WAR - A Primer (1 Viewer)

[oso diablo]

In this year's AL MVP debate, many supporters of Miguel Cabrera's candidacy have taken to bashing the WAR statistic. I could be uncharitable and give my armchair psychiatric view on that, but instead let's just take a look at what WAR is.

Here are two links that attempt to describe WAR in plain english.

Link 1 - with a look at Alan Trammell as one example

Link 2

Put in simple terms, WAR attempts to measure all aspects of a player's performance - batter, baserunning, and defense (including pitching), comparing that performance to the level of a "replacement level" player. That is, the sort of player a typical team might turn to, the baseline of value.

Now, at the end of the day, i don't really care whether Cabrera wins the MVP, but it would be a mistake if we ignored or dismissed WAR just because it doesn't support our favored conclusion. It is an extremely helpful measurement. Teams that want to win baseball games will pay attention to metrics like these, because they do the best job (at least among those tools currently available) at predicting what leads to winning baseball.

It's not voodoo just because it's complex.

 

[Doctor Detroit]

Jhonny Peralta is the second best fielding shortstop in the American League according to UZR; better than Elvis Andrus, Alcides Escobar (who has a brutal UZR), and his UZR/150 is better than JJ Hardy. Those who watch baseball understand that Johonny Peralta is no JJ Hardy. Now UZR gets it right with guys like Derek Jeter but it is not a perfect indicator by any means.

You want to present WAR as something that can be demonstrated in "plain English" without really getting into specifics. If you want to defend it break down the UBR component and explain what variables are and why you defend it, I'd love to hear it. I liked the Trammell explanation but again we don't know who is fielding those balls, where they are fielding them and what the conditions are. While the component does its best its still flawed to a great degree don't you think?

I like WAR for the most part but there are too many people claiming it is some sort of statistical anchor that actually does exactly as it promises (that is my bottom line issue with it and I've always said this about this particular measure). Evan Longoria's WAR was 3.2 this year yet the Rays were 47-27 with him, and 43-45 without him. Now there may be other factors at play with the Rays and some of that might be coincidence, but Tampa is a 102 win team with him and a 90 win team without him. A projected 6.5 WAR doesn't cover the difference.

I guess I take issue with people quoting the Angels record with Trout (leaving out Pujols' stats while doing it), but not considering what the Tigers record would be without Cabrera (because it's unknown). Well if you watch the Tigers every day, you'd know they are a 75 win team without Cabrera. The Angels without Mike Trout the whole season? Still in the 80s IMO. WAR isn't voodoo but I contend that a statement like that is suggesting that those who support it and get behind it really know what it is. I will tell you that most of them have no idea, and you know it to be true.

 

[oso diablo]

I guess I take issue with people quoting the Angels record with Trout (leaving out Pujols' stats while doing it), but not considering what the Tigers record would be without Cabrera (because it's unknown). Well if you watch the Tigers every day, you'd know they are a 75 win team without Cabrera. The Angels without Mike Trout the whole season? Still in the 80s IMO.

I agree that a lot of people don't know what WAR is; that's why i posted the topic.As for the above, this really has nothing to do with WAR. It's more of a faith-based statement. Though even with that, you say that the Tigers sans Cabrera is an unknown, and then turn around in the next sentence and say that it in indeed a known thing. Not really impressed with these guesses, no matter who does them. They're not really helpful to a meaningful discussion, because they are purely opinions.

 

[Doctor Detroit]

I guess I take issue with people quoting the Angels record with Trout (leaving out Pujols' stats while doing it), but not considering what the Tigers record would be without Cabrera (because it's unknown). Well if you watch the Tigers every day, you'd know they are a 75 win team without Cabrera. The Angels without Mike Trout the whole season? Still in the 80s IMO.

I agree that a lot of people don't know what WAR is; that's why i posted the topic.As for the above, this really has nothing to do with WAR. It's more of a faith-based statement. Though even with that, you say that the Tigers sans Cabrera is an unknown, and then turn around in the next sentence and say that it in indeed a known thing. Not really impressed with these guesses, no matter who does them. They're not really helpful to a meaningful discussion, because they are purely opinions.

Ok that's fine. But do you not admit WAR does have a lot of guessing in the formulas? Maybe "assumptions" is a better word actually. This is generally my issue with sabermetrics. I enjoy them, I enjoy getting to know them but I don't enjoy arguing with people who treat them as the final verdict, the decision to guideuis to the truth. We are getting folks that are poo-pooing raw numbers or not using the most important statistics (OPS, SLG, WHIP, K/9, K/BB) to create their opinions. I'm just merely suggesting that we continue to look at all the ingredients when making the soup, not just to rely on one or two set sabers that leave questions to be answered. To me you use these stats to bolster an argument, not to be the central point of one.

 

[Smack Tripper]

The meat and potatoes valuations of WAR are opinion based, and mechanically as out of context as traditional stats are.

A routine grounder in cincy may be a termed an extraordinary range play in Pittsburgh. It's left up to, well, who? The point is, the same person is not watching all these games as to establish a baseline control sample.

Even elements, like baserunning, which is hard to quantify, but tries to based on things like how often a player takes an extra base. Which, to my knowledge, doesn't account for the fact that if you have a Napoli or a prince fielder on base in front of you, the next base may be occupied.

All of that said, I like and respect what it tries to do. It's another dimension to use with traditional stats but I don't see what about it improves traditional stas in any regard as to render them irrelevant.

