were available going back to 1871, I'm sure we'd have a lot more outlier defensive seasons to compare to than we do right now.
   55. cardsfanboy Posted: July 20, 2012 at 12:02 PM (#4187978)
But, yes, I don't think you can necessarily swap positions willy-nilly as WAR implies ... and I doubt the WAR creators really think that either. In that sense, it gets back more to the "value" sort of question -- whether a 2B would produce the same value at 1B is kinda moot until he actually moves to 1B; while at 2B, here's how much value he's producing.


I wasn't saying that you could swap positions. I was saying I don't think that each position is worth the same. I don't think teams consider it. Second base is where you throw your replacement level players, because it's a relatively unimportant position. I think first base and catcher are inherently more valuable positions than the other positions on the diamond, especially second base, right field and left field. The average first baseman provides more value than the average second baseman in terms of wins added to the team.

I don't agree with that. There are ebbs and flows but the average 2B historically hits about the same as the average CF and average 3B which is right around league average. That doesn't sound like punting to me. And that makes sense -- those are the positions for guys who can't handle SS defensively but hit better than the typical SS (LH CF excepted in a way). It's selection bias in that we rarely see the 2B who can't hit better than a SS (at least not for very long).


Quite possibly, right now it's an ebb period at second base. Teams are punting the position or allowing guys to hang out there longer than they should because it's an unimportant(perceived) as a replacement position.

'm not sure what you're asking. What I think you're trying to get at is that there's a difference between "league average" and "average starter" at each position because league average is a mix of starter, bench and emergency replacement performance. That's true and saying that Barney is an average 2B is not the same as saying he's an average starting 2B. We, including I, should be more careful about that.


What I'm saying, is that these numbers include replacement level performance in their numbers. Let say that most of the positions in baseball have roughly 10% of the performance played by replacement level players. If it's consistent across the board, at all the positions then it wouldn't really matter overall, but let's say that at second base 30% of the performance is being taken up by replacement level players (argument is that the position is injured more and teams don't generally have a backup second baseman other than a generic utility fielder, or that teams are willing to put a Daniel Murphy, or Skip Schumaker out there) This would mean that since the numbers are compared across the board from the performance of everyone who played that position, that the league average is being brought down, making the players who are truly league average, appear to be better than league average. This is how you make a Darwin Barney look good, the position he plays at is weaker than it should be because teams are going to replacement level players (or in the case of the Brewers, sticking with one out of loyalty) more frequently.


For the last decade or so, at the end of the year a full time player in the AL should have 22 replacement runs, and in the NL 18. With the AL going 142-110 in interleague once again, there's no need to revisit that.


What is the purpose of this, how does it affect the numbers? Does this mean a .300/.400/.500 average defensive first baseman in the al would get more war than the same numbers in the nl?
   56. McCoy Posted: July 20, 2012 at 12:19 PM (#4188000)
Some of these excuses are pretty lame, and of course it wouldn't be difficult to come up with other fluke performers that didn't have these contextual factors.

Well, those were the examples I was given to work with and I doubt you'll find very many high end performances that don't have a lot contextual factors. Extremes happen because conditions are right to make them happen. Babe Ruth doesn't hit 60 homers in 1906, Hack Wilson doesn't get 181 RBI in 1968, so on and so on.
   57. Ron J2 Posted: July 20, 2012 at 12:51 PM (#4188049)
With fielding you can't really see that or really why would flukey conditions exist for fielding? There are no juiced gloves


You're overlooking a pretty big factor here. It's very common for a player to (attempt to) play through injuries.
   58. McCoy Posted: July 20, 2012 at 01:04 PM (#4188072)
It's very common for a player to (attempt to) play through injuries.

So are we going to chalk all these extreme fielding fluctuations to that? Darwin Barney sucked with the bat last year and he sucks with the bat this year but somehow he was injured last year so it hurt his fielding but now that he is healthy his glove is outstanding?

Are all these players constantly injured year in and year out so that it depresses their fielding runs and then have one healthy year that sends it skyrocketing?

Ben Zobrist put up a 25, 17, and 29 runs as a second baseman because he was healthy and is now putting up 3 runs because he is injured?
   59. TDF didn't lie, he just didn't remember Posted: July 20, 2012 at 01:26 PM (#4188103)
Some thoughts:

WAR is a conversation starter, not a conversation ender.
This is true because flawed as it is, it's much less flawed than any other stat we use to measure "value". Things like AVG, OBP, SLG, HR, ERA, WHIP are all more exact measures of events, but since there is no context they don't measure value very well at all. By adjusting for park, league, position, defensive and baserunning ability and so on, WAR is a much better conversation starter than anything else we have. Because of the flaws, though, it should never be a conversation ender.

