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

Sunday, November 19, 2017

dWAR to end all WARs – Joe Blogs – Medium

People might not remember but I created one of the first WAR, along with G. Jay Walker. Although I leave the evaluation of real MLB players to others now, I still use an offshoot of the system to evaluate players for my simulation league. For my system I essentially agree with Mitchel Lichtman’s thoughts at the bottom of the article. (Excluding tone, of course.)

I don’t think you’re really understanding how this works. I’m not sure Sean [Forman] does either. There’s no such thing as defensive wins or runs above “replacement” because replacement is defined by offense and defense. And as it turns out (not that it really matters) replacement players on the average are around average on defense. All their deficiency is in offense.

There is also no such thing as positional adjustment offense. You CAN give an offensive value relative to other players at that position but you have to specify over what time period.

All that being said, you are 100% correct in that most people want to know how good defensively a player is compared to others at their position and presenting a position adjusted number confuses that. Bottom line is this: Defense should be presented relative to the average at that position AND serious people SHOULD know how to compare players ACROSS positions (by applying positional adjustments).

Offense should be presented compared to average of ALL positions. Most people intuitively know that different positions have different offenses because of the size of the pool of players that can play that position (and the physical characteristics necessary to play them). But it’s not at all necessary to know those differences or include them in the offensive numbers.

Finally it IS necessary to include positional adjustments in the “final comprehensive number” (like WAR) in order to be able to compare all players AND because when we want to know “how good” a player is we MUST incorporate his defensive position.

I know Sean understands how this all works. He disagrees, however, on how positional adjustments should fit into a WAR framework. He’s not wrong but his choice makes WAR answer too narrow a question.

Jim Furtado Posted: November 19, 2017 at 07:00 AM | 53 comment(s) Login to Bookmark
  Tags: sabermetrics, war

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.

   1. Sean Forman Posted: November 20, 2017 at 11:21 AM (#5578738)
https://twitter.com/sean_forman/status/932642913754013697
   2. Blanks for Nothing, Larvell Posted: November 20, 2017 at 11:43 AM (#5578756)
Keith Hernandez’s CAREER dWAR is 0.6.


Return to sender.
   3. KJOK Posted: November 20, 2017 at 12:17 PM (#5578773)
I'm sure MGL is wrong about Sean not understanding how this works. Seems like it would be simple to have: oWAR (Positionally Adjusted) oWAR (not adjusted) dWAR (positionally adjusted) dWAR (NOT adjusted) and finally POSITIONAL ADJUSTMENT. That should make it clear to everyone and end the sillier parts of this debate.

   4. Fancy Crazy Town Banana Pants Handle Posted: November 20, 2017 at 12:49 PM (#5578800)
Defense should be presented relative to the average at that position AND serious people SHOULD know how to compare players ACROSS positions (by applying positional adjustments).

If you want to know the defense relative to the average at that position, all you have to do is glance over to the rField column. SERIOUS people SHOULD know how to do that before spouting off.
   5. shoewizard Posted: November 20, 2017 at 01:05 PM (#5578816)
Keith Hernandez’s CAREER dWAR is 0.6.



Return to sender.

Why ? The metric identifies him as the best first baseman of the modern era. Is your position that the positional adjustment unduly penalizes first base, so you don't agree with the dWAR career total relative to other positions ?

I'm not sure what your objection here is.


Rk              Player dWAR   PA From   To
1      Keith Hernandez  0.6 8553 1974 1990
2        Anthony Rizzo  0.3 3895 2011 2017
3            Sid Bream 
-0.3 3531 1983 1994
4        Mark Teixeira 
-0.4 8029 2003 2016
5       Casey Kotchman 
-0.6 3412 2004 2013
6         Brandon Belt 
-0.7 3149 2011 2017
7            Vic Power 
-0.8 6459 1954 1965
8    Doug Mientkiewicz 
-0.9 3844 1998 2009
9     Paul Goldschmidt 
-1.3 4018 2011 2017
10         Pete OBrien 
-1.8 6168 1982 1993 


Provided by Baseball-Reference.com: View Play Index Tool Used
Generated 11/20/2017.
   6. Misirlou doesn't live in the restaurant Posted: November 20, 2017 at 02:12 PM (#5578866)
I'm not sure what your objection here is.