 

[boubucarow]

I guess I take issue with people quoting the Angels record with Trout (leaving out Pujols' stats while doing it), but not considering what the Tigers record would be without Cabrera (because it's unknown). Well if you watch the Tigers every day, you'd know they are a 75 win team without Cabrera. The Angels without Mike Trout the whole season? Still in the 80s IMO.

I agree that a lot of people don't know what WAR is; that's why i posted the topic.As for the above, this really has nothing to do with WAR. It's more of a faith-based statement. Though even with that, you say that the Tigers sans Cabrera is an unknown, and then turn around in the next sentence and say that it in indeed a known thing. Not really impressed with these guesses, no matter who does them. They're not really helpful to a meaningful discussion, because they are purely opinions.

Ok that's fine. But do you not admit WAR does have a lot of guessing in the formulas? Maybe "assumptions" is a better word actually. This is generally my issue with sabermetrics. I enjoy them, I enjoy getting to know them but I don't enjoy arguing with people who treat them as the final verdict, the decision to guideuis to the truth. We are getting folks that are poo-pooing raw numbers or not using the most important statistics (OPS, SLG, WHIP, K/9, K/BB) to create their opinions. I'm just merely suggesting that we continue to look at all the ingredients when making the soup, not just to rely on one or two set sabers that leave questions to be answered. To me you use these stats to bolster an argument, not to be the central point of one.

WAR does have a lot of assumptions as the value of base running and defense will never be certain. However, your some of your important statistics aren't the most important.SLG, K/9, and K/BB are pretty clear. OPS on the other hand is an over simplified way to address a player's offensive contribution. OBP has been shown to be more important than SLG so adding them together is flawed. wOBA is the best statistic we have since it includes yearly determined rates on the various offensive statistics to gauge a players value in the current run scoring environment. It even accounts for SB and CS in addition to the at the plate numbers. So either use OBP and SLG (and AVG still has its value since a single is better than a walk) or use wOBA with OBP and SLG on the side to determine where a hitter is best suited in a lineup.As for WHIP, the defense behind the pitcher has way too much impact for it to be pointed to as one of the most important statistics.

 

[rodg12]

The meat and potatoes valuations of WAR are opinion based, and mechanically as out of context as traditional stats are.

A routine grounder in cincy may be a termed an extraordinary range play in Pittsburgh. It's left up to, well, who? The point is, the same person is not watching all these games as to establish a baseline control sample.

Even elements, like baserunning, which is hard to quantify, but tries to based on things like how often a player takes an extra base. Which, to my knowledge, doesn't account for the fact that if you have a Napoli or a prince fielder on base in front of you, the next base may be occupied.

All of that said, I like and respect what it tries to do. It's another dimension to use with traditional stats but I don't see what about it improves traditional stas in any regard as to render them irrelevant.

 

[bicycle_seat_sniffer]

sounds made up

 

[TobiasFunke]

The meat and potatoes valuations of WAR are opinion based, and mechanically as out of context as traditional stats are.

A routine grounder in cincy may be a termed an extraordinary range play in Pittsburgh. It's left up to, well, who? The point is, the same person is not watching all these games as to establish a baseline control sample.

Even elements, like baserunning, which is hard to quantify, but tries to based on things like how often a player takes an extra base. Which, to my knowledge, doesn't account for the fact that if you have a Napoli or a prince fielder on base in front of you, the next base may be occupied.

All of that said, I like and respect what it tries to do. It's another dimension to use with traditional stats but I don't see what about it improves traditional stas in any regard as to render them irrelevant.

What you say about the faults of WAR is true- and is why I try not to use it as a stand-alone in any conversation, or cite both Fangraphs and Baseball Reference when I do. But isn't ANY effort to incorporate defense and baserunning better than ignoring them completely or discussing them independent of one another? So The idea of WAR is to synthesize all of those things into a single number. Obviously it's not perfect, no single stat is or ever will be. You don't throw out the baby with the bathwater. Don't let the perfect be the enemy of the good. Insert whatever trite saying you find applicable here.

What single "traditional" stat would you say tells you more about a non-pitcher's value than WAR? Heck, what five traditional stats combined? OBP, slugging ... what else?

 

[rodg12]

The meat and potatoes valuations of WAR are opinion based, and mechanically as out of context as traditional stats are.

A routine grounder in cincy may be a termed an extraordinary range play in Pittsburgh. It's left up to, well, who? The point is, the same person is not watching all these games as to establish a baseline control sample.

Even elements, like baserunning, which is hard to quantify, but tries to based on things like how often a player takes an extra base. Which, to my knowledge, doesn't account for the fact that if you have a Napoli or a prince fielder on base in front of you, the next base may be occupied.

All of that said, I like and respect what it tries to do. It's another dimension to use with traditional stats but I don't see what about it improves traditional stas in any regard as to render them irrelevant.

What you say about the faults of WAR is true- and is why I try not to use it as a stand-alone in any conversation, or cite both Fangraphs and Baseball Reference when I do. But isn't ANY effort to incorporate defense and baserunning better than ignoring them completely or discussing them independent of one another? So The idea of WAR is to synthesize all of those things into a single number. Obviously it's not perfect, no single stat is or ever will be. You don't throw out the baby with the bathwater. Don't let the perfect be the enemy of the good. Insert whatever trite saying you find applicable here.

What single "traditional" stat would you say tells you more about a non-pitcher's value than WAR? Heck, what five traditional stats combined? OBP, slugging ... what else?