An "average" 2B should be able to move to 1B and you'd gain enough in defense to offset the offensive hit.
I think this is a misrepresentation of what WAR is telling us, and I think those behind WAR haven't explained it properly. It seems to me that WAR is giving us a measure of the average offensive difference in different defensive positions, and that is the "position adjustment". Say the average 1B is generating 10 runs of offense, and the average 2B is generating 5, which would lead to a position adjustment of +5 for 2B. WAR isn't saying that the average 2B is 5 runs better defensively than the average 1B; it's saying only that the average 1B is a better hitter. If the league has a bunch of crappy hitting 2b, then you assume they're all great fielders (relatively), but that's not what the position adjustment in WAR is saying. There could just be a bunch of crappy hitting 2B in the league (I subscribe to the theory that a good SS wouldn't necessarily be a great 2B or 3B.)

That being said...

I don't think Ryan (.190/.287/.287, 65 OPS+ in 290 PA) is having a better year than Fielder (.308/392/.504, 141 OPS+ in 403 PA)
Could you imagine Fielder at SS? Even at Ryan's level of hitting, I think the number of runs Fielder would give up with his "glove" would easily outstrip the runs a good (not Gold Glove, but good) SS would give up with Ryan's batting line.
   60. TDF didn't lie, he just didn't remember Posted: July 20, 2012 at 01:37 PM (#4188119)
Ben Zobrist put up a 25, 17, and 29 runs as a second baseman because he was healthy and is now putting up 3 runs because he is injured?
I once asked Tango a question on The Book Blog relating to this very subject (outliers in defensive measurements, and what it means to WAR) and mgl said
Any sample of UZR, whether it be an outlier or not, has two characteristics. One, the performance does not necessarily correspond with the number because of errors and bias in the data and the methodology. Two, the performance does not necessarily correspond with the true talent, just like with offensive and pitching metrics because of random sampling error. Both of those things are always present in any sample of UZR. The larger the sample, the smaller those errors, although biases will tend to persist in any size sample by definition.

Specifically, for the +19 player it is likely that his actual performance that year was better than his true talent (that would be true even if we had no other years, which is one reason why we regress all sample values) and it is also likely that he did NOT perform at a +19 level.
In other words, Zobrist this year (1) may not be as good as in the past defensively (or Barney may be much better than in the past), but (2) there is too much noise (bias and errors in data collection) to say with any certainty. A few years' worth of data would be more accurate because it would smooth the errors.

Yet another reason WAR is a conversation starter, not ender.
   61. tjm1 Posted: July 20, 2012 at 01:38 PM (#4188121)
The defensive adjustments are basically about how difficult it is to find a good hitter who can field the position. I agree with TDF that good SS's often don't make even better fielders at 2B/3B. A good part of the 2B/SS difference is throwing ability. The issue is that a good SS should be able to handle 2B or 3B, and the reverse may not be true. The defensive adjustment issue becomes even clearer when you consider catchers.

   62. Der-K: downgraded to lurker Posted: July 20, 2012 at 01:41 PM (#4188123)
60: That encapsulates why much of why I doubt Ryan's been a +22 on D.
   63. cercopithecus aethiops Posted: July 20, 2012 at 01:47 PM (#4188127)
I suppose that Barney could be having a lot of balls hit to the exact edge of his range, instead of just beyond it, or something.


More likely he's not having any balls hit anywhere near the edges of his zone. IOW, he's benefiting from an unusually easy set of defensive chances. Think of what Joey Votto might do if you could somehow get him all of his PA against bad pitchers.
   64. McCoy Posted: July 20, 2012 at 01:50 PM (#4188132)
Could you imagine Fielder at SS?

Could you imagine Ryan as a first baseman? Would he still be +22 runs on defense at first? Would his offense improve? I don't think he would be a +22 runs as a first baseman.
   65. tjm1 Posted: July 20, 2012 at 01:51 PM (#4188134)
It's really not comparable. Sean is now using John Dewan's defensive numbers, and the spread of performance in those are higher than what you'll find in the pre - 2002 numbers. The farther you go back in time, the smaller the spread in defensive ratings, because of the data available to make those ratings.


Hmm. If the spread of performance is getting bigger and isn't repeatable, then maybe all the new stats are really adding is noise. Mazeroski and Belanger, for example, had very consistent ratings from year to year.
   66. Sean Forman Posted: July 20, 2012 at 01:52 PM (#4188137)
Ben Zobrist put up a 25, 17, and 29 runs as a second baseman because he was healthy and is now putting up 3 runs because he is injured?


I will never understand why there is an assumption that hitting variance is natural and fielding variance isn't. Adam Dunn's last four seasons are 32, 27, -26, 12. And yet we have no problem believing one and are incredulous at the other.