The word troll get's thrown around far to liberally here, but that's was bear is doing with this. He knows full well that no one, including WAR, is making a claim that Hernandez was merely an average fielder. Ultimately, dWAR is a calibration tool. It means that Hernandez's offense can be directly compared to an average fielding CF, or a poor fielding SS (who would both have similar dWAR) to assess overall value.
   7. cardsfanboy Posted: November 20, 2017 at 08:30 PM (#5579124)
MGL still manages to look like a condescending ####### no matter what he does.
   8. cardsfanboy Posted: November 20, 2017 at 08:35 PM (#5579129)
3. KJOK Posted: November 20, 2017 at 12:17 PM (#5578773)
I'm sure MGL is wrong -----


You can usually stop there and be in the right a vast majority of the time, without bothering to explain anything beyond that.
   9. cardsfanboy Posted: November 20, 2017 at 08:48 PM (#5579132)
If you want to know the defense relative to the average at that position, all you have to do is glance over to the rField column. SERIOUS people SHOULD know how to do that before spouting off.


Not to defend MGL, but his point is that serious people do know that, and that a stat like dWar is used by non-serious people, and that it should be more readily accessible to their point of view.... I'm not disagreeing with him here. We've had the dWar debate for years around here, and for the most part, I'm in the camp that dWar is pretty much one of the most useless stats out there because of the issues Pos brought up.

I don't know how to explain it properly from my viewpoint, but you have the mainstream stats, you have secondary mainstream stats, you have the saber stats, and secondary saber stats....those are stats that come up a lot of times in different conversations, then you have the tertiary stats, these are stats that are brought up for various reasons, but something that not many people know about unless they are interested in it, or are using for an article.... dWar as it is presented should be a tertiary stat, but because it is a component of a primary stat like War, it's getting lumped into a secondary stat status, and it really shouldn't. rField should be the secondary stat that is used more often(or it's equivalent in other systems) but instead we are looking at a positional adjusted stat, that is frequently misinterpreted, as a stat used in a discussion.... for the most part I generally turn off someone listing dWar in a discussion unless they are talking players at the same position, which goes contrary to it's purpose, but because it's so often mis-used, that is the only real choice you have in the matter when discussing dWar.
   10. Blanks for Nothing, Larvell Posted: November 20, 2017 at 08:59 PM (#5579135)
The word troll get's thrown around far to liberally here, but that's was bear is doing with this. He knows full well that no one, including WAR, is making a claim that Hernandez was merely an average fielder. Ultimately, dWAR is a calibration tool. It means that Hernandez's offense can be directly compared to an average fielding CF, or a poor fielding SS (who would both have similar dWAR) to assess overall value.


And, as Posnanski said, that's dumb. Both in the regular barstool sense that no one really cares about doing that, but in the sense that it's silly to think that Keith Hernandez's defense led to less than a single win over a replacement level 1B for his career. The notion is just absurd on its face.

On the technicalities, it's dumb to port a shitty fielding SS to 1B and compare him to Hernandez because -- well, for a lot of reasons -- but one major one is that none of the shitty fielding SS's are left-handed. Keith Hernandez being left-handed in and of itself was worth more than 0.6 career dWAR.

Someone made a comment on one of these threads, I think it was PreservedFish, that Bill James wasn't that great at math, but he thought deeply and philosophically about this stuff. The Keith Hernandez 0.6 thing is just a perfect example of not doing that. If the idea and the math spits out "Keith Hernandez, career, 0.6," at that point like I said, it's return to sender time.



   11. Greg K Posted: November 20, 2017 at 09:15 PM (#5579141)
And, as Posnanski said, that's dumb. Both in the regular sense that no one really cares about doing that, but in the sense that it's silly to think that Keith Hernandez's defense led to less than a single win over a replacement level 1B for his career. The notion is just absurd on its face.

But is that what dWAR is measuring? Wouldn't a replacement level 1B have a massive negative dWAR?

I guess I don't entirely understand what the criticism is here.

To take a roughly replacement level 1B, Jim Spencer (5408 PA, 2.8 WAR), BB-Ref has him as an average fielding 1B (0 Rfield for his career, though he did win two Gold Gloves). He has a -9.4 dWAR. Keith Hernandez was a much, much better 1B (117 Rfield), which spits out that dWAR of 0.6.

In other words, he's not less than a win better than a replacement level 1B for his career. He's around 10 wins better than an average fielding, replacement level 1B for his career.

Or to take someone with a more comparable career length, Charlie Grimm has around 8000+ PA. +18Rfield, and dWAR of -8.2. Again, gives you around the same figure, Hernandez being 9 wins better on defence than a slightly above average fielding 1B. That sounds pretty reasonable to me. 10 wins is a lot.