No. Not when that method has inherent flaws. What difference does it make if you try to combine them if the end result can't be trusted and therefore provides no value?

 

[TobiasFunke]

The meat and potatoes valuations of WAR are opinion based, and mechanically as out of context as traditional stats are.

A routine grounder in cincy may be a termed an extraordinary range play in Pittsburgh. It's left up to, well, who? The point is, the same person is not watching all these games as to establish a baseline control sample.

Even elements, like baserunning, which is hard to quantify, but tries to based on things like how often a player takes an extra base. Which, to my knowledge, doesn't account for the fact that if you have a Napoli or a prince fielder on base in front of you, the next base may be occupied.

All of that said, I like and respect what it tries to do. It's another dimension to use with traditional stats but I don't see what about it improves traditional stas in any regard as to render them irrelevant.

What you say about the faults of WAR is true- and is why I try not to use it as a stand-alone in any conversation, or cite both Fangraphs and Baseball Reference when I do. But isn't ANY effort to incorporate defense and baserunning better than ignoring them completely or discussing them independent of one another? So The idea of WAR is to synthesize all of those things into a single number. Obviously it's not perfect, no single stat is or ever will be. You don't throw out the baby with the bathwater. Don't let the perfect be the enemy of the good. Insert whatever trite saying you find applicable here.

What single "traditional" stat would you say tells you more about a non-pitcher's value than WAR? Heck, what five traditional stats combined? OBP, slugging ... what else?

No. Not when that method has inherent flaws. What difference does it make if you try to combine them if the end result can't be trusted and therefore provides no value?

I don't understand this argument at all. EVERY stat is flawed. Should we disregard OBP because it doesn't account for power? Should we disregard HR totals because they don't account for park effects? Should we disregard Ks because it doesn't account for variances in umpiring? I'm trying to grasp what you're saying here. Is the argument that the flaw is so deep and pervasive that the defense/baserunning aspects of WAR have no value whatsoever? For that to be true there would have to be zero correlation between the defense/baserunning aspects of WAR and the player's true performance at those things. Is that what you're saying? That if you had to choose a player to play defense or run the bases, you'd just as soon throw darts at a list as base it on defense and baserunning WAR?

 

[Eephus]

One season is still a very small sample size for any statistic, no matter how well established or contrived it is. The problem with WAR IMHO is how some people use it at the exclusion of everything else. It's a fun single stat to throw around because it theoretically measures everything. But if every stat includes some degree of imprecision, combining a whole bunch of stats into a single metric increases the variability of that number. If B-R and Fangraphs can't agree on how to compute it, it reduces WAR credibility a bit.

As always, fielding stats are a big part of the problem. I think everybody agrees fielding is an important part of the game but damn it's tough to measure.

 

[jon_mx]

Jhonny Peralta is the second best fielding shortstop in the American League according to UZR; better than Elvis Andrus, Alcides Escobar (who has a brutal UZR), and his UZR/150 is better than JJ Hardy. Those who watch baseball understand that Johonny Peralta is no JJ Hardy. Now UZR gets it right with guys like Derek Jeter but it is not a perfect indicator by any means.

You look at Hardy numbers and they are just incredible. He has 529 assists compared to Peralta having 359. Hardy gets 1 plus assist more than Peralta per game! That is an huge advantage. Any metric which fails to recognize that is fatally flawed. Peralta is one of the best at not making errors, but he just does not lunge for balls or dive for balls. He is great at making the easy plays. But give me any of the other half dozen guys who get to 50 more balls than Peralta who might happen to make a few more errors.

 

[cheese]

One season is still a very small sample size for any statistic, no matter how well established or contrived it is. The problem with WAR IMHO is how some people use it at the exclusion of everything else. It's a fun single stat to throw around because it theoretically measures everything. But if every stat includes some degree of imprecision, combining a whole bunch of stats into a single metric increases the variability of that number. If B-R and Fangraphs can't agree on how to compute it, it reduces WAR credibility a bit.

As always, fielding stats are a big part of the problem. I think everybody agrees fielding is an important part of the game but damn it's tough to measure.

Fielding is a big hangup for me with embracing stats like WAR and maybe even advanced stats in general. It hurts the credibility of the offensive portion which I honestly believe in otherwise.If I say I rank these 3 offensive players 1. A 2. B 3. C and the players are close, I will accept it when a stat like OPS suggests I've ordered them wrong. In fact, it makes me believe the stats probably know better than I do. However, if the stats tell me that Edgar Martinez is a better defensive 3rd baseman than Brooks Robinson, I'm not interested in stats anymore. At that point, they're insulting the intelligence of the common baseball fan.

 

[Dickie Dunn]

I have never understood the purpose of simply adding OBP and SLG together to get OPS. Never mind the fact that almost every factor of OBP (aside from walks, HBP and the like) already is accounted for in one's SLG.

Why not a hybrid OBP/SLG number ... which factors in total bases per plate appearance (much like SLG, but also including BB, HBP, etc.)?

For all of the complexities of some sabermetrics, the crudeness of OPS always has baffled me.

 

[Doctor Detroit]

Now we're rolling.

 

[rodg12]

The meat and potatoes valuations of WAR are opinion based, and mechanically as out of context as traditional stats are.

A routine grounder in cincy may be a termed an extraordinary range play in Pittsburgh. It's left up to, well, who? The point is, the same person is not watching all these games as to establish a baseline control sample.

Even elements, like baserunning, which is hard to quantify, but tries to based on things like how often a player takes an extra base. Which, to my knowledge, doesn't account for the fact that if you have a Napoli or a prince fielder on base in front of you, the next base may be occupied.