If you don't like the fielding stats use oWAR. ESPN actually is using this on the site as well.

Also, I appreciate Walt's comments about serious people attempting to do this work. Between the WAR formulation and the creation of DRS I'm guessing the overall system design alone has multiple man-years of effort from folks who take this sort of thing very, very seriously. This is not a half-assed effort.

Regarding fielding variance being larger now. Michael Humphries believes that most historical defensive estimators are much too conservative and that +/- 30 run swings are not out of the ordinary. FWIW.

Additionally, regarding Barney's defensive improvement, there were some pretty major changes in the management of the Cubs. Sveum came from the Brewers who were very innovative using defensive positioning and I'm sure you've heard of Theo Epstein. Both Castro and Barney have seen dramatic improvements in the field of 20+ runs so far. I wouldn't be surprised if 1) Sveum worked a lot with them during spring training to improve their defensive play, and 2) their positioning has gone from among the worst in the league to among the best.

Now you can argue that positioning is a management rather than player skill and you'd be somewhat right, but we need to credit someone with that run prevention.
   67. Rally Posted: July 20, 2012 at 01:53 PM (#4188140)
What is the purpose of this, how does it affect the numbers? Does this mean a .300/.400/.500 average defensive first baseman in the al would get more war than the same numbers in the nl?


Yes. Other things being equal, that stat line might get you 5.0 WAR in the AL, and 4.6 WAR in the NL. The effect would be the opposite in the 50's/60's, when the National League had more talent.
   68. cardsfanboy Posted: July 20, 2012 at 01:53 PM (#4188141)
Could you imagine Ryan as a first baseman? Would he still be +22 runs on defense at first? Would his offense improve? I don't think he would be a +22 runs as a first baseman.


Exactly, Ryan would more than likely lose on defense (his skills don't really help there) and relative to position his offense would look even worse.

I don't think anyone who thought about it, would argue that war is predictive of performance if you moved player through the defensive spectrum. Again, it assumes that all positions are equally as valuable. I don't see how anyone could believe that to be the case, but that is the basic assumption that war is built upon.
   69. McCoy Posted: July 20, 2012 at 01:56 PM (#4188146)
I will never understand why there is an assumption that hitting variance is natural and fielding variance isn't.

Well, I don't assume that the offensive components don't share some of the same problems that the defensive side does but again it is a lot easier to see why the offensive numbers are what they are.


Now you can argue that positioning is a management rather than player skill and you'd be somewhat right, but we need to credit someone with that run prevention.


It doesn't need to be the player especially if it is the manager that is telling them where to stand.
   70. cardsfanboy Posted: July 20, 2012 at 01:58 PM (#4188150)
Yes. Other things being equal, that stat line might get you 5.0 WAR in the AL, and 4.6 WAR in the NL. The effect would be the opposite in the 50's/60's, when the National League had more talent.


Regardless of the criticism, I do love bWar, but I do not like this bit at all. I've seen people argue for Mays being better than Mantle because of war plus a league adjustment. I've seen people argue against Pujols using the same adjustment. Not your fault that people are doing it wrong, but that is like timelining to me, it's something I would prefer to make on my own, instead of having made for me.
   71. Tippecanoe Posted: July 20, 2012 at 02:02 PM (#4188157)
I will never understand why there is an assumption that hitting variance is natural and fielding variance isn't. Adam Dunn's last four seasons are 32, 27, -26, 12. And yet we have no problem believing one and are incredulous at the other.


I was going to say it this way -- Why do we believe it when Andrew McCutchen's slugging goes up 200 points at the 90-game mark, but we can't believe Darwin Barney's increase in dWAR? The reality is probably that neither rate is sustainable, but in each case a young player has markedly improved.
   72. Sean Forman Posted: July 20, 2012 at 02:09 PM (#4188166)
And maybe I'm hard-headed or a Luddite or whatever, but I do NOT believe Brendan Ryan is having a better year than Prince Fielder. I just don't.


If you don't think defense has any bearing or you know better what magnitude of adjustment to apply, just start with Fielder's 2.4 oWAR and Ryan's 0.4 oWAR and take it from there.
   73. McCoy Posted: July 20, 2012 at 02:10 PM (#4188167)
When you watch Barney play does he look like a guy that is having one of the greatest defensive seasons of all time by a second basemen?
   74. Sean Forman Posted: July 20, 2012 at 02:13 PM (#4188171)
Not your fault that people are doing it wrong, but that is like timelining to me, it's something I would prefer to make on my own, instead of having made for me.


These numbers are made up or gut estimates. It's based on things like looking at every player who moves from one league to the other and aggregating their change in performance, or in current years--evidence from interleague play.