EDIT: In other words, if you want to know how much better Keith Hernandez was than an average fielding 1B, look at Rfield, don't look at dWAR. That's not what the stat is measuring.
   12. cardsfanboy Posted: November 20, 2017 at 09:15 PM (#5579142)
Someone made a comment on one of these threads, I think it was PreservedFish, that Bill James wasn't that great at math, but he thought deeply and philosophically about this stuff. The Keith Hernandez 0.6 thing is just a perfect example of not doing that.


Bullshit, sorry blanks, but a stat is a stat, they all have their purposes and because some simpletons can't fully understand a stat, doesn't mean the stat is for crap.

If someone misuses a stat that is on them, if someone doesn't understand a stat, that is on them.... I don't disagree with Joe, Bill or even you on this issue, but the fault isn't the stat, the fault is the random people understanding and usage of the stat.

One of the things that pisses me off more than anything is when people quote fellow Missourian, Mark Twain and say there are lies, damn lies and statistics.... the issue I have is that stats only lie to you, if you are not willing to understand the basics of the stat.
   13. cardsfanboy Posted: November 20, 2017 at 09:24 PM (#5579143)
But is that what dWAR is measuring? Wouldn't a replacement level 1B have a massive negative dWAR?


Not likely....the point of the article is that almost every first baseman is going to have a negative war. It doesn't matter their level...

Here is the entire list of first baseman who have played 500 or more games at first with a positive dwar. From 1871 to now....

Rk             Player dWAR
1        Darin Erstad 10.3
2        Roger Connor  6.3
3         Ernie Banks  4.9
4           Cap Anson  4.9
5    Charlie Comiskey  4.6
6        Patsy Tebeau  3.6
7        Perry Werden  3.6
8         Fred Tenney  3.5
9        Frank Chance  2.9
10          Joe Start  2.8
11          Joe Mauer  2.3
12        Dots Miller  2.0
13     Kevin Youkilis  1.6
14        Gene Tenace  1.6
15       Frank Isbell  1.5
16         Sid Farrar  1.5
17       John Morrill  1.4
18       Daric Barton  1.3
19           Dave Orr  1.1
20        John Ganzel  1.0
21    Keith Hernandez  0.6
22     George Stovall  0.6
23      Jiggs Donahue  0.6
24   Carl Yastrzemski  0.5
25         Wally Pipp  0.4
26      Anthony Rizzo  0.3
27       Chick Gandil  0.1
28         Dan McGann  0.1 


Note this list doesn't remove their years from other positions, which is why Erstad and Banks are at the top.

   14. Blanks for Nothing, Larvell Posted: November 20, 2017 at 09:25 PM (#5579144)
I don't disagree with Joe, Bill or even you on this issue, but the fault isn't the stat, the fault is the random people understanding and usage of the stat.


No, if it spits out 0.6 career for Keith Hernandez's defense, the fault is the stat.

   15. Greg K Posted: November 20, 2017 at 09:30 PM (#5579145)
Not likely....the point of the article is that almost every first baseman is going to have a negative war. It doesn't matter their level...

Well I take your point, but that's what I meant. A replacement level first baseman isn't going to have 0 dWAR. Well...unless maybe he played one game in his career or something.

A guy who had a long career spent at 1B with a 0.6 dWAR should be read as "this 1B was a historically great glove man".
   16. Blanks for Nothing, Larvell Posted: November 20, 2017 at 09:38 PM (#5579148)
A guy who had a long career spent at 1B with a 0.6 dWAR should be read as "this 1B was a historically great glove man".


But the 0.6 doesn't remotely show that.
   17. cardsfanboy Posted: November 20, 2017 at 09:41 PM (#5579150)

No, if it spits out 0.6 career for Keith Hernandez's defense, the fault is the stat.


Why? The stat isn't about him relative to other first baseman, it's about his relative defensive contribution in comparison to other positions on the diamond relative to average defenders at those position...

It's a stat that is misinterpreted by the small minded, but which most people can understand...an average first baseman over the course of the season is massively inferior defensive value than the average shortstop, to not understand that is insane.....and considering that the baseline of dWar is "average defender spread out over all positions." to think that an elite first baseman is only .6 above average in his average season, is not really that much of a stretch of the imagination.
   18. Greg K Posted: November 20, 2017 at 09:41 PM (#5579151)
No, if it spits out 0.6 career for Keith Hernandez's defense, the fault is the stat.

But that's like saying SB is a useless stat because Miguel Cabrera is at the bottom of the league list every year and we know he's a better hitter than that.

