Posted by bad_luck on 1/21/2015 10:31:00 AM (view original):

Posted by burnsy483 on 1/21/2015 10:29:00 AM (view original):
Sure, but even if they were considered equals, I'd disagree with that finding. There's no way Boston's defense was that good. Evans/Armas/Rice, if anything, screams "bad defense" in the outfield.  Boggs was good defensively. Aside from that, it doesn't make a ton of sense.

I don't think it's defense. I think it's park effects.

Was County Stadium a hitter's park? I know Fenway is pretty neutral. 

Maybe. I'll look when I get to work.

Posted by burnsy483 on 1/21/2015 10:32:00 AM (view original):

Posted by MikeT23 on 1/21/2015 10:28:00 AM (view original):
If the WAR was closer, it might be interesting.    That's more than 5% difference when it looks reasonably obvious that Clemens was better.   I know BL said 3.4 and 2.6 were "essentially the same" a couple of years ago but, as I said then, if that much of a difference, percentage-wise, is "essentially the same", how useful can said stat be when 10 is a huge number?

The WAR is pretty close.

I don't fully understand what you're trying to say.  

I'm saying none of the other stats are that close.

Clemens pitched more, gave up fewer hits, walks and runs while compiling more strikeouts.    Some of those 5 keys stats aren't close.    If it was a 5% difference in Clemens' favor, I'd think the stat is more useful even though I think Clemens was more than 5% more effective.

Yea, it appears the stat thinks too much of either Boston's defense or the park effects of the 2 stadiums, I agree. 

If a WAR is 5% higher than someone else, it doesn't mean they're "5% more effective." In the same way someone with a WAR of 2 isn't "100% more effective" than someone with a WAR of 1. Or is it 200%? I've confused myself.

Another thing to consider - Higuera faced Boston twice in 86, and Clemens faced Milwaukee twice. Clemens won those two games against Milwaukee, allowing 1 run in 18 innings. Higuera never got the opportunity to face Milwaukee, but did have to pitch against Boston. Maybe that adds to it.

I just used "5% more effective" as an example.    A 10 WAR is huge while a 1 WAR is low.   I don't think that means one guy was 10 times better than the other(though he might have been).  

I just don't like any metric that say Higuera was better than Clemens in 1986.    Even .00001 better would look off.

Yes, I agree. I don't see how he's possibly better, or equal. But in the same way where a pitcher with a 1.10 WHIP could be obviously worse than one with a 1.15 WHIP. Or 20 wins to 18 wins. Or a 3.00 ERA to a 3.15 ERA. Doesn't mean the stat itself is trash.

I've said many times it has use.   You should know I just like to point out when it completely sucks because WAR is the God of Stats to BL.   

And, in this case, it completely sucks.

Posted by burnsy483 on 1/21/2015 11:02:00 AM (view original):
Yes, I agree. I don't see how he's possibly better, or equal. But in the same way where a pitcher with a 1.10 WHIP could be obviously worse than one with a 1.15 WHIP. Or 20 wins to 18 wins. Or a 3.00 ERA to a 3.15 ERA. Doesn't mean the stat itself is trash.

FWIW, Clemens smokes Higuera according to Fangraphs WAR, 7.8 to 5.5. Mike Scott is the only pitcher that beats Clemens in 1986 by fWAR.

Yet more proof that WAR is a massively retarded stat.  Different sites using different calculations, but calling it the same thing.

And no one really able to explain why something is something.

Posted by burnsy483 on 1/21/2015 11:02:00 AM (view original):
Yes, I agree. I don't see how he's possibly better, or equal. But in the same way where a pitcher with a 1.10 WHIP could be obviously worse than one with a 1.15 WHIP. Or 20 wins to 18 wins. Or a 3.00 ERA to a 3.15 ERA. Doesn't mean the stat itself is trash.

I'm not anti-WAR, but there's a big difference between all these stats you're talking about and a metric like WAR.  These other stats are simply quantitative measures.  They don't offer any particular inherent judgment they just are.  If pitcher A had a 1.10 WHIP and B had a 1.15, that actually happened, there's no disputing it.  What value you assign, if any, to that happening is up to you.  There's nothing about WHIP that's inherently meant to say Pitcher A > Pitcher B.

WAR, on the other hand, tries to take all the base quantitative measures and quantify the overall value of a player.  So if those comparisons consistently don't pass the smell test, then that's potentially a big deal.

Posted by burnsy483 on 1/21/2015 10:46:00 AM (view original):
Another thing to consider - Higuera faced Boston twice in 86, and Clemens faced Milwaukee twice. Clemens won those two games against Milwaukee, allowing 1 run in 18 innings. Higuera never got the opportunity to face Milwaukee, but did have to pitch against Boston. Maybe that adds to it.

Looking at the BR calculations, it looks like Higuera faced tougher offenses (RA9opp was slightly higher), in front of a worse defense (RA9def was lower), in tougher ballparks (PPFp was higher).

The average pitcher pitching against the same offenses, in front of the same defenses, and in the same ballparks as Higuera would have allowed an estimated 5.35 RA9. Compared to his actual RA9 of 3.04, Higuera was 63 runs better than average, 88 runs better than replacement.

Compared to Clemens RA9 of 2.73, the average pitcher (same offenses, same defenses, same ballparks) gives up an estimated 4.82 RA9, making him 58 runs better than average and 83 runs better than replacement.

It certainly isn't perfect, but it isn't ridiculous to say that those two had similar seasons, relative runs allowed, when you account for the quality of their opponents, the skill of the defenses behind them, and the ballparks they pitched in.

Posted by The Taint on 1/21/2015 11:58:00 AM (view original):
And no one really able to explain why something is something.

What?  You mean when someone says "I'm just going to divide by 3.28977665 then add 4", you want to know why?

The calculations and explanations are available on BR for anyone that wants to see them.