All of that said, I like and respect what it tries to do. It's another dimension to use with traditional stats but I don't see what about it improves traditional stas in any regard as to render them irrelevant.

What you say about the faults of WAR is true- and is why I try not to use it as a stand-alone in any conversation, or cite both Fangraphs and Baseball Reference when I do. But isn't ANY effort to incorporate defense and baserunning better than ignoring them completely or discussing them independent of one another? So The idea of WAR is to synthesize all of those things into a single number. Obviously it's not perfect, no single stat is or ever will be. You don't throw out the baby with the bathwater. Don't let the perfect be the enemy of the good. Insert whatever trite saying you find applicable here.

What single "traditional" stat would you say tells you more about a non-pitcher's value than WAR? Heck, what five traditional stats combined? OBP, slugging ... what else?

No. Not when that method has inherent flaws. What difference does it make if you try to combine them if the end result can't be trusted and therefore provides no value?

I don't understand this argument at all. EVERY stat is flawed. Should we disregard OBP because it doesn't account for power? Should we disregard HR totals because they don't account for park effects? Should we disregard Ks because it doesn't account for variances in umpiring?

This isn't like every stat, though. This is supposed to be an end-all summation of a player's value. One stop shop to see who's the best. That's the reason it was constructed. For that sole purpose. Nobody says...Player X has the highest OBP in baseball, he's the best player in the game. Nobody says Player X has the most HR in baseball, he's the best player in the game. Tons of people are now saying Player X has the highest WAR in baseball, he's the best player in the game. If you can't trust that it's accurate in defining baserunning/defensive value...I guess it can still be used, but not close to how it's intended to be or is in most cases. And, at that point, what value does it add to the conversation?

 

[Premier]

Evan Longoria's WAR was 3.2 this year yet the Rays were 47-27 with him, and 43-45 without him. Now there may be other factors at play with the Rays and some of that might be coincidence, but Tampa is a 102 win team with him and a 90 win team without him. A projected 6.5 WAR doesn't cover the difference. .

This is also my biggest problem with it. Also, I think I'm going to go cry again.

 

[oso diablo]

Defensive WAR metrics for qualifying AL SS in 2012:

1. Brendan Ryan - 3.6

2. JJ Hardy - 2.8

3. Yunel Escobar - 2.3

4. Alexei Ramirez - 2.3

5. Mike Aviles - 2.0

6. Cliff Pennington - 1.9

7. Elvis Andrus - 1.6

8. Erick Aybar - 1.0

9. Jhonny Peralta - 0.7

10. Alcides Escobar - 0.6

11. Asdrubal Cabrera - 0.1

12. Derek Jeter - (1.6)

 

[Spartans Rule]

I have never understood the purpose of simply adding OBP and SLG together to get OPS. Never mind the fact that almost every factor of OBP (aside from walks, HBP and the like) already is accounted for in one's SLG.Why not a hybrid OBP/SLG number ... which factors in total bases per plate appearance (much like SLG, but also including BB, HBP, etc.)?For all of the complexities of some sabermetrics, the crudeness of OPS always has baffled me.

OPS caught on because it's simple. Take two basic, readily-available stats and add them together. It's far from perfect, but it was a big improvement over batting average and home runs being held up as the holy grail for measuring offensive production.wOBA is a much better stat but it's also harder to calculate.

 

[jon_mx]

Defensive stats in general stink. I really don't want to know about assists and putouts. For a first basemen an assist could just mean he is too slow to get to first base and throws to the pitcher. I want to know how many ground balls a shortstop gets to, and how many line drives and pop-ups he catches. Those are the numbers I want to know to evaluate range. Assists and putouts really don't tell a good story. A shortstop number of put outs is more dependant upon how many base runners the team allows and how good his second basemen is.

 

[jon_mx]

Defensive WAR metrics for qualifying AL SS in 2012:1. Brendan Ryan - 3.62. JJ Hardy - 2.83. Yunel Escobar - 2.34. Alexei Ramirez - 2.35. Mike Aviles - 2.06. Cliff Pennington - 1.97. Elvis Andrus - 1.68. Erick Aybar - 1.09. Jhonny Peralta - 0.710. Alcides Escobar - 0.611. Asdrubal Cabrera - 0.112. Derek Jeter - (1.6)

That is not bad, I just don't see how any metric could put Ryan over Hardy. Hardy has a huge advantage over Ryan on assists and makes fewer errors to boot. That part is just crazy, IMHO. Jeter at 12 and Peralta at 9 are about right.

 

[oso diablo]

Defensive stats in general stink. I really don't want to know about assists and putouts. For a first basemen an assist could just mean he is too slow to get to first base and throws to the pitcher. I want to know how many ground balls a shortstop gets to, and how many line drives and pop-ups he catches. Those are the numbers I want to know to evaluate range. Assists and putouts really don't tell a good story. A shortstop number of put outs is more dependant upon how many base runners the team allows and how good his second basemen is.

you are far from the first person to have these inklings and wishes. there are some really smart people trying to figure out this stuff.

 

[jon_mx]

Defensive stats in general stink. I really don't want to know about assists and putouts. For a first basemen an assist could just mean he is too slow to get to first base and throws to the pitcher. I want to know how many ground balls a shortstop gets to, and how many line drives and pop-ups he catches. Those are the numbers I want to know to evaluate range. Assists and putouts really don't tell a good story. A shortstop number of put outs is more dependant upon how many base runners the team allows and how good his second basemen is.

you are far from the first person to have these inklings and wishes. there are some really smart people trying to figure out this stuff.

