Fangraphs: The Orioles Don't Care About Our Expectations

[SilentJames]

FIP is a stat for simpletons. It just uses home runs, walks and strikeouts to judge a pitcher. If that was so meaningful why don't they use it to rate batters? It is for people who don't understand statistics.

What? I mean, what?

[atomic]

I think it's primarily due to its (relatively) high correlation with future performance. However, it's worth noting that the reasons for this correlation are (as far I as know) not well understood. It's not supposed to tell you how a pitcher is performing right here and now. ERA does that, but ERA also is very context dependent, which makes it a bad predictor of future performance (parks, defense, etc.)

Why not just use OPS against with a defense and park adjustment? Wait that would require someone to do actual statistical calculations of some complexity.

[SilentJames]

Why not just use OPS against with a defense and park adjustment? Wait that would require someone to do actual statistical calculations of some complexity.

but defensive metrics are still in flux, where as the things that make up FIP are not subjective, Ks, BBs, and HRs - the only things that a pitcher has direct control over. Why does it need to be "complex"? Why do you need complexity for the sake of complexity? What does your method do that FIP does not?

[El Gordo]

It seems to me that even the SP with the best FIP still end up getting 2/3 of their outs on balls in play. So either they are lucky, have a great defense behind them, or there is a skill to getting batters to put balls in play weakly. Like cans of corn FB, or routine GB. A lot of pitchers as they age and lose velocity, seem to stay effective by changing speeds, locating, and having good movement on their pitches. They don't miss as many bats but they consistently miss the barrels of those bats. I don't see FIP reflecting that skill in it's measurement. Maybe that skill is just a myth?

[VeveJones007]

It seems to me that even the SP with the best FIP still end up getting 2/3 of their outs on balls in play. So either they are lucky, have a great defense behind them, or there is a skill to getting batters to put balls in play weakly. Like cans of corn FB, or routine GB. A lot of pitchers as they age and lose velocity, seem to stay effective by changing speeds, locating, and having good movement on their pitches. They don't miss as many bats but they consistently miss the barrels of those bats. I don't see FIP reflecting that skill in it's measurement. Maybe that skill is just a myth?

Greg Maddox once told John Smoltz that he was going to get a batter out on a fly ball to the warning track and that was exactly what occured. While that's certainly the extreme case, I believe pitchers can have a degree of control over the contact they induce.

[SilentJames]

There will always be a gap between FIP and ERA. But what FIP measures is the stuff that a pitcher can completely control - BB, K and HR.

It is no secret that a pitcher that tends to strike a lot of players out, not walk a lot of people, nor give up a ton of homers - is usually a pretty dang good pitcher. But a pitcher with a low ERA but has a low ERA, walks more than he should and gives up a greater number of homers than average, that tells us that the pitcher is likely primed to see their ERA jump.

The Orioles however, can take that FIP and chalk a portion of that discrepancy up to their superior defense. If the Orioles did not have such a strong defense, the Orioles' pitchers would not have such a solid ERA.

[atomic]

but defensive metrics are still in flux, where as the things that make up FIP are not subjective, Ks, BBs, and HRs - the only things that a pitcher has direct control over. Why does it need to be "complex"? Why do you need complexity for the sake of complexity? What does your method do that FIP does not?

Because it doesn't tell you how good a pitcher is pitching. Of course the same pitcher will give up more home runs pitching for the Orioles than the Mets. Different stadiums, different competition. I don't care if the pitcher strikes out the batter or the guy hits a pop up to third.

There is a difference between a pitcher who is giving up a ton of doubles and the guy who is getting a ton of ground balls. To think there is not is comical. Mark Buerhle doesn't strike out a lot of guys and gives up a decent amount of home runs, yet he seems quite effective.

And a stats has to be complex enough to meaningful.

[flyhighbirds]

I am kind of old fashioned and don't pay attention to these modern stats. I watch just about every game and use the eyeball test. I catell is a hitter is hitting the ball hard but having bad luck and so on. I know what obp is. Somebody tell me why I need to pay attention to these modern stats? I would really like to know

[Loophole]

Because it doesn't tell you how good a pitcher is pitching. Of course the same pitcher will give up more home runs pitching for the Orioles than the Mets. Different stadiums, different competition. I don't care if the pitcher strikes out the batter or the guy hits a pop up to third.

This is why we have things like xFIP and park/league factors that go into fWAR. FIP is meant to be as simple as it can be because not every stat has to be all-encompassing and people using stats shouldn't be so ignorant to ignore every other stat that might be affecting it. People like FIP because they know and understand the ingredients while something like SIERA is harder to grasp at a base level. Different needs call for different stats and only fools and trolls think every one has to tell us everything. Meanwhile, it's proven to be more predictive than pretty much everything else, as things like batted ball data can take multiple seasons to become reliable while walks/Ks/HRs don't.

