WAR is the dumbest stat known to mankind (Trout vs. Betts)

[S5Dewey]

No, not true for the most part.

 

1. Stat geeks understand the shortcomings of WAR.

2. You are making a far bigger deal out of the 17% margin of of error than it is.

 

17% is 17% and that's a big number no matter how you slice it. We're talking about an acceptable(?) margin of error of almost 1/5. In what statistical world is that acceptable?

 

Therefore WAR is NOT as useful as most people think it is.

[greenwell]

17% is 17% and that's a big number no matter how you slice it. We're talking about an acceptable(?) margin of error of almost 1/5. In what statistical world is that acceptable?

 

Therefore WAR is NOT as useful as most people think it is.

 

Weather forecasting

[Bellhorn04]

17% is 17% and that's a big number no matter how you slice it. We're talking about an acceptable(?) margin of error of almost 1/5. In what statistical world is that acceptable?

 

Therefore WAR is NOT as useful as most people think it is.

 

In a field of science as imprecise as measuring the performance of a baseball player I think 17% is probably acceptable.

 

The underlying question is, can the margin of error be made any lower? If yes, I'm sure it will be done.

 

If no, we all just have to accept it for what it is.

Edited August 17, 2018 by Bellhorn04

[S5Dewey]

In a field of science as imprecise as measuring the performance of a baseball player I think 17% is probably acceptable.

 

The underlying question is, can the margin of error be made any lower? If yes, I'm sure I will be done.

 

If no, we all just have to accept it for what it is.

 

Usually when someone says "something is (or is not) as useful as 'most people' think it is" that statement is based on speculation as to what 'most people' think.

 

Since I see many comparisons here saying that Player A has a WAR of X.x and Player B has a WAR

 

I think it is too, but only outside that 17% parameter. I see it as 'most people' believing it's useful within that 17% when even Fangraphs says it's not.

[Bellhorn04]

Usually when someone says "something is (or is not) as useful as 'most people' think it is" that statement is based on speculation as to what 'most people' think.

 

Since I see many comparisons here saying that Player A has a WAR of X.x and Player B has a WAR

 

I think it is too, but only outside that 17% parameter. I see it as 'most people' believing it's useful within that 17% when even Fangraphs says it's not.

 

There's a big difference between believing that WAR is a useful tool and believing that it's an absolutely reliable determinant in comparing players.

 

On this forum we see a lot of different numbers put up. moonslav puts up the most numbers and I think it's more often OPS or ERA rather than WAR.

 

For me WAR is kind of a starting point. If I see a WAR that seems oddly high or low I want to look into it and see what's going on.

 

Examples:

 

At first I was a bit surprised when I looked at Big Papi's WAR, much lower than I expected considering his huge numbers. But the explanations of course were that he didn't play defense and he was slow as molasses.

 

Last year I was surprised at how low Pedroia's WAR was. Looking into it I found it was because of his terrible baserunning numbers. It really highlighted how much of his speed he had lost.

 

Just little nuggets like that make it useful for me.

[moonslav59]

There's a big difference between believing that WAR is a useful tool and believing that it's an absolutely reliable determinant in comparing players.

 

On this forum we see a lot of different numbers put up. moonslav puts up the most numbers and I think it's more often OPS or ERA rather than WAR.

 

For me WAR is kind of a starting point. If I see a WAR that seems oddly high or low I want to look into it and see what's going on.

 

Examples:

 

At first I was a bit surprised when I looked at Big Papi's WAR, much lower than I expected considering his huge numbers. But the explanations of course were that he didn't play defense and he was slow as molasses.

 

Last year I was surprised at how low Pedroia's WAR was. Looking into it I found it was because of his terrible baserunning numbers. It really highlighted how much of his speed he had lost.

 

Just little nuggets like that make it useful for me.

 

Well said.

 

I don't look at WAR first.

 

I don't think anyone here feels it is the be all- end all.

 

Harmony uses it more than anybody, and he's not even a Sox fan.

[notin]

...or postings!

 

Do people really think postings on message boards are useful?

[S5Dewey]

Do people really think postings on message boards are useful?

 

Apparently they're useful to some of us.