I've done some studies of player opponent quality and due to playing in the NL Central the Cards batters had a good run where they faced the worst pitching of any team in the majors and I think you are seeing that play out a bit now with the drop in Pujols and Fielder's numbers moving to the AL.
   75. Sean Forman Posted: July 20, 2012 at 02:18 PM (#4188181)
When you watch Barney play does he look like a guy that is having one of the greatest defensive seasons of all time by a second basemen?


Is this really the nub of your complaint about Barney's defensive measure? How is that any different than saying the 40 games I watched the Phillies, Freddy Galvis hit a lot of line drives, so he must be having a great season at the plate.

Actual trained/paid people are watching every single play of every single game with stop watches to track every play that is being made as carefully as possible, so yes, if you were to watch the play of every major league 2Bman this year and carefully record the results you would probably argue Barney is having a great defensive season. It certainly isn't infallible, but it goes a long ways beyond how most fans or even mlb coaches who vote for gold gloves look at things.
   76. McCoy Posted: July 20, 2012 at 02:30 PM (#4188188)
Highly trained paid people also were saying that Brett Lawrie was having the greatest defensive season by anyone ever as well.

Why is Darwin Barney having one of the greatest seasons by second basemen of all time? Do you know? Does ARom know? Does Walt? Does anyone besides a black box and the guy cranking the handle to that box know?

I'm not sure why you would need a stopwatch for Barney though. He isn't ranging to his left or right by 15 steps or something like that. It basically boils down to either his coaches are positioning him really well or a lot of balls are getting hit near him.
   77. Jose is an Absurd Doubles Machine Posted: July 20, 2012 at 02:44 PM (#4188202)
It basically boils down to either his coaches are positioning him really well or a lot of balls are getting hit near him.


That's too simplistic for my taste. There are a number of other things that could be impacting Barney;

- He could be positioning himself better
- He could have done something to improve his throwing strength or accuracy this off-season
- He could have improved his ability to field the baseball this off-season

and probably a host of other things I'm not thinking of. We know for a fact that players often improve or regress defensively. I think in years past where the most widely used measures of defense were errors and Gold Gloves it took a little more time to become obvious. I wouldn't be at all surprised that Barney has improved defensively in a major way. I think it's a bit unlikely that he's having a historically great defensive season but I see no reason that substantial year to year improvement isn't possible for him or any other player.
   78. Rally Posted: July 20, 2012 at 02:44 PM (#4188203)
Not your fault that people are doing it wrong, but that is like timelining to me, it's something I would prefer to make on my own, instead of having made for me.


If you prefer to do it yourself, the components of WAR are broken out, so you can always remove the replacement level part and sub in a constant that is not league dependent.

I think there's a difference between this and timelining. We have a pretty good idea that the quality of play in baseball is higher now than it is in say, 1912. But we don't know for sure, and we can't test it. While some people have made interesting efforts to quantify it (David Gassko, GuyM), it can't be done with any certainty.

League quality is different. We can test it every June. If the adjustment is true and accurate, then the AL should win about 55% of interleague games. And give or take a few games every year, that is what they have done. The scientific method at work.
   79. McCoy Posted: July 20, 2012 at 02:53 PM (#4188206)
We know for a fact that players often improve or regress defensively

Not usually to the tune of greatest of all time though.

and probably a host of other things I'm not thinking of.

And that's the thing. We can't look at the WAR fielding numbers and figure out what he is doing that is causing him to put up those numbers. So you can't have a conversation about defense and really use WAR numbers.
   80. Ron J2 Posted: July 20, 2012 at 03:04 PM (#4188213)
#63 Sure. There are two very difficult issue to deal with in any form of fielding evaluation.

1) There are a certain number of discretionary chances. Easy chance that two (or more) players can make. Chris Dial used to argue that the big disconnect between the grid based systems (which saw him as nothing special) and the adjusted range factor systems (which saw him as a defensive god) WRT Andruw Jones was due to Jones taking an unusual number of the discretionary plays.

2. There's no particular reason to think that the number of difficult chances are constant. PMR attempts to address this, but I think PMR will run into problems given that teams are doing very different things with their infield alignment.

   81. BDC Posted: July 20, 2012 at 04:43 PM (#4188312)
in years past where the most widely used measures of defense were errors

Funny enough, Barney is fielding .998 at second in 2012, and if he keeps it up all year, it will be an all-time record for a full season. Last year, he fielded .981 at second, which is not even all that good nowadays. So even by the most primitive of measures (and one in which, unless there's some weird bias going on, there is no particular "black box" at work, just routine scoring decisions), it's possible to go from mediocre to world-beating in a little over half a season.
   82. villageidiom Posted: July 20, 2012 at 04:59 PM (#4188324)
I think it's likely that a site like bb-ref should not publish in-season WAR. The defensive component is too misleading.
Mark Melancon's ERA this season is 5.89 in 18 appearances. If you start with just his return from the minors in June he has had an 0.55 ERA in 14 appearances. By the same token one could argue ERA is too misleading, and should not be published in-season.