This isn't rocket science. I could produce better metrics than what are being used, at least the public ones which seem rather simplistic and not so valuable.

 

[Doctor Detroit]

Defensive WAR metrics for qualifying AL SS in 2012:

1. Brendan Ryan - 3.6

2. JJ Hardy - 2.8

3. Yunel Escobar - 2.3

4. Alexei Ramirez - 2.3

5. Mike Aviles - 2.0

6. Cliff Pennington - 1.9

7. Elvis Andrus - 1.6

8. Erick Aybar - 1.0

9. Jhonny Peralta - 0.7

10. Alcides Escobar - 0.6

11. Asdrubal Cabrera - 0.1

12. Derek Jeter - (1.6)

Jeter gold glove?

Offensive players – Take wRAA, UBR, and UZR (which express offensive, base running, and defensive value in runs above average) and add them together. Add in a positional adjustment, since some positions are tougher to play than others, and then convert the numbers so that they’re not based on league average, but on replacement level (which is the value a team would lose if they had to replace that player with a “replacement” player – a minor leaguer or someone from the waiver wire). Convert the run value to wins (10 runs = 1 win) and voila, finished!

Defensive Ability UZR

Gold Glove Caliber +15

Great +10

Above Average +5

Average 0

Below Average -5

Poor -10

Awful -15

Name	Team	Pos	Inn	rSB	    rGDP	rARM	rGFP	rPM	DRS	BIZ	Plays	RZR	OOZ	CPP	RPP	TZL	FSR	ARM	DPR	RngR	ErrR	UZR	UZR/150Brendan Ryan	SEA	SS	1170.2	 	4	 	2	21	27	309	252	.816	87	 	 	 	 	 	3.8	10.5	1.3	15.5	18.2Clint Barmes	PIT	SS	1159.0	 	1	 	1	11	13	368	306	.832	61	 	 	 	 	 	-0.1	14.1	1.0	15.0	16.5J.J. Hardy	BAL	SS	1439.0	 	0	 	4	14	18	420	350	.833	109	 	 	 	 	 	0.1	3.5	7.9	11.4	10.3Jhonny Peralta	DET	SS	1277.2	 	0	 	-1	0	-1	307	262	.853	49	 	 	 	 	 	0.1	5.5	4.3	9.9	12.0Zack Cozart	CIN	SS	1163.2	 	0	 	0	12	12	313	258	.824	66	 	 	 	 	 	0.1	6.8	2.7	9.5	11.5Elvis Andrus	TEX	SS	1333.0	 	1	 	-2	9	8	349	283	.811	83	 	 	 	 	 	0.8	7.2	0.4	8.4	8.9BrandoCrawford	SFG	SS	1101.0	 	1	 	1	10	12	365	303	.830	49	 	 	 	 	 	2.1	6.7	-1.0	7.9	9.0Mike Aviles	BOS	SS	1107.2	 	2	 	-2	14	14	292	245	.839	71	 	 	 	 	 	2.2	4.7	-0.9	6.0	7.1Alexei Ramirez	CHW	SS	1392.0	 	2	 	0	12	14	369	309	.837	77	 	 	 	 	 	1.2	1.7	3.1	6.0	5.9Yunel Escobar	TOR	SS	1250.2	 	0	 	-2	17	15	369	312	.846	72	 	 	 	 	 	-1.9	4.4	3.3	5.8	6.5Jimmy Rollins	PHI	SS	1364.0	 	-1	 	3	-10	-8	352	287	.815	48	 	 	 	 	 	0.7	1.6	3.3	5.5	6.2Ian Desmond	WSN	SS	1139.1	 	1	 	0	-7	-6	269	230	.855	40	 	 	 	 	 	0.5	3.3	-0.3	3.4	4.9Ruben Tejada	NYM	SS	964.2	 	0	 	3	-3	0	253	210	.830	33	 	 	 	 	 	0.0	0.3	1.1	1.3	2.1

fangraphs

 

[Smack Tripper]

The meat and potatoes valuations of WAR are opinion based, and mechanically as out of context as traditional stats are.

A routine grounder in cincy may be a termed an extraordinary range play in Pittsburgh. It's left up to, well, who? The point is, the same person is not watching all these games as to establish a baseline control sample.

Even elements, like baserunning, which is hard to quantify, but tries to based on things like how often a player takes an extra base. Which, to my knowledge, doesn't account for the fact that if you have a Napoli or a prince fielder on base in front of you, the next base may be occupied.

All of that said, I like and respect what it tries to do. It's another dimension to use with traditional stats but I don't see what about it improves traditional stas in any regard as to render them irrelevant.

What you say about the faults of WAR is true- and is why I try not to use it as a stand-alone in any conversation, or cite both Fangraphs and Baseball Reference when I do. But isn't ANY effort to incorporate defense and baserunning better than ignoring them completely or discussing them independent of one another? So The idea of WAR is to synthesize all of those things into a single number. Obviously it's not perfect, no single stat is or ever will be. You don't throw out the baby with the bathwater. Don't let the perfect be the enemy of the good. Insert whatever trite saying you find applicable here.

What single "traditional" stat would you say tells you more about a non-pitcher's value than WAR? Heck, what five traditional stats combined? OBP, slugging ... what else?