For a century before WAR baseball teams have always treated guys who field 1B as well as Keith Hernandez as around the same value as guys who play a below average 2B, if they hit the same.

I don't know....say Jeff Kent and Keith Hernandez. Both in the same ball park for WAR, both in the same ballpark for dWAR, both in the same ballpark for OPS+. I think they generally have the same reputation in terms of being great ball players. The main difference is one was a historically great 1B, the other was a below average 2B.

That's a decision teams have been making since the beginning of the sport. These guys are equals as hitters, how good does this guy have to be at first to make him more worthwhile than this mediocre 2B? dWAR is taking a stab at quantifying that age old balancing act.

EDITed to conform to the second half of the post
   19. Greg K Posted: November 20, 2017 at 09:43 PM (#5579152)
But the 0.6 doesn't remotely show that.

It does though. You're getting too caught up in the number itself rather than what it represents. If that 0.6 puts you on the short list of highest totals for 1B all time...then that means that 0.6 is saying you're an all-time great 1B.

0.6 can be a lot of things. 0.6 is a great batting average, but 0.6 is a terrible OPS.
   20. cardsfanboy Posted: November 20, 2017 at 09:44 PM (#5579153)
But the 0.6 doesn't remotely show that.


Really? It absolutely does show that, just because you don't know the stat, doesn't mean it isn't doing it's job... How many first baseman in baseball history have over a .6 dWar? I posted a list of ALL the first baseman who have more dWar over 0 and played 500 or more games at first base above...but that doesn't remotely tell the story, since it includes a ton of players who's defensive value was acculturated at other positions..


Again, your inability to grasp a simple stat, isn't the fault of the stat, it's the fault of the user.
   21. PreservedFish Posted: November 20, 2017 at 09:49 PM (#5579157)
But that's like saying SB is a useless stat because Miguel Cabrera is at the bottom of the league list every year and we know he's a better hitter than that.


Well, no, because we know that Keith Hernandez was a phenomenal defender. SB is measuring the wrong thing. dWAR is supposed to be measuring the right thing.

But the Jeff Kent / Keith Hernandez example is an interesting one.

No, if it spits out 0.6 career for Keith Hernandez's defense, the fault is the stat.

0.6 does sound awful and wrong, but if 0.6 is actually the 755 of 1B dWARs, then it's really only a superficial problem.
   22. cardsfanboy Posted: November 20, 2017 at 09:52 PM (#5579161)

0.6 does sound awful and wrong, but if 0.6 is actually the 755 of 1B dWARs, then it's really only a superficial problem.


Exactly, just because the viewer/reader was expecting a different number, doesn't invalidate the stat, there is a reason why many stats attempt to normalize the numbers to the viewers perception, but if the stat doesn't do that, it doesn't invalidate the stat.
   23. Lance Reddick! Lance him! Posted: November 20, 2017 at 09:53 PM (#5579163)
But the 0.6 doesn't remotely show that.

I find it difficult to believe you get through any day without being beaten senseless.
   24. BDC Posted: November 20, 2017 at 09:54 PM (#5579164)
Yeah, Bear, it's like you're saying that diamond can't possibly be worth anything, it's only three carats.
   25. Greg K Posted: November 20, 2017 at 09:55 PM (#5579165)
SB is measuring the wrong thing. dWAR is supposed to be measuring the right thing.

Yeah, the analogy is a bit off because at least dWAR is in the realm of measuring defence.

But I think it's a case of people assuming dWAR is measuring the thing they want measured, but it's not. I mean, I'm sure someone out there uses dWAR for something, but I know I have never looked at it outside of these threads. If I want to know how good BB-Ref thinks Keith Hernandez was at defence I look at Rfield.
   26. Greg K Posted: November 20, 2017 at 09:59 PM (#5579169)
But the Jeff Kent / Keith Hernandez example is an interesting one.

My math sucks (I'm only about 25% I actually understand what I'm saying in this thread), so I find I can only make sense of these discussions when we're dealing with actual players I can wrap my head around.
   27. cardsfanboy Posted: November 20, 2017 at 10:08 PM (#5579173)
But I think it's a case of people assuming dWAR is measuring the thing they want measured, but it's not. I mean, I'm sure someone out there uses dWAR for something, but I know I have never looked at it outside of these threads. If I want to know how good BB-Ref thinks Keith Hernandez was at defence I look at Rfield.