Or, as an alternative, they can publish the numbers and people just should know not to take them at face value. With reliever ERA, they already do, mostly. With defensive stats they haven't learned to do it. If you don't publish the numbers, (a) they don't learn, and (b) the people who know better can't look at them easily.

I'm not sure why you would need a stopwatch for Barney though.
Then learn about the damn stats before disparaging them.

They use the stopwatch not to time Barney, but to time the ball he's fielding so they can tell how fast it was traveling. If they know trajectory and speed they can compare the batted ball to other batted balls in the historical record (which admittedly only goes back so far) and tell how more often Barney converts that to an out than they had been in the historical record.

There could be positioning effects that make him appear better, and this should be studied more. Brett Lawrie in the shift is the extreme case that makes this obvious. If they also controlled for starting position of the fielder - not simply "3B" or "2B" but the coordinates on the playing field - they could potentially make much more accurate measurements.
   83. Jose is an Absurd Doubles Machine Posted: July 20, 2012 at 05:14 PM (#4188334)

And that's the thing. We can't look at the WAR fielding numbers and figure out what he is doing that is causing him to put up those numbers. So you can't have a conversation about defense and really use WAR numbers.


This doesn't make sense to me. Maybe I'm missing your point but just because we don't know WHY a player has dramatically improved defensively doesn't mean it hasn't happened. When I read about Roger Maris' 1961 season I don't read anything about WHY he suddenly started hitting home runs just that he did.
   84. Ron J Posted: July 20, 2012 at 05:20 PM (#4188337)
#82 What you're measuring there is range. It's not quite the same as fielding ability. Cal Ripken for instance wouldn't grade out particularly well. He played the position differently than anybody else. He made a lot of seemingly easy plays because he played deeper than anybody else (and was very good at positioning himself)

Dunno. It seems that the management team (meaning mostly coaches and the manager) are far more involved in the positioning decisions than they were in Ripken's day. Maybe the objection isn't valid, but I'd hate to just adjust away a guy who makes a lot of seemingly easy plays that somebody with a weaker arm couldn't make.
   85. cardsfanboy Posted: July 20, 2012 at 05:29 PM (#4188344)
What you're measuring there is range. It's not quite the same as fielding ability. Cal Ripken for instance wouldn't grade out particularly well. He played the position differently than anybody else. He made a lot of seemingly easy plays because he played deeper than anybody else (and was very good at positioning himself)

Dunno. It seems that the management team (meaning mostly coaches and the manager) are far more involved in the positioning decisions than they were in Ripken's day. Maybe the objection isn't valid, but I'd hate to just adjust away a guy who makes a lot of seemingly easy plays that somebody with a weaker arm couldn't make.


Agreed, but if you don't then a player is getting extra value for not actually being better, he's effectively receiving a beneficial park adjustment that you aren't adjusting for. I do not want to take positioning away either, but I don't think a player should get the benefits of extreme positioning that if he was on another team he wouldn't have been doing.

I'm hoping that the eventual compromise, for the systems we have now, is to divide the field into rough positions and grade from there. There is really no reason that the current systems have to lock themselves into just the 7 defensive positions. Realistically shortstop, second base and centerfield could all be captured by three separate starting positions and the corner positions by two starting positions and then judging relative to how other players in those rough zones performed. Players still position themselves where they want relative to where the coaches tell them, so there is some of their own skill involved.

You could also completely ignore plays made with the extreme shift on.
   86. tjm1 Posted: July 20, 2012 at 06:02 PM (#4188356)
Agreed, but if you don't then a player is getting extra value for not actually being better, he's effectively receiving a beneficial park adjustment that you aren't adjusting for. I do not want to take positioning away either, but I don't think a player should get the benefits of extreme positioning that if he was on another team he wouldn't have been doing.


His point is that Ripken was able to position himself deep because he could make throws other shortstops couldn't. That's not a function of his managers or coaches. It's one of the things that make baseball so interesting - Cal Ripken and Ozzie Smith were probably the two best defensive shortstops of the 1980s, and they did it with almost completely opposite skill sets.
   87. McCoy Posted: July 20, 2012 at 06:19 PM (#4188362)
Then learn about the damn stats before disparaging them.

Come again?


This doesn't make sense to me. Maybe I'm missing your point but just because we don't know WHY a player has dramatically improved defensively doesn't mean it hasn't happened. When I read about Roger Maris' 1961 season I don't read anything about WHY he suddenly started hitting home runs just that he did.

