I’ve had folks tell me lately that even if Cabrera wins baseball’s first Triple Crown since 1967, Trout remains the MVP pick. My apologies. I simply cannot comprehend that.
But fans of Sabermetrics, these new-age stats that attempt to give you an all-around worth of a player, actually will tell you — and with a straight face! — that batting average and RBIs are irrelevant today.
There might — might — be the case with batting average, because that doesn’t necessarily tell you how clutch or how much of a run producer a player is. There were some in Seattle who thought Ichiro Suzuki could have hit many more home runs, but didn’t because he wanted his batting average to stay sky high.
But RBIs? Really? Irrelevant? The critics like to say that’s only a measurement of how often guys get on ahead of you. Well, to that I have a two-word response: Delmon Young.
...Now, I ask you, is that an MVP? That’s for 28 of my peers to decide; I don’t have an MVP vote this year.
But if that is an MVP, then let me say this: Voters better be consistent and make Tigers ace Justin Verlander the Cy Young winner for a second straight year, because he leads the major leagues in WAR, too, not contenders Felix Hernandez or David Price.
By the way, here’s another thing FanGraphs’ WAR tells us, folks: Anibal Sanchez (3.1) and Rick Porcello (2.9) are worth more to their team than Angels ace Jered Weaver (2.7) is to his. Never mind that Weaver is 18-4 with a 2.79 ERA. What do those stats matter anymore?
Repoz
Posted: September 23, 2012 at 10:23 AM |
180 comment(s)
Login to Bookmark
Tags:
awards,
mvp
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.
Deal. Now shut up.
Has something with the ballpark changed in Anaheim? Three years ago it was netural and now it's an obscene pitcher's park.
I know WAR is considered holy around here, but if this is true, it's at least kind of weird.
BB-Ref has Weaver at 3.3, Porcello at 1.5, and Sanchez at 1.7 (1.4 for Miami, 0.3 for Detroit).
FaGraphs WAR is useless for pitchers. FIP does not reflect the reality of run prevention. It is not a value stat.
BRef WAR has Weaver at 3.3 (still seems low), Porcello at 1.5, and Sanchez at 1.7.
Edit: half a Coke to Cooper!
Slugging is a geek stat?
Truthfully I don't have a problem with people who want to make Cabrera the MVP; I understand the point of view. But arguing that there is no basis for choosing Trout instead is ridiculous and close minded.
Not that this changes the overall point at all, however ... I know I could look this up, but I'm being lazy today. Isn't fWAR based on xFIP rather than straight FIP? That is it is based on actual Ks and BBs, but instead of actual HRs is based on estimated HRs from number of fly balls allowed?
Does Fangraphs believe that the Tigers' defense is so bad that everything good that happens in Porcello's starts is solely attributable to HIM?
His HR and BB numbers are pretty good, but he leads the league in hits allowed, thanks in part to a very high BABip, but also the fact that he only strikes out 5.4 per 9 IP. He's allowed 11 unearned runs.
Their Glossary says FIP. Had to look it up, b/c I never look at FanGraphs for pitcher value.
Does Fangraphs believe that the Tigers' defense is so bad that everything good that happens in Porcello's starts is solely attributable to HIM?
His HR and BB numbers are pretty good, but he leads the league in hits allowed, thanks in part to a very high BABip, but also the fact that he only strikes out 5.4 per 9 IP. He's allowed 11 unearned runs.
It's purely based on FIP. K/BB/HR. It has no idea how many runs, unearned or earned Porcello has allowed.
Porcello has a similar FIP to Weaver, and pitches in a less friendly park. That's all there is to it.
EDIT: Sorry, Porcello hasn't been "significantly" better by fWAR, they've been approximately equal. But I still don't believe it.
Nor should you. FIP is a predictive stat, not a value stat.
In pretty much every thread of this kind, at least one person states that WAR is nonsense or that they don't believe the numbers or whatever. It's not "considered holy" around here at all.
It's FIP-based, so it removes all influence of the fielders (and just uses SO, BB, HR). If you look at Fangraphs' RA9-wins on the same Porcello page, it's 1.1, which is a lot closer to the BRef numbers (similarly, they allow you to play around with the components if you want). I have a lot of problems with Fangraphs WAR, but they have been doing a great job with including components and such so that if you take the time to read some definitions, you'll get a lot of what you need.