Exactly.... dWar has a purpose, it's actually a purpose very rarely used except outside of statistical studies or arguments. As I tried to point out, it's not a primary or a secondary argument for anything, but it has a basis in the war framework.. The issue has been that people have tried to take it and do things with it, that it wasn't ever intended for. dWar is more or less useless in ANY And ALL discussions with the mainstream or any type of public discourse.... it's entire value is in the back end discussion with people who know what the #### they are talking about.... sadly people have used it for other reasons... and it should never be used in those ways.
   28. PreservedFish Posted: November 20, 2017 at 10:25 PM (#5579178)
I will say that Keith Hernandez's defensive numbers do look surprisingly light on BR, even acknowledging that they have him as the best defensive 1B of all time. 10-13 Rfield in a year is good but not terribly impressive - Jeff Kent himself had a year of 15 Rfield, which tops Keith's best, if you can believe that. Mark Teixeira was a strong fielder but he shouldn't be within spitting distance of Keith Hernandez, who might well have been Cal Ripken had he been born right-handed.
   29. Fred Garvin is dead to Mug Posted: November 21, 2017 at 12:13 AM (#5579214)
I’m wondering if part of the confusion is due to the stat being labeled dWAR and not something else.
   30. Misirlou doesn't live in the restaurant Posted: November 21, 2017 at 12:29 AM (#5579215)
3% is just a number. 3% could be a very large number, as in 3% BAC. It could be a very small numbers, as in 3% body fat. It could be an average size number, as in current 30 year morthgage rates or corporated divident rate. But by itself, with no other context, you have no idea if 3% is large or small. Trolls like bear want to tell you 3% is meaningless, because if 3% body fat is very low, why is it illegal to drive with 3% BAC?

I’m wondering if part of the confusion is due to the stat being labeled dWAR and not something else.


It's not confusion. bear is trolling for comments like #23. He knows perfectly well what dWAR measures.
   31. Misirlou doesn't live in the restaurant Posted: November 21, 2017 at 12:36 AM (#5579218)
I will say that Keith Hernandez's defensive numbers do look surprisingly light on BR, even acknowledging that they have him as the best defensive 1B of all time. 10-13 Rfield in a year is good but not terribly impressive


2 things. Becuase we don't have PBP data, Hernandez's stats are heavily regressed. Hernandez has the 16th best season at 1B in terms of rField. Of the 15 ahead of him, only 3 were pre 1997. All the older guys will look worse because of this.

The other reason is, there's only so much a 1B can do. 90% of his job is to accept throws from professional infielders.
   32. Fred Garvin is dead to Mug Posted: November 21, 2017 at 12:55 AM (#5579221)
It's not confusion. bear is trolling for comments like #23. He knows perfectly well what dWAR measures.

Yes, but I’m not sure Pos does.
   33. Cooper Nielson Posted: November 21, 2017 at 02:03 AM (#5579228)
I don't really disagree with anything posted above (except the tone, in some places), but I would like to say:

1. Baseball-Reference today is extremely influential in terms of propagating baseball stats. Nearly every serious baseball fan refers to it regularly, and casual fans probably visit it as well. It is the authoritative source.

2. In the Player Value--Batting section, dWAR is given a rather prominent place -- it arguably draws the eye to it even more than WAR does. And because it's right next to oWAR, and symmetrically named, it's pretty natural for a user to think that these are two sides of the same coin, and that oWAR + dWAR = WAR.

3. In the past, I have, and I'll bet most of you also have, made the mistake in #2 above. I know better now, but it's not 100% obvious/clear to non-intensive users of the site.

4. So, dWAR may indeed be showing exactly what its creators intended it to show. However, the nomenclature and the presentation are clearly suboptimal to the point of being confusing. If BB-Ref doesn't want to change how the stat is calculated, they should change the name or the presentation, IMHO.

A similar problem was resolved a few years ago in the postseason stats, where the team's record in postseason series was constantly being mistaken for (and even quoted as) the individual pitcher's record until the presentation was fixed to be more consistent with the other stats. So I think Forman can perhaps take care of this "issue" with HTML rather than math or words.
   34. Morty Causa Posted: November 21, 2017 at 06:46 AM (#5579233)
The comment that Hernandez's dWAR of 0.6 is the value of an average shortstop raises a question, perhaps a stupid question, in my mind: if Ozzie Smith had been a first baseman, would he have been astronomically better than Keith? Or even simply better. If not, why not? Is it simply because that would be impossible for a player to achieve as a first baseman? Like I say, I realize this may be a dumb observation, but comparing values across positions seems to foster it--in my mind, anyway.
   35. Blanks for Nothing, Larvell Posted: November 21, 2017 at 07:12 AM (#5579237)
3% is just a number. 3% could be a very large number, as in 3% BAC.