 

Bellhorn: Just little nuggets like that make it useful for me.

Are they useful to you?

[notin]

Do you really want to get into a discussion of "useful", what's useful to who, and why?

 

Yes. That would be very useful...

[S5Dewey]

Yes. That would be very useful...

 

Ok. You start.

[Bellhorn04]

The usefulness of anything in the realm of sports is purely subjective.

 

And from a fan's perspective, can only be useful in the sense of providing some sort of pleasure.

[User Name]

The usefulness of anything in the realm of sports is purely subjective.

 

And from a fan's perspective, can only be useful in the sense of providing some sort of pleasure.

 

FO's use WAR as a measuring tool, since that is what it is, imperfect as it may be. Those young whippersnappers need to stop using WAR and start watching the games!

[dannycater]

WAR is Hell.

[OH FOY]

Always wanted to know why don't they use it in Minors? If it is such a good tool, I would think that's the best place to use it.

[kenmeister]

Another article on who should win the AL MVP: https://www.cbssports.com/mlb/news/why-mookie-betts-deserves-the-al-mvp-amid-a-crowded-field-packed-with-deserving-candidates/

 

Look how far down the list JDM is!

[notin]

FO's use WAR as a measuring tool, since that is what it is, imperfect as it may be. Those young whippersnappers need to stop using WAR and start watching the games!

 

Are you the Goose or the Wolfman?

[Kimmi]

17% is 17% and that's a big number no matter how you slice it. We're talking about an acceptable(?) margin of error of almost 1/5. In what statistical world is that acceptable?

 

Therefore WAR is NOT as useful as most people think it is.

 

Incorrect on both counts.

[S5Dewey]

Incorrect on both counts.

 

Untrue.

[notin]

Ok. You start.

 

Well I tried continuing the discussion. You said WAR was not perfect, which it isn't and everyone knows that. I said "What stat is?" Then you started attacking my posts...

[sk7326]

WAR tells us what we already know so we tend to believe it. It tells us that JDM is a better player than, say, Christian Vazquez and that Mookie is better than Brock Holt.

 

However, we as consumers tend (want?) to believe that any value expressed as an integer is concise and WAR doesn't work that way. Fangraphs in it explanation of how WAR works says that "Given the imperfections of some of the available data and the assumptions made to calculate other components, WAR works best as an approximation. A 6 WAR player might be worth between 5.0 and 7.0 WAR..." and then goes on to say that player in that group is probably an All-Star, etc.

 

That's a 17% margin for error and it means that a player with a WAR of 3.0 may actually be worth somewhere between 2.5 & 3.5. Yet when we see a player with a WAR of 3.5 we assume that he's 1/4 again as good as that 2.5 player when actually they may be the same player once the variables are accounted for. What we do know from WAR is that Mookie is better than a player with a WAR of 3.0 - but we already knew that from watching them.

 

Yes, WAR is valuable in confirming what we already know but when we're comparing two players whose WAR are within a couple of points WAR is essentially meaningless.

 

WAR is more directionally accurate than precisely accurate.

 

Now what is a 17% error. Think of it this way. If a player registered 6 wins above replacement - that is actual measured performance. It is true. Now, does that mean he is a 6-win player? No. THAT is where the range of error kicks in. Similary if a player goes 3 for 10, is he REALLY a .300 hitter? No. Heck, even at 80 at-bats (a 24 for 80 stretch) you are still looking at a spread between .198 and .402. Get to 160 at-bats and it's .228 to .372.

 

Once you take that sort of sampling into account - a 17% spread is actually pretty good. It's not the 6.0 WAR which is "wrong" - it is exactly what its components say it is, but what it says about the underlying contribution of the player.

 

If you look at the AL fWAR rankings it tells you the obvious - the AL MVP is a 4 man race between Ramirez, Lindor, Trout and Betts. I'd pick Betts but none of those 4 are bad choices.

Edited August 23, 2018 by sk7326

[Kimmi]

WAR is more directionally accurate than precisely accurate.