Mike Trout has scored 121 Runs to lead Cabrera, and the whole league, by about a month's worth. There is no reason even to divide one number into another to make him a strong MVP candidate.
Sorry, I've just been beaten down a lot over the past week in other threads for expressing doubt that the difference between Trout and Cabrera is really on the order of 4 wins. So I'm a bit timid in expressing any additional doubt.
Really? War is possibly overrated as a stat for position players, but I don't think anyone really looks at pitchers war and says that is the beginning and end of the argument.
I think you're definitely right that the gap isn't that big.
Trout is getting +25 from fielding, which seems unbelievable in only 120 Gs. He's also getting a big park adjustment; Trout 953 OPS = 168 OPS+, Cabrera 1014 OPS = 170 OPS+).
I would guess the gap is closer to 2 WAR than 4. That said, Trout has still been better.
As smarter folks than me have noted, WAR doesn't have anything to do with it. Either defense counts towards "valuable" or it doesn't.
Yeah, isn't Verlander likely the favorite among SABR folks for Cy Young at this point?
4 Wins might be a bit high, and some of that could be argued that it is an overrating of the park adjustments... But I find it hard to believe that someone can honestly think that Cabrera has been more valuable than Trout. A gold glove quality centerfielder, exceptional runner, posting similar offensive numbers as a less than average defensive first baseman who has played about 20 games more looks as the more valuable commodity.
We are talking about on raw numbers a difference of .007 avg, .003 obp, and .048 slg. Basically a few more extra base hits that are more than wiped out by the 46-4 sb difference
Edit: or what Snapper said in post 18. (mind you I think the real issue is the park adjustment, something screwy is going on there)
First time I looked at AL Cy Young in about a month, and yes, I don't even see any reason to argue for anyone else over him at this point in time. Sale has 30 innings fewer with a minor advantage in era+(within margin of error at least) And maybe an argument could be made for Hernandez, but that is really the only other person in the conversation.
That's the thing: I never said that Cabrera was more valuable than Trout (this year) but because I questioned the gap — it really does seem overly large — I guess everyone assumed I was on the "Cabrera is better than Trout" side and felt they had to quiet that nonsense immediately.
I think there's probably a lot of pent-up rage from people who watch ESPN or listen to sports talk radio or work with morons, where I'm sure there are people talking all the time about how much better Cabrera is than Trout. But on this board, it seems that even the most pro-Cabrera guys are guys like me, who root for the Tigers, think Cabrera is awesome, and would be happy if he won the MVP. And even we have never said he "deserves" it over Trout. I see the defense, and the SB, and the GIDP. I still think Cabrera's awesome, and always will.
But Porcello over Weaver? Now that's actual nonsense.
Let me make it clear that this is not my choice (I'd vote for Verlander if the season ended today), but do you think Fernando Rodney is in the conversation?
Kimbrel and Chapman are getting a lot of attention in the NL Cy Young race, and in some ways Rodney has been better than them: 0.65 ERA, more saves and a better save percentage, more innings (especially compared to Kimbrel). There's not really a starting pitcher who has separated himself from the pack, though Verlander could still do it with a couple of good starts. Anyway, when there's not an OBVIOUS selection from the starters, that's when closers slip in.
The thing going against him is that he's Fernando Freakin' Rodney, and we're so used to him sucking that no one thinks this is "for real," whereas Chapman and Kimbrel have always been this good.
That is a chant that has been no these boards for close to a decade.
The rage is that the people who are supposed to know these things aren't able to look at how close the numbers are and see that they are basically the same offensive player.
He's in the conversation, not because he's worthy, but because the writers are idiots.
I wonder, knowing what we know about the writers, do you think there's a chance he could actually win?
Let's say Price, Verlander, Sale all get lit up in their last 2-3 starts, and all the other top starters are no better than mediocre. If no one exceeds 18 wins, the ERA leader is around 2.80-2.90, and Rodney gets close to 50 saves (he has 43 now) without giving up any more runs... could idiocy rule the day?
It's 28 writers so it's hard to make a prediction. Here is ESPN's Cy Young predictor(created by Bill James/Rob Neyer) It has Rodney 4th on that prediction. It's arguable that the writers have gotten better and ignoring wins so it might not be as accurate as it could be. I think that it's right now Verlander's to lose.