I don't think there is a single number that tells you this, any more than seeing a guy play one game tells you what kind of player someone is. They all work together and tell a story in total.WAR tells us that a five tool player is going to dial up a good WAR number. This isn't exactly something new we've learned in baseball. Players that can suceed and do lots of things are generally better than one dimensional guys. UNLESS, they are outstanding in one category. Strike me dead, but over 500, 1000, 2500 abs, I think batting average still tells you quite a bit. Putting the bat on the ball, with the general conceit, the higher average guy will typically draw walks due to his overall hitting, is something of value to me. And yes, I place a higher value on it than OBA because with two outs and a man on second, a walk is not as good as a hit.

But thats not to say anything or anyone should be imbalanced, but as a 25 year baseball fan, I didn't need WAR to tell me that a guy who can work the count is going to generally get a better pitch to hit, and drive, than a guy who flails at everything and is in 0-2, 1-2 protect mode constantly.

No one number puts a bow on anything however. Including WAR, despite its valiant efforts. Ben Zobrist is proof of that last year.

 

[Please See Mine]

Defensive stats in general stink. I really don't want to know about assists and putouts. For a first basemen an assist could just mean he is too slow to get to first base and throws to the pitcher. I want to know how many ground balls a shortstop gets to, and how many line drives and pop-ups he catches. Those are the numbers I want to know to evaluate range. Assists and putouts really don't tell a good story. A shortstop number of put outs is more dependant upon how many base runners the team allows and how good his second basemen is.

you are far from the first person to have these inklings and wishes. there are some really smart people trying to figure out this stuff.

This isn't rocket science. I could produce better metrics than what are being used, at least the public ones which seem rather simplistic and not so valuable.

 

[Eephus]

Defensive WAR metrics for qualifying AL SS in 2012:1. Brendan Ryan - 3.62. JJ Hardy - 2.83. Yunel Escobar - 2.34. Alexei Ramirez - 2.35. Mike Aviles - 2.06. Cliff Pennington - 1.97. Elvis Andrus - 1.68. Erick Aybar - 1.09. Jhonny Peralta - 0.710. Alcides Escobar - 0.611. Asdrubal Cabrera - 0.112. Derek Jeter - (1.6)

That is not bad, I just don't see how any metric could put Ryan over Hardy. Hardy has a huge advantage over Ryan on assists and makes fewer errors to boot. That part is just crazy, IMHO. Jeter at 12 and Peralta at 9 are about right.

Ryan is a very good glove. You can make the argument that the basic metrics give the edge to Hardy but you can't discard the methodology simply because Ryan comes out on top.

 

[jon_mx]

Defensive WAR metrics for qualifying AL SS in 2012:1. Brendan Ryan - 3.62. JJ Hardy - 2.83. Yunel Escobar - 2.34. Alexei Ramirez - 2.35. Mike Aviles - 2.06. Cliff Pennington - 1.97. Elvis Andrus - 1.68. Erick Aybar - 1.09. Jhonny Peralta - 0.710. Alcides Escobar - 0.611. Asdrubal Cabrera - 0.112. Derek Jeter - (1.6)

That is not bad, I just don't see how any metric could put Ryan over Hardy. Hardy has a huge advantage over Ryan on assists and makes fewer errors to boot. That part is just crazy, IMHO. Jeter at 12 and Peralta at 9 are about right.

Ryan is a very good glove. You can make the argument that the basic metrics give the edge to Hardy but you can't discard the methodology simply because Ryan comes out on top.

But when one guy makes a 100 more plays against the same competition in a similar number of games, it is difficult to convince me that player isn't a more effective defensive player. Maybe he reacts quicker or anticipates better, but to me it is clear, he has a much better range.

 

[Doctor Detroit]

I guess I take issue with people quoting the Angels record with Trout (leaving out Pujols' stats while doing it), but not considering what the Tigers record would be without Cabrera (because it's unknown). Well if you watch the Tigers every day, you'd know they are a 75 win team without Cabrera. The Angels without Mike Trout the whole season? Still in the 80s IMO.

I agree that a lot of people don't know what WAR is; that's why i posted the topic.As for the above, this really has nothing to do with WAR. It's more of a faith-based statement. Though even with that, you say that the Tigers sans Cabrera is an unknown, and then turn around in the next sentence and say that it in indeed a known thing. Not really impressed with these guesses, no matter who does them. They're not really helpful to a meaningful discussion, because they are purely opinions.

Ok that's fine. But do you not admit WAR does have a lot of guessing in the formulas? Maybe "assumptions" is a better word actually. This is generally my issue with sabermetrics. I enjoy them, I enjoy getting to know them but I don't enjoy arguing with people who treat them as the final verdict, the decision to guideuis to the truth. We are getting folks that are poo-pooing raw numbers or not using the most important statistics (OPS, SLG, WHIP, K/9, K/BB) to create their opinions. I'm just merely suggesting that we continue to look at all the ingredients when making the soup, not just to rely on one or two set sabers that leave questions to be answered. To me you use these stats to bolster an argument, not to be the central point of one.

WAR does have a lot of assumptions as the value of base running and defense will never be certain. However, your some of your important statistics aren't the most important.SLG, K/9, and K/BB are pretty clear. OPS on the other hand is an over simplified way to address a player's offensive contribution. OBP has been shown to be more important than SLG so adding them together is flawed. wOBA is the best statistic we have since it includes yearly determined rates on the various offensive statistics to gauge a players value in the current run scoring environment. It even accounts for SB and CS in addition to the at the plate numbers. So either use OBP and SLG (and AVG still has its value since a single is better than a walk) or use wOBA with OBP and SLG on the side to determine where a hitter is best suited in a lineup.As for WHIP, the defense behind the pitcher has way too much impact for it to be pointed to as one of the most important statistics.