A pitcher who is giving up a lot of doubles can suddenly stop giving up a lot of doubles when his luck readjusts. A lot of doubles come from ground balls down the line.

Mark Buerhle is frequently not very good. FIP-based fWAR still likes him.

We don't use FIP for hitters because individual hitters have much more control over their batted ball profiles. We still use Ks/BBs/HRs more than most other statistics to judge them. .AVG/.OBP/.SLG are influenced by those individual stats more than anything else. And I don't know about others, but the first stats I look at when I want to see how a hitter hits are K%, BB% and ISO.

[Hallas]

Why not just use OPS against with a defense and park adjustment? Wait that would require someone to do actual statistical calculations of some complexity.

RA9-WAR (the baseball reference version) does something like this. However, Dave Cameron has commented that he doesn't like it because, as people have noted, people aren't being consistent with how they are applying fielding modifiers. Furthermore, the data behind even advanced fielding metrics is very subjective in many cases.

The data from MLBAM could fix this issue, but I suspect that is a long way from here.

[awannabegeek]

Why not just use OPS against with a defense and park adjustment? Wait that would require someone to do actual statistical calculations of some complexity.

I'm not sure I get the rationale here. You want to use a (simple) stat that weighs two statistics cumulatively as if they provide equal value as the basis of a stat, but people that use FIP are "simpletons"? FIP is meant to be simple, not perfect, and has shown pretty good predictive capabilities, with some outliers here and there.

Dismissing things with terms like "simpleton" is usually a sign of...well...a simpleton.

[waroriole]

There will always be a gap between FIP and ERA. But what FIP measures is the stuff that a pitcher can completely control - BB, K and HR.

.

Complete control is overselling it a lot, especially if catcher framing data is to be believed at all.

[DrungoHazewood]

FIP is a stat for simpletons. It just uses home runs, walks and strikeouts to judge a pitcher. If that was so meaningful why don't they use it to rate batters? It is for people who don't understand statistics.

This may be among the top ten posts in the history of the Hangout. Rarely does a statement so delicately walk the line between sarcasm, satire, misunderstanding, and willful ignorance.

[Phantom]

FIP is a stat for simpletons. It just uses home runs, walks and strikeouts to judge a pitcher. If that was so meaningful why don't they use it to rate batters? It is for people who don't understand statistics.

You're 100% right. Dave Cameron, the life blood of FanGraphs (and Felix Hernandez advisor) should really brush up on his statistics.

Why not just use OPS against with a defense and park adjustment? Wait that would require someone to do actual statistical calculations of some complexity.

It's almost like hitters have more control over where the ball goes off the bat. Imagine that.

It seems to me that even the SP with the best FIP still end up getting 2/3 of their outs on balls in play. So either they are lucky, have a great defense behind them, or there is a skill to getting batters to put balls in play weakly. Like cans of corn FB, or routine GB.

Don't let the perfect be the enemy of the good. Are some pitchers able to get more in play outs as a skill? Probably, but let's not outliers ruin the whole statistic set.

A lot of pitchers as they age and lose velocity, seem to stay effective by changing speeds, locating, and having good movement on their pitches. They don't miss as many bats but they consistently miss the barrels of those bats. I don't see FIP reflecting that skill in it's measurement. Maybe that skill is just a myth?

I think it's mostly a myth. I doubt that a pitcher (or a statistically significant amount of pitchers) can confuse hitters enough to miss their barrels, but not enough to miss the bat entirely. Generally, if you are keeping hitters off balance, even with slower stuff, it will show in Ks. These pitchers also tend to lower their walk and HR rates, which are counted in FIP.

[Phantom]

Complete control is overselling it a lot, especially if catcher framing data is to be believed at all.

Very good point. Level of competition is also a huge problem here.

For those who want to know more, here is a great article on the differences between Fip, xFip and SIERRA:

http://www.fullspectrumbaseball.com/2012/08/03/sabermetric-mining-fip-xfip-and-siera/

Which Is Best?

When SIERA first came out, Matt Swartz at Fangraphs tested the different DIPS metrics to see which was the best predictor of future ERA. If you are a math person, I suggest checking out the article for the specifics, but allow me to summarize his results as they apply to using these stats to predict the next year’s ERA:

1) SIERA

2) xFIP

3) FIP

The gap between SIERA and xFIP is relatively small, while the gap between xFIP and FIP is more pronounced. Basically, using xFIP or SIERA will both be helpful, while using FIP will help to a lesser degree.

With that said, when it comes to predicting rest of season ERA, it should be noted that FIP will do better than these tests show. This is because FIP does not normalize for home run rate, which may be in part impacted by ballpark factors. While xFIP will apply a normal home run rate for pitchers, a pitcher pitching in, say, Petco Park may exhibit a low HR/FB for the rest of the season, making FIP slightly more valuable than in end-of-season analysis.