He's only blown 2 saves, and in one of them, both of the runs he gave up were unearned. He's been scored on in 7 games: the 2 blown saves, 3 successful saves, 1 easy win, and 1 game where the Rays were behind by 2 when he came in. So he only really "contributed" to 2, maybe 3 losses. He has allowed only 2 of 17 inherited runners to score.
I have no idea how he's done it, but he's been really, really good.
Cabrera is the best player in baseball at the most visually appealing position player skill, hitting. If you don't love watching him hit, you don't love baseball.
But the things he does poorly are less clear to the naked eye. He doesn't steal, sure that's visually obvious but base stealing opportunities are not as frequent as plate appearances and hits. He doesn't take extra bases, grounds into a lot of double plays, but again, those don't happen often and aren't something that jump out at most fans. And you'd have to be Ted Williams with enhanced laser surgery to be able to definitely say whether a decent third basemen would have made some of the plays that he just misses. Watching him with a fans eye it's easy to say he's not that bad.
But this is what his bat has produced this year
fRuns: 55.1
bRuns: 53
This is what his base-running & GIDPs have cost this year
fRuns: -2.9
bRuns: -5
This is his offensive value
fRuns: 52.2
bRuns: 48
Now let's look at Mike Trout's bat
fRuns: 52.4
bRuns: 48
His base running
fRuns: 6.1
bRuns: 11
His offensive value
fRuns: 58.5
bRuns: 59
In total offensive value, not a rate stat, this is total actual value that incorporates playing time and without any positional adjustments, Mike Trout is the best offensive player in all of baseball according to both FanGraphs and Baseball Reference. And neither sees it as particularly close.
And the sad fact is, it's obvious Trout has been at least as valuable on offense even if you don't have advanced stats. Miggy has 356 total bases, Trout 290. But total bases doesn't include steals, HBP, or GIDP. If you move stolen bases & HBP into total bases, and caught stealing into GIDP, now Miggy leads 363 to 341 in total bases, but in GIDP 29-11. Including GIDP, Miggy has made 415 outs, while Trout has made 362 outs.
So Miggy produces 363 bases at a cost of 415 outs, while Trout produces 341 bases at cost of 362 outs, and this still shorts Trout any extra bases he took running the bases. Miggy most certainly produced less than 22 extra bases at a cost of about 53 more outs, that's not the type of extra production that makes Miggy more valuable than Trout, it makes him worse.
Given their relative defensive positions and clear skills at each, it's blatently obvious that Trout is way more valuable than Miggy this year, easily more than 2 WAR better and making 4 WAR not only believable, but likely.
If Miggy wins the MVP it probably won't be the most undeserving ever, but on an actual value basis it would have to be one of the bigger discrepancies and Trout would have to have one of the most valuable non-MVP seasons ever. And you don't need advanced stats or WAR to make that glaringly obvious.
What I can't comprehend is that Cabrera's MVP worthiness is based on one HR.
There's an interesting argument to be made for Cabrera involving his moving 3B and doing a good job there. That kind of flexibility is not reflected in WAR. But the argument OMG TEH TRIPLE CROWN is not a good one.
2 war is a pretty large discrepency, 4 war is huge. You are basically saying you take Cabrera and add one of Fielder/Pujols/Hamilton to the mix.
Not really. ERA measures what actually happened(ignoring the unearned run aspect) and assigns full value to the pitcher. Fip says this is what should have happened based upon the number of batters faced, strikeouts and walks allowed, with an average defense. It not only ignores the runs that scored, it also ignores the order of events, and it ignores the actual type of hits(other than hr) that happened. It's multiple levels removed from reality. It's a theoretical model of a theoretical model using some real events as a base.
Why? Future doesn't matter for a value award based upon past events. I mean would we argue against an MVP because he hit an unsustainable .420 babip? Of course not, why should pitchers be any different? Yes I understand you want to remove defense from the equation, but that isn't as easy as fip says it is. I mean if you have a great infield, and you intentionally throw balls that lead to a higher ground ball percentage, shouldn't the pitcher be credited for using his available tools? There is enough evidence to support that pitchers can affect when they get strikeouts and avoid walks which fip doesn't take into account, but era(sequence of events) does. There is also evidence that pitchers do affect some type of specific outs(Infield fly balls is one of them, and again that is a sequence event thing)
Yes, but there are much better measures of actual events than RBI.(average, obp, slg etc) On top of that you can look at avg risp etc to more accurately gauge the players performance. I'm not saying ERA is the perfect evaluation tool, but it's massively better than fip. Fip treats a single and triple exactly the same.