It's not "just a number" because other players have the same number.
   36. Fancy Crazy Town Banana Pants Handle Posted: November 21, 2017 at 08:09 AM (#5579250)
The comment that Hernandez's dWAR of 0.6 is the value of an average shortstop raises a question, perhaps a stupid question, in my mind: if Ozzie Smith had been a first baseman, would he have been astronomically better than Keith? Or even simply better. If not, why not? Is it simply because that would be impossible for a player to achieve as a first baseman? Like I say, I realize this may be a dumb observation, but comparing values across positions seems to foster it--in my mind, anyway.

First of all, an average shortstop will produce something in the range of 0.6 dWAR per season. Over the course of the same amount of playing time as Hernandez's whole career, it will be much higher than that.

Secondly, if you move Ozzie, or another good SS to first, they will produce fewer dWAR than they do at short. Mostly because their value is a function (among other things) of opportunities, and shortstops have the most opportunities of anybody in the infield, and especially compared to 1B, which has the fewest.
And 1B is somewhat of an outlier infield position, because what makes a good 1B valuable is not exactly the same as what makes other infielders valuable. The ability to scoop balls, and stretch to receive throws is much more important. Range and throwing arm is less valuable.

That said, I think Ozzie at first would have had better range. (Not that Hernandez was a slouch when it comes to that.) But how much that translates in terms of total value is debatable.
   37. Greg K Posted: November 21, 2017 at 08:20 AM (#5579254)
The comment that Hernandez's dWAR of 0.6 is the value of an average shortstop raises a question, perhaps a stupid question, in my mind: if Ozzie Smith had been a first baseman, would he have been astronomically better than Keith? Or even simply better. If not, why not? Is it simply because that would be impossible for a player to achieve as a first baseman? Like I say, I realize this may be a dumb observation, but comparing values across positions seems to foster it--in my mind, anyway.

I might be the wrong person to answer, but it's my understanding that positional adjustments are about scarcity rather than skill.

In other words, it's not that Ozzie Smith has X defensive ability which will always come out as X wherever the plays. (ie. Rfield + Positional Adjustment = X) As the positional adjustment goes down from SS to 1B, the Rfield goes up by the same amount. But rather the positional adjustment is independent of the player himself. Ozzie Smith may well have been a worse 1B than Keith Hernandez, but Smith's ability to play SS allowed his teams to play someone else at 1B. Or in this case, they played on the same team for a while, so another way of looking at it is, you couldn't have swapped Smith and Hernandez at SS and 1B and had the same quality of team defence.

Which I think has two outcomes:
A] While WAR assumes values for different positions, it implicitly recognizes that players have a diverse set of skills that match them to various positions unequally. Two guys score equal Rfields at SS, aren't necessarily equally good 2B.

B] Part of WAR is out of the hands of the player. For instance, maybe you're a rookie outfielder coming up, pretty good in left field. But the team's veteran 1B has been injured, so that's where you get your chance. You're an ok 1B, so dWAR treats you as the equivalent as a shoddy LF. Not a true reflection of who you are, but thems the breaks. I suspect Ryan Goins gets a little short-changed by this. He's a great 2B (and WAR treats him as such), but he is also a fantastic SS. He's just played most of his career games at 2B because that's where the Jays have needed him. It's possible that his Rfield ratings are such that it evens out, but he hasn't been playing at his optimal position for most of his career, which I think (more often than not) shaves off some potential WAR you could have achieved.
   38. manchestermets Posted: November 21, 2017 at 09:01 AM (#5579264)
2. In the Player Value--Batting section, dWAR is given a rather prominent place -- it arguably draws the eye to it even more than WAR does. And because it's right next to oWAR, and symmetrically named, it's pretty natural for a user to think that these are two sides of the same coin, and that oWAR + dWAR = WAR.

3. In the past, I have, and I'll bet most of you also have, made the mistake in #2 above. I know better now, but it's not 100% obvious/clear to non-intensive users of the site.


I have, and I still don't entirely understand why it's not the case. I think I'm correct in saying that it's because oWAR is relative to players at the same position, and dWAR is relative to all players, right? But then I get slightly confused as to what exactly WAR itself is relative to - is that all players, or players at the same position? It's definitely counterintuitive that there is a stat composed of offensive and defensive values, but the exact values that combine to make that stat don't seem to be visible anywhere.
   39. snapper (history's 42nd greatest monster) Posted: November 21, 2017 at 09:08 AM (#5579268)
I have, and I still don't entirely understand why it's not the case. I think I'm correct in saying that it's because oWAR is relative to players at the same position, and dWAR is relative to all players, right? But then I get slightly confused as to what exactly WAR itself is relative to - is that all players, or players at the same position?