OPS is fairly easy to understand even if it does overlap slugging and if you look at the all-time OPS list, the really great hitters are on top. But wOBP is better, no argument there. I'm going to disagree with you on WHIP. Again let's look at the top ten list in WHIP for career:

Rank 	Player (yrs, age) 	WHIP	Throws1.	Addie Joss+ (9)	        0.9678	R2.	Mariano Rivera (18, 42)	0.9978	R3.	Ed Walsh+ (14)	        0.9996	R4.	Monte Ward+ (17)	1.0435	R5.	Pedro Martinez (18)	1.0544	R6.	Christy Mathewson+ (17)	1.0581	R7.	Trevor Hoffman (18)	1.0584	R8.	Walter Johnson+ (21)	1.0612	R9.	Mordecai Brown+ (14)	1.0658	R10.	Charlie Sweeney (6)	1.0673	R

1, 3, 4, 5, 6, 8 and 9 are probably considered by many as top 10 guys all-time. 2 and 7 are considered by most the two best closers of all-time. WHIP is somewhat team dependent but some of these guys like Walter Johnson and Addie Joss played on some pretty crappy teams. The pitcher controls walks, this is certain and the measure dictates derivative raw categories like K/BB per 9, ERA and saves. It's not just all-time guys either. Today's best pitchers Cain, Verlander, Hamels, Halliday, Weaver and Johan (ok well he is still recovering) are all in the top 100 all-time (and mostly make up the top of the active WHIP list). Walks reduce the amount of potential runs a pitcher has to face in any given game thus reducing pressure situations which add stress to the arm and mind. When eras cross we have to exclude BABIP which is designed for the modern game, but could change the formula to highlight extra base hits. Verlander for example plays with a defense that most would consider substandard, but he still had just a .273 BABIP this past year. Less baserunners equals less pressure situations which allows pitchers of this ilk to concentrate on the guy at the plate. Opp average is impacted and even the Tigers get softer ground balls and pop ups behind Verlander. Great pitchers power or otherwise have great control. Even a guy like Randy Johnson who lacked control in his early years finished in the top 100 in WHIP for his career, because he was a great pitcher. There are outliers like Nolan Ryan to the extreme side and Bob Gibson and Warren Spahn outside the top 100 but most of the truly great guys had good WHIP numbers. It goes to the art of pitching IMO and is the most easy to digest pitching ratio that correlates with greatness. WHIP to me is the most important single stat in pitching for those reasons.

 

[Dickie Dunn]

OPS is fairly easy to understand even if it does overlap slugging and if you look at the all-time OPS list, the really great hitters are on top. But wOBP is better, no argument there.

I definitely like the concept of wOBA, and it doesn't seem that much more difficult to calculate. And I assume that a lot of research was put into devising the formula, but do we trust that the values placed on each category are the correct ones? For instance, why does a player get more "credit" for reaching base on an error, which could be as simple as a second baseman bobbling a routine grounder, than he gets for a single? Or even a walk? That doesn't make sense to me.How much harder would it be to calculate an on-base percentage and simple add in the doubles, triples, and HRs to come up with an "adjusted" OBP?I agree that the best OPS values usually go to the best players, but can there be a better way?Using just a single game as a basic example:Batter A goes 1-for-4 with a 3B; Batter B goes 1-for-3 with a pair of BBs. Who had the "better game?"In this case, Batter A had an OPS of 1.000 (.250 OBP, .750 SLG), while Batter B had an OPS of .933 (.600 OBP, .333 SLG). But, even though Batter A had the same number of total bases in one fewer at-bat, I'd find it hard to argue that he was the more effective player.

 

[Eephus]

Defensive WAR metrics for qualifying AL SS in 2012:1. Brendan Ryan - 3.62. JJ Hardy - 2.83. Yunel Escobar - 2.34. Alexei Ramirez - 2.35. Mike Aviles - 2.06. Cliff Pennington - 1.97. Elvis Andrus - 1.68. Erick Aybar - 1.09. Jhonny Peralta - 0.710. Alcides Escobar - 0.611. Asdrubal Cabrera - 0.112. Derek Jeter - (1.6)

That is not bad, I just don't see how any metric could put Ryan over Hardy. Hardy has a huge advantage over Ryan on assists and makes fewer errors to boot. That part is just crazy, IMHO. Jeter at 12 and Peralta at 9 are about right.

Ryan is a very good glove. You can make the argument that the basic metrics give the edge to Hardy but you can't discard the methodology simply because Ryan comes out on top.

But when one guy makes a 100 more plays against the same competition in a similar number of games, it is difficult to convince me that player isn't a more effective defensive player. Maybe he reacts quicker or anticipates better, but to me it is clear, he has a much better range.

Ryan had a extremely high UZR this year at SS. It's probably an outlier even for him. I've read the explanations of UZR and followed them conceptually but even with 15 credits of college calculus, I don't really understand the inner workings of the stat.1000 defensive innings give or take is a small sample. Ryan and Hardy has comparable three year UZR aggregates. Both are good shortstops. Ryan's one year rating doesn't make the stat horribly flawed.Pretend it's Paul Ryan instead of Brendan and maybe you'll like the ranking better

 

[the moops]

Now there may be other factors at play with the Rays and some of that might be coincidence, but Tampa is a 102 win team with him and a 90 win team without him. A projected 6.5 WAR doesn't cover the difference.