Home runs aren't a black box stat. Nobody says that Player X had 70 BB and 20 2B therefore he hit 45 homers. A home run is an event that is tracked. Defensive runs is a formula.
   88. Drew (Primakov, Gungho Iguanas) Posted: July 20, 2012 at 07:05 PM (#4188374)
Defensive runs is a formula.


Based on events that are tracked. A defensive play is also an event.
   89. McCoy Posted: July 20, 2012 at 07:11 PM (#4188378)
Based on events that are tracked. A defensive play is also an event.

That are then put into different buckets and then compared to average. All of which happens in a black box.
   90. Moeball Posted: July 20, 2012 at 08:45 PM (#4188469)
The thing I think is interesting is how measurement methods would apply in managerial strategy as it applies to lineup selection.

For example - I was really excited when "Hidden Game of Baseball" came out in 1984 because one of the things I noticed right away was that players such as Rabbit Maranville and Bill Mazeroski could be shown to be strongly positive players overall (significantly better than average)despite being poor (negative) players offensively. This was the first time I saw anyone put an actual numerical value on a player's defensive contributions in terms of runs saved. I had heard much of my baseball fan life about players who were "good field, no hit" and how some managers would insist on playing this guy in spite of a weak stick because he was supposedly saving many more runs with his glove than he was costing the team with his bat. The whole HOF case for guys like Mazeroski and Maranville was based upon the premise that they might cost a team 10-20 runs a season compared to an average hitter but that they were often saving 40 runs or more defensively over an average fielder, meaning they could conceivably be 20 or 30 runs a season better than an average player overall (approximately +2 or +3 WAA per season).

Fast forward to today - Defensive Linear Weights runs have given way to newer measurements based on play-by-play data that are (in theory) much more accurate. Yet these newer measurement systems (such as dWAR) are still beset by many problems and variables that make them, uh, questionable as to the validity of the results as can be seen by many of the comments above in this thread.

So one thing I'm wondering is if there are some basic philosophical differences at issue here. One of the main issues is this - just how much of a difference can a great fielder or a poor fielder actually make? It is clear that many people on this site believe that the gap between the absolutely greatest fielders and the poorest fielders is only about 20-30 runs a season (good fielders can be up to only about 10-15 runs above average, poor fielders about 10-15 runs below average). If this is the case - then not only is a Mazeroski or a Maranville not playing at a consistent All Star level (potential HOF candidate), but perhaps they are actually below average players overall, since their offensive costs may outweigh their defensive benefits. You see, while opinions of the defensive valuations are all over the place, no one is arguing that much about the offensive measurements and they stay fairly constant. So while we may now think Maz was only worth +15 runs a season defensively instead of +40, we may still think he was at -20 offensively.

There are many people that react with disbelief any time they hear that a player may be 40 or 50 runs better than average defensively in a season. They simply refuse to believe that any player could be that good. Yet 50 run gaps between two players offensively are commonplace and accepted by just about everyone.

So back to my original point - how does a manager's perception impact his choices in lineup selection?

Every manager would love to have a Johnny Bench or an Albert Pujols in the lineup at each position - someone who is much better than average both offensively and defensively. Unfortunately, many times a manager is stuck with having to choose between the following two options: a player who is "good hit, no field" or a player who is "good field, no hit".

If a manager believes that there is only a narrow range between great fielders and terrible fielders, then he will virtually always choose the "good hit, no field" player over the "good field, no hit" player because the impact defensively cannot possibly outweigh the offensive gap.

On the other hand, if the manager believes that a player's defensive impact can truly be large, he might go with the "good field, no hit" player (or, if a terrible fielder, the manager may think "this guy's a born DH and I don't dare put him anywhere in the field because he will cost me too many games").

I'm curious as to how others view this. Any thoughts?
   91. Jittery McFrog Posted: July 20, 2012 at 09:01 PM (#4188474)
#82 What you're measuring there is range. It's not quite the same as fielding ability.


It would be great to have both. This would at least put a lower bound on how many runs of "fielding ability" could be attributed to the fielder.
   92. Ron J Posted: July 20, 2012 at 09:03 PM (#4188476)
#89 All offensive events are put into buckets too in any metric that converts the events into value. Take runs created. At the team level, the value of a home run varies with the team's OBP. (And the final values are adjusted for offensive context. If the imprecision of defensive numbers bothers you, it's worth noting that a healthy chunk of the standard error in offensive metrics comes from park effect calculations.)