All three are relative to all players. Both oWAR and dWAR have the position adjustment. That's why you can't simply add them. You have to substract the duplicate positional adjustment.
   40. Greg K Posted: November 21, 2017 at 09:45 AM (#5579278)
Yeah, the way I think of it is...

The point of oWAR is to get everything in the WAR package except for how good a fielder the player is. It's useful if you have a metric you think is better than Rfield, or if you want to treat all players as effectively average at their positions.

dWAR is useful if you want to know what WAR thinks of the player's entire contribution on defence (both how good they are at the position, and positional adjustment).

It's pretty straightforward when you hover over the stat and get the glossary, but I think for anyone just taking a glance at a B-Ref page they are misleadingly named.
   41. Misirlou doesn't live in the restaurant Posted: November 21, 2017 at 09:48 AM (#5579280)
It's not "just a number" because other players have the same number.


Right. One person has 3% BAC (and is dead), the other has 3% body fat. Therefore they are the same? 0.6 dWAR is outstanding for a long term firstbaseman. It is lousy for a long tern SS. It's average for a long term CF.

Part of WAR is out of the hands of the player. For instance, maybe you're a rookie outfielder coming up, pretty good in left field. But the team's veteran 1B has been injured, so that's where you get your chance. You're an ok 1B, so dWAR treats you as the equivalent as a shoddy LF. Not a true reflection of who you are, but thems the breaks.


Or, handedness in the case of Hernandez. Had he been right handed, he might have been a GG 3B, or even a passable SS.
   42. snapper (history's 42nd greatest monster) Posted: November 21, 2017 at 10:05 AM (#5579292)
The point of oWAR is to get everything in the WAR package except for how good a fielder the player is. It's useful if you have a metric you think is better than Rfield, or if you want to treat all players as effectively average at their positions.

I think the point of WAR is that it's a framework for a set of replaceable stats, rather than being a stat itself.

Each of the factors is additive and separable, so you can substitute your own preferred version. If you like UZR better than DRS, sub it in. If you want to use a different positional adjustment you can.
   43. Greg K Posted: November 21, 2017 at 10:05 AM (#5579293)
How good would a lefty have to be, to play a decent 3B? I suppose, if Keith Hernandez wasn't good enough, then it's an academic question.
   44. RMc's Unenviable Situation Posted: November 21, 2017 at 07:38 PM (#5579857)
People might not remember but I created one of the first WAR

Hmph! Back in my day, ol' Specs would get out his abacus and figger out who was better than whom! And the coach would smash the abacus over his head, saying, "There ain't no numbers in baseball, ya poindexter!"
   45. Rally Posted: November 21, 2017 at 08:14 PM (#5579868)
How good would a lefty have to be, to play a decent 3B? I suppose, if Keith Hernandez wasn't good enough, then it's an academic question.


Hernandez never got a chance to play there, so we don't know how good he could have been. I guess there was no reason to try it since he was so great at first and also hit more than enough (despite lack of power) to be great at the position.

Don Mattingly did fine at 3B in a super small sample (18 innings, 11 assists, 1 error, 2 DP, +1 TZ). Mattingly was a very good defender at first, but Hernandez was a lot better, and particularly had a better arm. My null hypothesis, if someone had given him a chance, would have been that Hernandez would have been able to play a pretty good 3B despite being left handed.
   46. PreservedFish Posted: November 21, 2017 at 08:26 PM (#5579869)
I think that "how would Ozzie do at first" is a great question. I wish I had some insight into the answer. His range would have been mind-blowing.

What happens to the left-handed Ozzie Smiths? I mean, we know what happens to them - they play centerfield - but it's a bit sad to reflect on the number of players with superlative infield skills that have never played the infield at all. For all we know someone like Brett Gardner might have made a boffo shortstop in another universe.
   47. Mike Emeigh Posted: November 21, 2017 at 09:35 PM (#5579883)
I think the point of WAR is that it's a framework for a set of replaceable stats, rather than being a stat itself.


WAR is a property appraisal. The WAR value represents what the player's performance is "theoretically" worth based on the factors that give players value in the first place, and how much those factors contribute to the overall value to the team.

A property appraisal doesn't mean that you can either buy a property or sell a property for that amount. It gives you an idea of what you can "expect" if you try to buy or sell the property, but you will almost never be able to buy it or sell it for exactly that amount.

