I need an analytics refreshment course lol

What's the better stat, FIP or xFIP?

 

Also, we couldn't even get a lottery ticket prospect for Lindstrom last year. I'm not saying he's a horrible reliever, but he's not a closer and that's not just based on his performance so far this year.

 

Lindstrom's mediocre, he won't net anything great.

 

And they're more or less the same stat except xFIP normalizes home runs based on fly balls. On average, 11% of fly balls are HRs. So instead of counting home runs, 11% of all fly balls the pitcher gives up (outside of infield flies) are counted as home runs in the xFIP calculation.

 

FIP is a better indicator for that year and maybe the next year while xFIP is a better indicator for a year or two in the future.

Edited April 22, 201411 yr by chw42

What's the better stat, FIP or xFIP?

I prefer xFIP. It corrects guys getting lucky with HR/FB percentage. Although it's not perfect because some guys can sustain relatively low HR/FB. But it is more predictive long-term than FIP. The best of all is SIERA I think, not only is it a bit more predictive than xFIP, but it tells us more about the complexity of pitching. It doesn't ignore balls in play, but does attempt to correct them.

Edited April 22, 201411 yr by Feeky Magee

Forgive me, what's FIP?

 

Fielding Independent Pitching. Basically, it's ERA based on only the "true outcomes" of pitching -- HR, BB, and K. It's an attempt to strip the effects of defense and batted ball luck from pitcher evaluation.

 

It's typically used to identify regression candidates. If a guy's ERA is drastically lower than his FIP, he's probably getting lucky from good defense, conveniently placed balls in play, and/or fortunate hit sequencing, like if all the HRs he gives up are with the bases empty, or he strands way more runners than usual. If a guy's ERA is drastically higher than his FIP, the opposite is true and he's a good candidate to start seeing much better results given the same input.

 

Over time, FIP has shown to be among the better predictors of statistical regression on pitchers, but there are exceptions -- guys that seem to consistently post FIPs higher or lower than their ERAs year after year. The guys that "beat" FIP and defy negative regression are those that seem to have a knack for inducing weak contact, like Bronson Arroyo, or those that seem to have an inhuman ability to never give up homeruns, like Matt Cain. The guys on the other end are those that never seem to live up to their potential, like Ricky Nolasco or Bud Norris, or those that are good but really seem like they should/could be elite on better teams, like Zack Greinke or Javy Vazquez.

What's the better stat, FIP or xFIP?

 

Also, we couldn't even get a lottery ticket prospect for Lindstrom last year. I'm not saying he's a horrible reliever, but he's not a closer and that's not just based on his performance so far this year.

 

They're just different. FIP is a descriptive statistic, xFIP is a predictor. Or, another way: FIP is what happened, xFIP is what SHOULD have happened. The major difference in the calculation is that xFIP regresses HR rate, which is by far the most unstable factor involved.

It's cool that we have a new Matt The Bullpen Guy Who Everyone Abuses Constantly Despite His Statistical Decency so soon after Thornton left.

 

Do you want a mulligan on this comment? He's been horrible. But we did win the game so I'm very very happy and give him credit for the save.

Do you want a mulligan on this comment? He's been horrible. But we did win the game so I'm very very happy and give him credit for the save.

Horrible this season or for the White Sox or for his career?

The Sox seem to get a lot of games on the U. Maybe because not a lot of people watch.