On the defensive side we're not guessing at the cost of a play not made, nor are we guessing about the value of a double play turned. And we've got a decent first order estimate as to the value of an outfielder's arm

Black box bothers you? There are decent defensive metrics that are completely open. Most are essentially adjusted range factor systems but there's enough info in the retrosheet data files that you can make a pretty decent context adjustment. Sean's made the decision to go with a proprietary system in WAR (presumably because he feels it's a better estimate of defensive value), but you're not bound by Sean's decision. It wouldn't be nuts to look at (say) Charlie Saeger's system and see if you can incorporate the info in the Retrosheet data files to tighten up the context adjustments.

I'm not in love with any of the black box systems. I'd like to see the raw numbers and the adjustments (I've seen too many smart guys blow context adjustment calculations) but I accept them because I don't currently feel like doing the heavy lifting.
   93. cercopithecus aethiops Posted: July 20, 2012 at 09:26 PM (#4188486)
There are many people that react with disbelief any time they hear that a player may be 40 or 50 runs better than average defensively in a season. They simply refuse to believe that any player could be that good. Yet 50 run gaps between two players offensively are commonplace and accepted by just about everyone.


So far this season, about 30% of all PA in MLB have resulted in one of the three true outcomes. A substantial percentage of all hits on BIP are not preventable by any defender. This tells me that there simply are many more runs scored than there are runs saved. A system that told me that the gaps on defense were as big as the gaps on offense wouldn't pass my personal smell test.
   94. Everybody Loves Tyrus Raymond Posted: July 20, 2012 at 10:23 PM (#4188513)
A substantial percentage of all hits on BIP are not preventable by any defender. This tells me that there simply are many more runs scored than there are runs saved. A system that told me that the gaps on defense were as big as the gaps on offense wouldn't pass my personal smell test.


Further a substantial percentage of outs on BIP are routine plays that don't require anything special of the defender.
   95. Sunday silence Posted: July 21, 2012 at 03:50 AM (#4188582)
And we've got a decent first order estimate as to the value of an outfielder's arm


if this is really true, than could answer one basic hypothetical? Let's say it's the 1960s and in RF CLemente has 15 assists and Aaron has 10 assists. Both play for a full season. How many baserunners would CLemente have prevented, if any, from moving up a base during that season?
   96. Sunday silence Posted: July 21, 2012 at 04:16 AM (#4188585)

So far this season, about 30% of all PA in MLB have resulted in one of the three true outcomes. A substantial percentage of all hits on BIP are not preventable by any defender. This tells me that there simply are many more runs scored than there are runs saved. A system that told me that the gaps on defense were as big as the gaps on offense wouldn't pass my personal smell test.



OK I understand and agree with the first two sentences. HOw does it follow that "more runs are scored than saved? "

THis does not seem to make any logical sense in a game where everything that contributes to a run is zero sum.

As for the last sense, when you say "gaps" you are referring to the range from worst to best offensive or defensive players? But are including pitchers in this? If pitchers are large part of defense, it might be true that position players have less "spread" than hitters. If you include pitchers, they probably have as much a spread as hitters.

Hell the more I think about it, I see no reason in theory why defensive position players couldnt vary as much or more than hitters. What is your basis for this statement?
   97. Sunday silence Posted: July 21, 2012 at 04:28 AM (#4188586)
Sure all of them had a good bit of flukish luck as well as environment help them out but the point is we can see the conditions that made it possible for them to do what they did. With fielding you can't really see that or really why would flukey conditions exist for fielding? There are no juiced gloves, they don't change the rules on what is an out or not, so on and so on. That isn't to say fielding should be consistent from year to year but it does make it hard to accept numbers that vary wildly from year to year.


In the particular case of Davey Johnson and Brady ANderson I sort of agree with what you are saying. But in the general sense what makes you think fielding conditions are more inscrutable than batting conditions?

Say Rod Carew comes up to bat 600 times. We sort of accept that .390 could very well be an accurate estimate of his ability.

Andy Van Slyke stands in CF for a whole season and 500 balls are hit to him. Why is this stat any less accurate than hitting? It seems to me that the same problems are inherent in both. Aside from sample size, perhaps 20% larger sample size for batting, other than that they seem both susceptible to the same criticism.

Before you can bring up park effects; you could use park effects to adjust for fielding range couldn't you? I dont know if anyone has but that part seems to be also be a wash for both sides.

What does that leave us with? In terms of factors that would cause fielding numbers to any less than accurate than hitting.
   98. Swedish Chef Posted: July 21, 2012 at 05:11 AM (#4188591)
Andy Van Slyke stands in CF for a whole season and 500 balls are hit to him.

Well, let's just begin there, how do you know if a ball was hit to him? That's a judgement call.
   99. tjm1 Posted: July 21, 2012 at 05:19 AM (#4188592)
It's probably not possible to come up with a rigorous mathematical proof that says that the spread in fielding ability is smaller than the spread in hitting ability, but I think it is straightforward to come up with a good argument that says that that's likely to be the case.