-- MWE
   48. Morty Causa Posted: November 21, 2017 at 09:58 PM (#5579890)
Oops, I didn't have this thread bookmarked (don't bookmark much). But, I did want to acknowledge the studious replies to my post 34. They've helped clarify some of my questions and qualms. Thank'ee.
   49. Meatwad Posted: November 21, 2017 at 11:34 PM (#5579918)
I was told there was a wilco discussion going on here. Lies.
   50. Steve Parris, Je t'aime Posted: November 27, 2017 at 10:14 AM (#5581368)
I was told there was a wilco discussion going on here. Lies.


This was premature - we're all waiting for the release of the Being There deluxe edition.
   51. bookbook Posted: November 27, 2017 at 12:55 PM (#5581482)
How much career value did Bagwell lose playing 1b rather than 3b? Or did he get too big to maintain his range? He was capable of stealing ten bases a year or so for much of his career...
   52. Rally Posted: November 27, 2017 at 01:59 PM (#5581530)
Was Bagwell's shoulder hurt in a specific incident late in his career that prevented him from playing? My recollection is that while it wasn't as bad as it got at the end, it was a chronic issue. He probably would have had to move to first at some point in mid career.
   53. Rally Posted: November 27, 2017 at 02:06 PM (#5581536)
How much career value did Bagwell lose playing 1b rather than 3b? Or did he get too big to maintain his range? He was capable of stealing ten bases a year or so for much of his career...


Here are 3 scenarios. As is, Bagwell had 80 WAR

1. 94 WAR. Bagwell plays 3B for his whole career. He is above average there, saving just as many runs at third as he did at first

2. 89 WAR. Bagwell plays 3B and gets the better position adjustment, but is a league average defender at third.

3. 84 WAR. Bagwell plays 3B through age 30, is an average defender, and shifts to first at age 31.

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
Ray (RDP)
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Hot Topics

NewsblogOTP 11 December, 2017 - GOP strategist: Moore would have 'date with a baseball bat' if he tried dating teens where I grew up
(317 - 6:00pm, Dec 11)
Last: Count

NewsblogOT - NBA 2017-2018 Tip-off Thread
(1904 - 5:53pm, Dec 11)
Last: Fourth True Outcome

Gonfalon CubsLooking to next year
(315 - 5:47pm, Dec 11)
Last: Moses Taylor, aka Hambone Fakenameington

NewsblogJack Morris, Alan Trammell elected to Hall | MLB.com
(115 - 5:44pm, Dec 11)
Last: RMc's Unenviable Situation

NewsblogTexas Rangers: Shohei Ohtani or not, Rangers still considering nontraditional 6-man, 4-man rotations next season | SportsDay
(9 - 5:29pm, Dec 11)
Last: 6 - 4 - 3

NewsblogRyan Thibs has his HOF Ballot Tracker Up and Running!
(339 - 5:22pm, Dec 11)
Last: The Yankee Clapper

Sox TherapyA Container of Milk, A Loaf of Bread and a Dude Who Can Hit Home Runs
(9 - 5:17pm, Dec 11)
Last: PreservedFish

NewsblogOT: Winter Soccer Thread
(313 - 5:03pm, Dec 11)
Last: SPICEY WITH A SIDE OF BEER ON A BABYYYYYYY

NewsblogPrimer Dugout (and link of the day) 12-11-2017
(19 - 4:49pm, Dec 11)
Last: TDF didn't lie, he just didn't remember

NewsblogMets have talked to teams about trading Matt Harvey: sources - NY Daily News
(6 - 4:21pm, Dec 11)
Last: Walt Davis

Hall of Merit2018 Hall of Merit Ballot Discussion
(324 - 3:41pm, Dec 11)
Last: Kiko Sakata

Hall of MeritBattle of the Uber-Stat Systems (Win Shares vs. WARP)!
(375 - 3:36pm, Dec 11)
Last: Carl Goetz

NewsblogPeter Gammons: Splashes and notes from the Orlando Winter Meetings - GammonsDaily.com
(11 - 3:22pm, Dec 11)
Last: Batman

NewsblogTrade secrets: The key to covering baseball’s winter meetings is an adjoining door – The Athletic
(3 - 3:12pm, Dec 11)
Last: Tulo's Fishy Mullet (mrams)

Hall of Merit2018 Hall of Merit Ballot
(19 - 2:22pm, Dec 11)
Last: ronw

Page rendered in 0.3841 seconds
47 querie(s) executed