Question about WAR (Machado vs. Trout)

[TradeAngelos]

Close enough, assuming (as I did) he means the combined difference is worth about 0.6. 0.3 for the difference in games and 0.34 for the difference in baserunning.

I mean sorry if I didn't specify the .04 but yes this is correct.

[Frobby]

Park factors have a pretty big impact here, too. If you just look at raw numbers, Manny looks better, but his OPS+ is 158 compared to Trout's 163 because of the stadiums in which they've played.

[OrioleMagic79]

Haha ... the more complicated they make the stat, the harder it becomes to call it BS because there is always an excuse.

[BohKnowsBmore]

Park factors have a pretty big impact here, too. If you just look at raw numbers, Manny looks better, but his OPS+ is 158 compared to Trout's 163 because of the stadiums in which they've played.

I like the idea of park factors, but sometimes it makes me a bit suspicious of its impact on individual player comparisons. No way to tell if an individual benefits in line with the aggregate. In some ways I worry it over-normalizes on occasion.

Sent from my iPhone using Tapatalk

[El Gordo]

Lets see, who would I rather have, the best defensive and offensive 3B, possibly in the history of the game. Or a slightly defensively challenged, great offensive CF? Manny is Beltre's bat with Brooks' glove. Sorry Cal.

[ArtVanDelay]

Lets see, who would I rather have, the best defensive and offensive 3B, possibly in the history of the game. Or a slightly defensively challenged, great offensive CF? Manny is Beltre's bat with Brooks' glove. Sorry Cal.

[Phantom]

Haha ... the more complicated they make the stat, the harder it becomes to call it BS because there is always an excuse.

Who is making excuses?

[Moose Milligan]

Haha ... the more complicated they make the stat, the harder it becomes to call it BS because there is always an excuse.

Which is why I hate WAR. How good of a stat can it be when there's the bbref version and the fangraph version?

I'm almost at a point where anything Manny does like win an MVP or leads in WAR is just going to drive up any asking price moving forward and I'm not excited about that. Let Trout have a higher WAR and win an MVP this year, go right ahead.

[Phantom]

Which is why I hate WAR. How good of a stat can it be when there's the bbref version and the fangraph version?

bbref and fangraphs simply have different theories on what value means in practice. Is there a stat you like better for judging a players value?

[BohKnowsBmore]

Which is why I hate WAR. How good of a stat can it be when there's the bbref version and the fangraph version?

I'm almost at a point where anything Manny does like win an MVP or leads in WAR is just going to drive up any asking price moving forward and I'm not excited about that. Let Trout have a higher WAR and win an MVP this year, go right ahead.

bbref and fangraphs simply have different theories on what value means in practice. Is there a stat you like better for judging a players value?

I actually like having two flavors based on the somewhat subjective nature of the calculation, weighting, etc. I sort of wish we had a third source. It's imperfect, but I like seeing the range of estimates, rather than a single, seemingly-definitive number. Wide variations between different sources raise questions that inspire you to dig a bit deeper on what's causing it to pop. That's why I think there's value in WAR as well as the old counting stats. I like having as much information as possible.

[Phantom]

I actually like having two flavors based on the somewhat subjective nature of the calculation, weighting, etc. I sort of wish we had a third source. It's imperfect, but I like seeing the range of estimates, rather than a single, seemingly-definitive number. Wide variations between different sources raise questions that inspire you to dig a bit deeper on what's causing it to pop. That's why I think there's value in WAR as well as the old counting stats. I like having as much information as possible.

Well put.

[Tryptamine]

Which is why I hate WAR. How good of a stat can it be when there's the bbref version and the fangraph version?

I'm almost at a point where anything Manny does like win an MVP or leads in WAR is just going to drive up any asking price moving forward and I'm not excited about that. Let Trout have a higher WAR and win an MVP this year, go right ahead.

They both measure different things, especially when it comes to pitchers. Fangraph's WAR, which is called fWAR, is more predictive in nature. It's not necessarily about what you have done, it's about what your peripherals suggest you will do in the future.fWAR uses FIP to figure WAR The Bref version is based on what actually happened. For example, a pitcher with a great ERA but who has been exceedingly lucky with babip and lob% will have a very high rWAR because bbref uses ERA to calculate WAR. With fWAR you can have a poor ERA and still have a very good fWAR if you have strong peripherals, because it's predicting that you are likely to succeed in the future and that you were just unlucky to have such a high ERA despite the good peripherals.

[TradeAngelos]

They both measure different things, especially when it comes to pitchers. Fangraph's WAR, which is called fWAR, is more predictive in nature. It's not necessarily about what you have done, it's about what your peripherals suggest you will do in the future.fWAR uses FIP to figure WAR The Bref version is based on what actually happened. For example, a pitcher with a great ERA but who has been exceedingly lucky with babip and lob% will have a very high rWAR because bbref uses ERA to calculate WAR. With fWAR you can have a poor ERA and still have a very good fWAR if you have strong peripherals, because it's predicting that you are likely to succeed in the future and that you were just unlucky to have such a high ERA despite the good peripherals.

This is why I like bbref better. Leave the "projections" out of it and just tell me what the

"production" has been.

[Phantom]

This is why I like bbref better. Leave the "projections" out of it and just tell me what the

"production" has been.

I feel the exact opposite (and similarly hate ERA).

[TradeAngelos]

Stats measure what has happened, not what is going to.

You can look to any one of the "fine, always accurate, never that far off" projection models if you want to see the future. Allegedly.

That isn't what WAR, or any other counting stat using past performance is supposed to be IMO.