I wouldn't be surprised if there is a voter who decides whichever team makes the playoffs, that will be the deciding factor for him.
In that case, a hitter's MVP worthiness might end up being based on whether a PITCHER can get a save in the final game of the season (see "Game 162" last year) in a game which the MVP-candidate doesn't participate.
You have to ask yourself: why is it predictive?
If you're going to assert we should credit things to pitchers that they might not necessarily be doing but benefit based on the circumstances, it is essentially the same damn thing as looking at RBI.
I agree with fip as a stat for predictive, it tells you the components of a player. I think that fip does a good job of measuring "true talent level" while season long era is including hot/cold streaks that are being smoothed out by fip. Personally I would prefer a component era that looks at actual types of hits allowed converted to runs created and figured from there.
The definition of "valuable" has always been and remains intentionally ambiguous. Neither Trout nor Cabrera's teams are likely to be among the FIVE teams to make the playoffs.
My MVP vote goes to Derek Jeter.
Miguel Cabrera can have the Hank Aaron Award and Mike Trout can have the Jackie Robinson Award.
You have to ask yourself: why is it predictive?
If you're going to assert we should credit things to pitchers that they might not necessarily be doing but benefit based on the circumstances, it is essentially the same damn thing as looking at RBI.
No. We're crediting pitchers for things they did: successfully preventing hits, inducing pop-ups, inducing DPs, controlling the running game, stranding a lot of runners. FIP ignores these things, and, because they're not very stable over long stretches, it makes it slightly more predictive of future ERA than past ERA is.
But, all those things generate real value in a pitching season. I literally don't care if they are even 1% repeatable. A pitcher that strands 90% of baserunners deserves credit for it for awards that recognize that season. FIP is still waiting for Tom Glavine to regress.
And crediting for things their fielders did like make good defensive plays. And not crediting them when they give up runs after their fielders make errors. And giving them credit when a reliever strands their runners. ERA is nowhere near the pure stat that you're making it out to be. It only seems that way because we've been using for so long.
RA is good as well. The point is not that they're perfect, but they at least reflect what actually happened in terms of run prevention.
Of course, the fact that Cabrera plays third base changes things a little.
NOBODY who is championing for era cares about the ####### errors and runs. Count them, don't count them it doesn't make a bit of difference. Why do people ever harp on this issue, who ####### cares about those runs. If RA was more prevalent we would go by that, nobody is saying era is perfect, they are just saying it measure ACTUAL EVENTS something that fip doesn't seem to care about.
Again... In fip, a pitcher that allows 6 doubles over 6 innings is exactly the same as a pitcher that allows 6 singles. (assuming walks and strikeouts are the same)
I'm surprised you forgot 2009.
Strikeouts, walks, and home runs are not actual events?
I keep forgetting that point. And it does (another reason why I think that the 4 war difference is probably overstating the gap)
Sure, but it ignore all the others. Doubles, triples, singles, DPs, pop-ups, etc.
FIP sees the pitcher who gives up three doubles, 2 runs, and K's two, and having a better innings than the pitcher who gets 3 pop-ups.
Hmmm a stat that looks at 29% of the game or a stat that looks at 100%.... tough choice. Even with the flaws with ERA, it covers a larger portion of the game (and yes, argument is that fip could arguably cover 100% of the game if you argue that every other out is exactly equal and that every non-homerun hit is exactly equal)
Well, WAR doesn't forget what position Cabrera plays.
Edit: Nevermind
Ok. What is your point?
FIP = ((13*HR)+(3*(BB+HBP))-(2*K))/IP + constant
Yeah, so those doubles and runs don't count against the first pitcher. His score is ((13*0) + (3*(0-0)) - (2*2))/1 + constant. or (constant - 4). The 2nd pitcher score is just the constant.
ERA ignores those too, and it ignores the other 8 guys on the field. The only thing ERA cares about is earned runs. A pitcher does not ever have to get a batter out to have an ERA of 0.
True, but I don't think Fip is actually counting the guys on the field either. Personally I would prefer component era or something that actually does include type of actual hits allowed.