First let's look on a team level. The spread between the best teams in runs scored is pretty much the same as the spread among the best teams in runs allowed. We know that within a team, there will almost always be a large spread in the pitchers' runs allowed. This immediately implies that on a team level, fielding is less important than hitting, because fielding+pitching is roughly equally important as hitting. I think the general consensus is that pitching is the bigger part of the fielding+pitching side of the equation, and that this has become more true with time as the number of homeruns and strikeouts increase, and the number of balls in play decreases.

Also, regarding reliability of hitting versus fielding stats: out of Carew's 600 at bats in your example, the variation in his chance of getting a hit is probably no more than a factor of 2. Out of the 500 balls a year that might be hit to a centerfielder, probably 20% are going to be hits no matter what and 30% are high flies that anyone would catch. The number that actually could go either way is probably no more than half the total. The balls where every fielder will make or not make the play are part of the statistical record, but are not contributing any information about how good the fielder actually is. They can, however, contribute noise to the process.
   100. cercopithecus aethiops Posted: July 21, 2012 at 08:38 AM (#4188609)
HOw does it follow that "more runs are scored than saved?


I assumed that I would be understood to mean runs saved by position players' efforts in the field. Another way of looking at it is that a big chunk of run-saving is accomplished by pitchers in defense-independent ways. Ring a bell?

Hell the more I think about it, I see no reason in theory why defensive position players couldnt vary as much or more than hitters. What is your basis for this statement?


As a matter of fact, I personally assume that players' ability/skill varies exactly as much on defense as it does on offense (only because it's the simplest assumption to make). But if we guess that only 60% or so of run-saving is attributable to defense rather than pitching, then the absolute magnitude of defensive differences should be smaller than the absolute magnitude of offensive differences when we convert them all to runs. If an OPS+ of 150 roughly converts to around 4 oWAR, then being 50% better than league average on defense should only be worth 2 or 3 dWAR (depending on position, of course, since defensive opportunities are not distributed equally).
Page 1 of 2 pages  1 2 > 

You must be Registered and Logged In to post comments.

 

 

<< Back to main

News

All News | Prime News

Old-School Newsstand


BBTF Partner

Support BBTF

donate

Thanks to
Don Malcolm
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Hot Topics

NewsblogOT - NBA 2017-2018 Tip-off Thread
(2022 - 2:51am, Dec 16)
Last: stevegamer

NewsblogWinter Meeting Signings
(30 - 2:47am, Dec 16)
Last: stevegamer

NewsblogRyan Thibs has his HOF Ballot Tracker Up and Running!
(477 - 2:27am, Dec 16)
Last: vortex of dissipation

NewsblogCardinals trade for Marcell Ozuna of Marlins
(63 - 2:00am, Dec 16)
Last: Voodoo

NewsblogOTP 11 December, 2017 - GOP strategist: Moore would have 'date with a baseball bat' if he tried dating teens where I grew up
(2298 - 1:44am, Dec 16)
Last: zenbitz

NewsblogPrimer Dugout (and link of the day) 12-15-2017
(10 - 10:40pm, Dec 15)
Last: Der-K: downgraded to lurker

NewsblogDerek Jeter Was Once the Captain. But Now He’s the Apprentice. - The New York Times
(96 - 8:57pm, Dec 15)
Last: fra paolo

NewsblogTaking Back the Ballparks - Mariners voting thread
(20 - 8:41pm, Dec 15)
Last: Omineca Greg

NewsblogA's reportedly acquire OF Stephen Piscotty
(6 - 7:13pm, Dec 15)
Last: cardsfanboy

NewsblogOT: Winter Soccer Thread
(361 - 6:41pm, Dec 15)
Last: Biff, highly-regarded young guy

Gonfalon CubsLooking to next year
(347 - 5:46pm, Dec 15)
Last: Moses Taylor, aka Hambone Fakenameington

NewsblogESPN: Bob Costas wins Hall of Fame's Frick Award for broadcasting
(23 - 5:02pm, Dec 15)
Last: ERROR---Jolly Old St. Nick

NewsblogMets agree to two-year deal with Anthony Swarzak
(22 - 4:52pm, Dec 15)
Last: Dog on the sidewalk

NewsblogOT Gaming: October 2015
(717 - 3:52pm, Dec 15)
Last: GGIAS (aka Poster Nutbag)

Sox TherapyA Container of Milk, A Loaf of Bread and a Dude Who Can Hit Home Runs
(30 - 2:39pm, Dec 15)
Last: jmurph

Page rendered in 1.0418 seconds
47 querie(s) executed