 

Now what is a 17% error. Think of it this way. If a player registered 6 wins above replacement - that is actual measured performance. It is true. Now, does that mean he is a 6-win player? No. THAT is where the range of error kicks in. Similary if a player goes 3 for 10, is he REALLY a .300 hitter? No. Heck, even at 80 at-bats (a 24 for 80 stretch) you are still looking at a spread between .198 and .402. Get to 160 at-bats and it's .228 to .372.

 

Once you take that sort of sampling into account - a 17% spread is actually pretty good. It's not the 6.0 WAR which is "wrong" - it is exactly what its components say it is, but what it says about the underlying contribution of the player.

 

If you look at the AL fWAR rankings it tells you the obvious - the AL MVP is a 4 man race between Ramirez, Lindor, Trout and Betts. I'd pick Betts but none of those 4 are bad choices.

 

This post is music to my ears.

[notin]

17% is 17% and that's a big number no matter how you slice it. We're talking about an acceptable(?) margin of error of almost 1/5. In what statistical world is that acceptable?

 

Therefore WAR is NOT as useful as most people think it is.

 

So again, what is better?

[S5Dewey]

So again, what is better?

 

No thank you. I'm exhausted on this topic. Some people think its great in spite of the big margin of error and I don't. I'm good with that.

 

I'll even use WAR in a discussion occasionally but that doesn't mean I have a lot of faith in it. It's just, as you implied, I can't identify anything better. But I will say that "not able to identify anything better" is a pretty low bar.

 

If Fangraphs admitted to a 50% margin of error would you still like it? How about 80%? At what point would YOU say it's "not as useful as most people think it is"?

[sk7326]

No thank you. I'm exhausted on this topic. Some people think its great in spite of the big margin of error and I don't. I'm good with that.

 

I'll even use WAR in a discussion occasionally but that doesn't mean I have a lot of faith in it. It's just, as you implied, I can't identify anything better. But I will say that "not able to identify anything better" is a pretty low bar.

 

If Fangraphs admitted to a 50% margin of error would you still like it? How about 80%? At what point would YOU say it's "not as useful as most people think it is"?

 

I would submit that RBIs have a much much worse track record at identifying good hitters for example. Pitcher wins as well.

[notin]

No thank you. I'm exhausted on this topic. Some people think its great in spite of the big margin of error and I don't. I'm good with that.

 

I'll even use WAR in a discussion occasionally but that doesn't mean I have a lot of faith in it. It's just, as you implied, I can't identify anything better. But I will say that "not able to identify anything better" is a pretty low bar.

 

If Fangraphs admitted to a 50% margin of error would you still like it? How about 80%? At what point would YOU say it's "not as useful as most people think it is"?

 

There is no "margin of error". WAR is, was, and always meant to be an approximation, not a definitive value. That doesn't mean it isn't useful; it just means it isn't absolute. Is that really a flaw? That's like saying meteorology is flawed. I have heard the weatherman say "50% chance of rain." Well, it's either going to rain or it isn't. Should we disregard that entire field of science based on the weatherman being off by a full 50%?

 

WAR tries to take the unimaginable task of attempting to determine the overall contribution of a player. It hs many flaws. It uses a floating reference of a replacement player. It's projections into the future are pretty laughable. But the value of WAR doesn't and wasn't meant to directly translate to wins. But it does give some overall perspective of the value of one player as opposed to many other players many of us rarely if ever get to watch.

 

It's a tough concept to grasp because it isn't a concrete value like batting average. But then batting average is actually a pretty useless stat if you think about it. I mean, the definition of a "hit" is hardly universal. A player robbed of a home run doesn't get one, while the player who beats out the weak grounder does. but who was clearly the better hitter? Not to mention, all the players in MLB always fall into what is actually a very tight range. A .280 hitter, whom most fans think is good, only gets 4 more hits per 100 at-bats than a .240 hitter. That's maybe one hit per week. Why is the .280 hitter so revered over the .240 hitter? Especially when practically every player in the entire league will hit between .200 and .300. That small 10% range of success is so magnified by fans and sportswriters as if the differences are really of major significance.

 

But as batting average is a very simple concept, people grasp it and it is and will likely continue to be the most cited of all offensive statistics. And even I will continue to cite it as required, for those very reasons.