Ok, but to be consistent I hope you evaluate batters based on what percentage of runners they drive in rather than things like hits and walks.
The same reason you keep harping on the fact that one implementation of the FIP concept doesn't include doubles. You want doubles? Fine, assign a value to it and add it to the formula. It's not hard.
Nobody cares about any specific FIP implementation. We are talking about the concept. The point is that there is value to attempting to zone in on measuring the pitcher's actual performance by teasing out things like fielding. By eliminating the noise, you get a more predictive stat, yes, but you also get a stat that measures the ACTUAL VALUE.
We evaluate hitters by their hits...singles, doubles, triples and homeruns. Not their percentage of driven in. ERA does a better job of listing that result(of course you also look at whip, and other stats etc)
If you include doubles, triples etc in it's implementation, it's not FIP.... It becomes a version of component era, which is a great stat.
That's nonsensical. A batter can't control what happens before or after their PA. We evaluate them by their actual contribution.
Pitching is completely different. Since the pitcher many faces consecutive batters, bunching of hits and walks, and sequencing matters.
Except none of the DIPS stats do that.
BRef attempts that in its pitching WAR. They account for defensive quality. I'm fine with that.
I'm not fine with a pitching value measure that ignores the actual runs that score.
My point is that people seem to get offended when you bring up a pitcher evaluation that ignores runs allowed, yet they have no problem doing so for batters. How many people pay attention to runs and RBI? Should we? If a batter managed to hit .400 with 100 walks would we care that he never scored or drove in a run? Should we?
Edit: The bolded font in [67] is what I'm talking about.
No, it doesn't.
Runs allowed measures "actual events." You could stop right there.
Unfortunately runs allowed doesn't tell you how well a pitcher pitched, it doesn't tell you what VALUE the PITCHER provided.
Plain and simple, it doesn't. You know it doesn't.
It tells you the combined value of the starter, the reliever, the defense, the quality of the offense they were facing, the game conditions, the umpiring, etc.
If you are going to assign credit/blame for runs allowed to individual pitchers you have no choice but to decide how you are going to apportion that value among all those factors.
ERA and RA/9 are purely arbitrary means of assigning that value by giving all the value to the pitcher. That is idiotic on its face unless one of the following two things is true. (a) The pitcher is providing something like 95% of the true value so the rest isn't worth adjusting for. But we know this is false. (b) The rest of that stuff is random so the only "signal" in ERA/RA is the quality of the pitcher's performance -- which, oddly enough, leads you to FIP.
The difference between the two is not their assumptions about what is and isn't random but their decision whether to "reward" that randomness or not.
The less abstract way to look at is that ERA assumes the defense is never at fault (except for errors) while FIP assumes the defense is always at fault. Those seem like equally untenable assumptions to me but if I had to choose one for assessing the quality of a pitcher's performance, I'd probably choose the latter.
bWAR falls in-between I believe, attempting to adjust for defensive quality, thereby blaming the pitcher for "extra" hits given up. Just by virtue of falling in the middle of the "how much do we award randomness" question it almost has to be the most accurate of the three (on a large sample basis ... could be off by a mile for an individual pitcher).
I don't think we should ignore runs and RBI in awards. If a guy happens to have a very clutch season driving in runs, that should count in his favor.
Really? I disagree.
The pitcher is almost certainly closer to 100% responsible than 0%. We know the vast majority of chances are either always outs, or always hits. It's not conceivable that defense could MORE responsible for BIP than the pitcher.
Edit: In other words, there is far more variance in quality between individual pitchers than between team defenses.
I'm fine with adjusting for defense (like BRef does), but you've got to start from the point of view of accounting for all the runs allowed.
That while the narrative description of the difference in value between Trout and Cabrera that you posted implies that you were making an erroneous mental positional adjustment, the WAR difference can't be due to a mistake in positional adjustment. Unless you think that the positional adjustments for 3B and/or CF are just wrong. Was that your point?
Umm... actually, he does. While ERA might be listed as 0.00 for a guy who allows no runs in no innings, it isn't correct to do so, just as it isn't correct to list a BA of .000 for someone who has no hits in no at bats.
I meant something like that he could give up single after single all day, but have all the runners get caught stealing.
Maybe, but probably not.
How many wins do you think a good defensive player at a premium position can add in a season?