Yes, there are probably 100 other factors.But didn't Sean Rodriguez play a bunch of those games at 3B? he is terrible, and likely below replacement level terrible.

 

[Eephus]

I'm going to disagree with you on WHIP. Again let's look at the top ten list in WHIP for career:

Rank 	Player (yrs, age) 	WHIP	Throws1.	Addie Joss+ (9)	        0.9678	R2.	Mariano Rivera (18, 42)	0.9978	R3.	Ed Walsh+ (14)	        0.9996	R4.	Monte Ward+ (17)	1.0435	R5.	Pedro Martinez (18)	1.0544	R6.	Christy Mathewson+ (17)	1.0581	R7.	Trevor Hoffman (18)	1.0584	R8.	Walter Johnson+ (21)	1.0612	R9.	Mordecai Brown+ (14)	1.0658	R10.	Charlie Sweeney (6)	1.0673	R

Monte Ward's B-R page is ridiculous. His biography is even more impressive.

 

[Doctor Detroit]

I'm going to disagree with you on WHIP. Again let's look at the top ten list in WHIP for career:

Rank Player (yrs, age) WHIP Throws1. Addie Joss+ (9) 0.9678 R2. Mariano Rivera (18, 42) 0.9978 R3. Ed Walsh+ (14) 0.9996 R4. Monte Ward+ (17) 1.0435 R5. Pedro Martinez (18) 1.0544 R6. Christy Mathewson+ (17) 1.0581 R7. Trevor Hoffman (18) 1.0584 R8. Walter Johnson+ (21) 1.0612 R9. Mordecai Brown+ (14) 1.0658 R10. Charlie Sweeney (6) 1.0673 R

Monte Ward's B-R page is ridiculous. His biography is even more impressive.

Yeah I've read that one before, at least some of it. Definitely one of the true pioneers of the game. 3rd in pitcher WAR, nine years later...111 steals. Even Ricky Henderson would be impressed by that.

 

[bicycle_seat_sniffer]

cant wait to buy my Mike Trout WAR champion tee-shirt

 

[oso diablo]

This isn't rocket science. I could produce better metrics than what are being used, at least the public ones which seem rather simplistic and not so valuable.

go ahead. you could make quite a bit of $ if so.

 

[oso diablo]

cant wait to buy my Mike Trout WAR champion tee-shirt

why are you so negative about this? are you just that way in general, or is there something specific about advanced stats that gets a burr in your saddle?

 

[bicycle_seat_sniffer]

cant wait to buy my Mike Trout WAR champion tee-shirt

why are you so negative about this? are you just that way in general, or is there something specific about advanced stats that gets a burr in your saddle?

I know nothing about "advanced" stats. just want my boy miggy to win the MVP. Im not a baseball geek

 

[Premier]

This isn't rocket science. I could produce better metrics than what are being used, at least the public ones which seem rather simplistic and not so valuable.

this thread is amazing.

 

[jon_mx]

This isn't rocket science. I could produce better metrics than what are being used, at least the public ones which seem rather simplistic and not so valuable.

this thread is amazing.

Well let's look at range which just takes assists and putouts and divides by innings. An assist or putout can be things really don't apply to range, such as having the having the second baseman flip you the ball for an out or turning a double play or catching a pop up. What I want to see is how many ground balls an infielder gets to. That would make it a far more meaningful stat instead of including stuff that is more dependent upon the players around you which just mucks up the stat.

 

[Premier]

(For what it's worth, Zobrist, per baseball-reference.com, over 2009-12 has the highest WAR (wins above replacement), 26.6, of any player, ahead of Albert Pujols, Ryan Braun, Miguel Cabrera, etc.)

I did not know that.

 

[jon_mx]

(For what it's worth, Zobrist, per baseball-reference.com, over 2009-12 has the highest WAR (wins above replacement), 26.6, of any player, ahead of Albert Pujols, Ryan Braun, Miguel Cabrera, etc.) I did not know that.

He must be replacing somebody who really sucks

 

[oso diablo]

He must be replacing somebody who really sucks

shtick? or have you really not read the explanations?

 

[jon_mx]

He must be replacing somebody who really sucks

shtick? or have you really not read the explanations?

Lol, it was a joke. Sorry

 

[bicycle_seat_sniffer]

(For what it's worth, Zobrist, per baseball-reference.com, over 2009-12 has the highest WAR (wins above replacement), 26.6, of any player, ahead of Albert Pujols, Ryan Braun, Miguel Cabrera, etc.) I did not know that.

yeah Im sure Zorby is worth pujlos

 

[Thunderlips]

So the foundation of this stat is a comparison of a player to an imaginary player?....LOL!....OK.

These "hypothetical" situations are great mental masturbation exercises.....but in the end, can only hope to provide a very preliminary sketch of a players real talent.

 

[oso diablo]

So the foundation of this stat is a comparison of a player to an imaginary player?....LOL!....OK.

if it helps to understand, the replacement level is a composite, akin to an Obama girlfriend.

 

[Premier]

(For what it's worth, Zobrist, per baseball-reference.com, over 2009-12 has the highest WAR (wins above replacement), 26.6, of any player, ahead of Albert Pujols, Ryan Braun, Miguel Cabrera, etc.) I did not know that.

yeah Im sure Zorby is worth pujlos

He's not. He's worth more.