 

WAR has a few flaws, the least of which is the range. Like any value with a tolerance, the nominal value is the ideal one. For players too close within that range (which is never disclosed in the article you read, but only given as an example), it doesn't end the debate of who is better. It just adds fuel to it.

 

You claim it doesn't tell us anything we don't already know, but, really it does. For example. Khris Davis is on a home run tear and becoming a player some feel is a legitimate MVP candidate himself. But WAR says he is the fourth most significant position player on his team, behind Matt Chapman, Jed Lowrie and Marcus Semien. Whether he is really fourth or not, doesn't WAR tell you maybe his contributions beyond the home runs might not be so great? And maybe there is a better MVP candidate on that team?

 

And looking overall at the AL, is Jose Ramirez having a better season than Mookie? As of today, WAR says he is (8.2 fWAR vs 7.7 fWAR). That doesn't guarantee him the MVP, but it certainly helps his case. And like in Oaklamd, it is just a little bit of fuel for the debate...

[bosoxmal]

I totall agree, and I wonder how many people pay attention to it anymore@

[S5Dewey]

That's OK with me. I'm sticking with "not as valuable as most people think it is".

 

IMHO when "most people" see a list of 3 players, one with a WAR of 3.50, one with a WAR of 4.00 and a third with a WAR of 4.50 they don't say, "Let's see, 4.0 plus or minus 17% is a range of 3.32 to 4.68 so these three players are all statistically the same." Instead, "most people" are going to assume that the player with the 4.5 WAR is a better player than the 3.5 player. At least I would, and I think I'm like most people.

 

Not as valuable as "most people" think it is.

[S5Dewey]

 

And looking overall at the AL, is Jose Ramirez having a better season than Mookie? As of today, WAR says he is (8.2 fWAR vs 7.7 fWAR). That doesn't guarantee him the MVP, but it certainly helps his case. And like in Oaklamd, it is just a little bit of fuel for the debate...

 

I don't think you want to say that it helps his (Ramirez) case. It says what I said above, that the people voting for the MVP award are "most people".

[notin]

I don't think you want to say that it helps his (Ramirez) case. It says what I said above, that the people voting for the MVP award are "most people".

 

I would say that leading in WAR does help Ramirez' case. Definitely. However, at the end of the season, he may not still be leading Betts.

 

And the "most people" who vote are the BBWAA, and they as a collection appear to have varying degrees of acceptance of WAR. But there will be those who say it is too close to be a factor and those who ignore it completely in favor of traditional stats. In fact, I would not be surprised if more votes go to Khris Davis and his 2.6 fWAR over Matt Chapman and his 5.4 fWAR, simply because Davis got hot at the right time and could easily be the AL home run leader.

 

But WAR at least puts Chapman into the argument, which a lot of the traditional stats do not necessarily do. And if you like to argue this is a game played by human beings and not by statistics, WAR does a much better job of making that the case than simply looking at who leads the league in home runs. After all, what else is putting Chapman into the argument?

[notin]

That's OK with me. I'm sticking with "not as valuable as most people think it is".

 

IMHO when "most people" see a list of 3 players, one with a WAR of 3.50, one with a WAR of 4.00 and a third with a WAR of 4.50 they don't say, "Let's see, 4.0 plus or minus 17% is a range of 3.32 to 4.68 so these three players are all statistically the same." Instead, "most people" are going to assume that the player with the 4.5 WAR is a better player than the 3.5 player. At least I would, and I think I'm like most people.

 

Not as valuable as "most people" think it is.

 

That's because those people would be stupid to look at WAR that way. In fact, simply saying the player with 4.5 fWAR is having the best season would be the simplest and probably most common and maybe even most correct way to look at it.

 

This 17% tolerance appears to be your creation. The article you cited says a 6 fWAR player might be between 5 WAR and 7 WAR, but that doesn't mean they are even defining a range or a tolerance as much as they are trying to make a point, which is that it isn't anything specific. A big part of this is it isn't measuring anything, which really makes the notion hat it has a tolerance as useless.

 

Now if I asked you today who was the MVP of the A's, who would you say? If you are "most people", you probably say Khris Davis. ESPN says so. Certainly an argument can be made for Davis, but my point is that Chapman deserves a look, and might be the better candidate, which is one of the values WAR provides...