How many wins can a bad third baseman cost?
How many wins are an appropriate adjustment for putting up slightly better offense than Cabrera in a defense first position like centerfield?
Its pretty easy to think that Trout can get 4 wins from all three areas, his better offense, his positional adjustment, and their huge gap in defensive value.
Edit: but sinse fWar is only 2.6 apart maybe you should be happy with that?
It's pretty easy to think that, but it's another thing to actually prove that.
The fact that there are redundancies and distortions in the currently favored versions of WAR is enough of a reason to at least lower the level of one's voice when attempting to discuss this matter. (Note to KT: this last reference is not directed at your personally--you ask the absolutely pivotal questions above. The answers you get will, alas, mostly be dubiously applied theory.)
Oh, I see. You're thinking of that little lefty who was on my son's U14 travel team -- couldn't actually pitch to save his life but had an absolutely hellacious move to first. His version of three up, three down was three walks and three pickoffs. Alas, he doesn't play baseball anymore.
Also that excerpt was one of the worst things I've ever read.
It's not just the defense, it's also the batter, ballpark and umpire. Combined that's a huge effect. The only thing the pitcher can do is throw the ball - once he does that, the rest is up to factors completely out of his control. Strikeouts and walks are the best stats we have in a single season sample size at gauging pitcher input because they take up an entire count and are much less prone to random variation, if you rely heavily on hits allowed, you're gonna be putting a lot of noise in it.
I'm with with taking RA and doing a general accounting - look at quality of batters faced, how well relievers stranded runners, quality of defense (not at the team level, but behind the pitcher on that given day), quality of umpire.
Well, what he said was "A pitcher does not ever have to get a batter out to have an ERA of 0." Meaning he could let every batter he faced reach base, and still not give up a run. I understood what he meant.
I'm of course in favor of stadium adj. When I say ERA/RA I really mean ERA+/RA+. Sorry for any confusion.
Over the course of a season, I'd have to think the umpire, and opposing batter factors are not great differentiators among pitchers. You never have all the good hitting teams in one division.
I'm with with taking RA and doing a general accounting - look at quality of batters faced, how well relievers stranded runners, quality of defense (not at the team level, but behind the pitcher on that given day), quality of umpire.
That's a fine approach. It's basically BRef's ERA+; I think the added signal from the batters/umpires faced is going to be pretty tiny.
There is much less variation on the quality of a pitch vs. the quality of the batter outcome. Think about it, a pitcher will throw a fastball in the meat of the strikeone zone 10-15 times per game, and sometimes he gives up 7 hits and a home run and sometimes he gives up 4 hits and no homers. In 30 starts, that variation from the offense is still very large. Of the three factors that drive BABIP - pitching variation, defense variation and offensive variation - offense is by far the largest.
Strikeouts and walks are more stable because they occur over 4-5 pitches per at bat, rather than just one pitch for batted balls. And they don't put the defense into the equation, or the ballpark as much as with batted balls. So even though there are more BIP in a given season than strikeouts and walks, strikeouts and walks are going to be much more reflective of pitcher performance.
So I think combining xFIP (or even just kwERA, which is only strikeouts and walks) with bWAR is the best thing you can do. Maybe 60% bWAR and 40% kwERA. That's just simply giving more weight to the strikeout and walk portion of bWAR.
How can anyone determine the quality of defense on a given day? Is UZR or TZ on a per game basis available or meaningful? What would you do with the relievers stranding runners....adjust to some average reliever strand rate? Aren't pitchers having good days typically relieved by better relievers? How would you adjust for that? More broadly, what is the purpose of this? Are you trying to say that pitcher BABIP and Strand Rate is a skill for starting pitchers or that it doesn't matter whether it's a skill or not and we're just trying to measure what took place? In the first instance, isn't that provable or disprovable using regression analysis? In the second instance, why make any adjustments at all? Why not just go with Innings Pitched and RA+?
Exactly. If Howard drives in a lot more runs than Pujols, it's worth considering. One might find that Howard was a much better hitter, in which case (if defense and baserunning are a wash), the conversation is over. But one might also find:
(a) Pujols is a much better hitter than Howard, and would have driven in a lot more runs given Howard's much more numerous opportunities (MVP to Pujols)
(b) Pujols was a much better hitter than Howard, and had the same opportunities, but he got IBB'd so much because the team behind him was so weak, that he didn't drive in nearly as many runs (in which case one has to weigh the value of the IBBs, and settle the issue accordingly)
(c) They're even as hitters, but Howard had more opportunities (in which case toss a coin, or vote for a relief pitcher :)
(d) They're even as hitters and had equal opportunities, in which case Howard gets the MVP because dang it, he actually did stuff, as opposed to establish an even theoretical chance of doing it
And FIP ignores other aspect of run prevention at least partially under a pitcher's control. Popups/foulouts are almost 100% outs and should go as pure pitcher credit. The pitcher plays a huge role against the running game. The pitcher plays a huge role in DP frequency.
And not of least importance there's a big difference in expected runs allowed given balls in play, BB and home runs allowed and actual runs allowed. The standard error for a pitcher who pitched 200 innings has to be in the range of 10 runs. (IOW Fangraphs pitcher WAR has a standard error of something close to a win for a full-time starter)
I used to start from support neutral won/loss (SNWL). BP has changed what it's called and how it's derived but it's still a good start to the discussion since it considers not just the runs allowed but the distribution of those runs. (And deal with runners left on base in a way I like -- they're counted according to the expected runs allowed so bullpen support isn't an issue)
FIP sees the pitcher who gives up three doubles, 2 runs, and K's two, and having a better innings than the pitcher who gets 3 pop-ups.
FIP isn't intended to be used on one-inning sample sizes.
I think adjusting ERA/ERA+ for the quality of the defense when we know that there are differences in the defense makes sense. It's more complicated than a simple FIP calculation. There also needs to be differences in the stat if you're using it for predictive purposes or value purposes. You'd regress K, BB, and HR in addition to BABIP for predictive purposes, but you would want to leave the actual events as is when looking at historical value.
The portion of hits allowed that can't be easily attributed to the defense or the pitcher ("luck") is the toughest part. I'm inclined to be conservative and attribute it to the pitcher if we don't know any better.
And yet, past FIP predicts future ERA better than past ERA predicts future ERA. FIP is not the perfect indicator of a pitcher's performance, but it is better than ERA in finding his true talent.
This doesn't follow at all. The terms "sustainable" and "true" are not synonymous.
Right, but the Cy Young isn't supposed to reward true talent. It rewards actual performance in that year.
Right, but most people who use ERA+, or RA+, also look at Ks, WHIP, and K/BB ratios (as well as IP and wins). HRs directly show up in runs allowed.
FIP would be a better stat if it wasn't scaled to ERA, b/c it wouldn't be masquerading as a replacement.
Several times more removed than is warranted or necessary. The things that happened are HRs, SOs, and BBs. FIP did not "happen" in any real sense of the word.
Right, but the Cy Young isn't supposed to reward true talent. It rewards actual performance in that year.
It's hard to believe such an obvious thing needs to be repeated. "True talent," even if it could be measured accurately (which the commonly-proffered measurements don't) is of secondary or tertiary importance in baseball -- and that's being kind.
The point of today's baseball game is not to predict what will happen in tomorrow's baseball game.
This is true for all composite stats, down to OBP and SLG.
Yes, but you can also look at both ERA+ (or RA+) and FIP. There's no reason one has to be a "FIP person" or an "ERA person". I don't like FIP on its own, but it still contains a lot of information.
I'm not quite sure why we haven't moved to a linear weights (or weighted linear weights where H, 2B and 3B are scaled down) system for pitchers yet except that the raw data's not ubiquitous.
Actually, no. OBP and SLG aren't "scaled" to anything and there's no "13" or other multiplier.(*) And they make sense on their own terms in a way that FIP doesn't approach.
FIP is like multiplying gas mileage and the square of engine cylinders and pretending you're measuring something important about cars.
(*) Yes, I know doubles are multiplied by 2 to get SLG, but there's an underlying reason for that that isn't present with FIP.
Not exactly; FIP specifically is scaled in a distorted way to make it appear as something it is strictly not, namely an ERA equivalent. This is because the original DIPS theory was for ERA predition. Also, it is reverse engineered in a non-intuitive formula with superficially arbitrary weights for a specific purpose, unlike OBP and SLG which are ostensibly just measuring something basic in a very basic way.
You must be Registered and Logged In to post comments.
<< Back to main