Tribe relievers and HR/FB

by Peter Bendix on Jun 20, 2008 11:13 AM EDT

Indians relievers have given up an inordinately high amount of homers per fly balls so far this year.

Out of American League relief pitchers who have pitched at least nine innings this year, how many Cleveland Indians do you think are in the bottom 22 relievers, in terms of homers per fly ball? In other words, out of the 22 relievers who have given up the most homers per fly ball this year, how many are Cleveland Indians?

Well, there are 14 teams in the AL. So, considering that any deviation from the norm of around 11-12% homers per fly ball for a pitcher is generally due to luck, we would expect at least one, and perhaps two, Indians relievers to be in the bottom 22.

In actuality, FIVE Indians relievers are in the bottom 22.

Rafael Perez has given up homers on 25% of his fly balls, Edward Mujica on 16.7%, Joe Borowski on 15.8%, Jorge Julio on 15.8%, and Rafael Betancourt on 14.3%.

In comparison, Minnesota has three relievers in the bottom 22, and no other team has more than two.

As the amount of homers a pitcher gives up per fly ball fluctuates due to luck, it is reasonable to assume that these numbers will regress, and, as they do, these relievers will give up fewer runs. Hopefully this regression will play itself out over the course of the rest of the season, and our bullpen will improve.

0 recs | Comment 16 comments

Read Related

Story-email Email Printer Print

Comments

I know that track records matter and that over a long period of time, players tend to perform in consistent ways. But, I think one can take that “regression to the mean” argument too far. Players also have bad years and bullpens are notoriously inconsistent (as everyone keeps pointing out). Beyond that, curing the gopher ball problem won’t do away with the fact that the bullpen is horrible in BAA and in WHIP. Too many baserunners is a problem, whether or not you’re giving up home runs.

by peter m on Jun 20, 2008 11:21 AM EDT 0 recs

These metrics of BABIP and HR/FB are far from perfect. They begin with the assumption that each hitter or pitcher gets or gives an avg amount of wood on each ball put in play. A hitter who puts the ball in play but fails to hit with any authority will carry a low BABIP. A pitcher who isn’t fooling hitters will end up allowing too much wood on the ball and the ball carries further, over the fence. This is how pitch-to-contact pitchers make a living. Opposing batters put the ball in play, but fail to get a lot of wood on the ball.

by elsandito on Jun 20, 2008 12:31 PM EDT 0 recs

Intuitively, you seem right. But research has proven that this is not so. BABIP is not perfect – ground balls are more likely to become hits than fly balls, for example, and thus ground ball pitchers have a higher BABIP. And of course, the defense behind a pitcher matters.

But research has shown, over and over, that very few, if any, pitchers can consistently keep their BABIP below average (and by very few, I mean like six non-knuckleball pitchers over the last 30 years).

Additionally, homers per fly ball is a stat that stabilizes quickly, and very few pitchers demonstrate a consistent ability to keep this below normal. Thus, any deviation from the mean is due mostly to good or bad luck.

Rafael Betancourt is a perfect example. His BABIP, starting with this year and dating back to 2004:
.351
.240
.273
.296
.346

His HR/FB, starting with this year and dating back to 2004:
14.3%
4.7%
9.4%
8.0%
11.4%

What’s remarkable is that, essentially, Betancourt has been the same pitcher throughout his whole career. Here is his strikeouts/game and walks/game, starting with this year and dating back to 2004:
9.1/2.4
10.7/1.2
8.0/1.8
10.3/2.4
10.4/2.5

by Peter Bendix on Jun 20, 2008 12:52 PM EDT up 0 recs

“But research has shown, over and over, that very few, if any, pitchers can consistently keep their BABIP below average (and by very few, I mean like six non-knuckleball pitchers over the last 30 years).”
That’s pretty amazing, statistically speaking, that only 6 pitchers have varied from average on the low side. I guess about 6 pitchers have been above the average, then. The balance would be exactly average.

by elsandito on Jun 20, 2008 2:49 PM EDT up 0 recs

I have exaggerated by saying 6 pitchers, but I do not believe the number is much higher than that. The point remains that very few pitchers have demonstrated a consistent ability to supress hits on balls in play.

Tom Tippett wrote a great article about this, I recommend it highly: http://www.diamond-mind.com/articles/ipavg2.htm

by Peter Bendix on Jun 20, 2008 3:37 PM EDT 0 recs

I’m not convinced that this article’s conclusions support your assertions. You are focusing on the point that year to year fluctuations can mask pitchers’ true ability to suppress hits on balls in play. It doesn’t, however, suggest that Indians’ relievers are mostly better than this year’s showings in this category. Another point in the article is that it shows, for successful pitchers exampled, there is a shape to pitchers’ success curves for this stat, shows improvement toward a playeer’s prime and a subsequent degradation thereafter. I believe this supports the idea that a pitcher’s ability to suppress hits on ball in play, is very real and is a function of my previously stated variables.
Another way to think of this is: Think of the probability that a ground ball has of becoming a hit, as it is directed somewhere between the SS and 2B. The probability rises as the velocity of the ground ball rises. A weakly hit grounder won’t have much chance to get through. The pitcher’s ability to avoid wood determines the velocity of the ground ball.

by elsandito on Jun 20, 2008 4:12 PM EDT up 0 recs

Well, we have diverged from my original point about HR/FB, but I am interested in discussing BABIP.

First of all, here is Betancourt’s line-drive percentage, with his accompanying BABIP, starting with this season:

21.9% – .346
19.7% – .296
25.1% – .273
22.2% – .240
18.6% – .351

I believe that there is a good chance that you are at least somewhat correct – pitchers can control, to some extent, how much wood a hitter gets on a ball. However, I think a lot of that “control” shows up in a pitcher’s line-drive percentage. Furthermore, I think – especially for relief pitchers – that the sample sizes are much too small to be able to tell anything definitively.

Let’s say, for argument’s sake, that we KNEW that most pitchers had a BABIP of .300. And that we KNEW that Rafael Betancourt, for whatever reason, had a BABIP of .280. We’re talking about 2 hits per every 100 at bats. And because relievers only pitch 60-70 innings a year, we’re talking about extremely small sample sizes. Thus, even if we KNEW that Betancourt had an ability to prevent hits on balls in play, there’s a very good chance that the data would not support this, due to the small sample size.

Furthermore, we have evidence that Betancourt’s inherent skills have not declined, as judging from his walk and strikeout rates – both of which are directly in line with what we would expect, given his career averages coming into this season.

Thus, I believe that my original point – namely, that Betancourt’s high BABIP is mostly a result of poor luck and therefore is likely to regress – remains.

by Peter Bendix on Jun 20, 2008 4:43 PM EDT 0 recs

Oops, I meant to post that as a reply…my bad.

by Peter Bendix on Jun 20, 2008 4:43 PM EDT up 0 recs

Your original point was about the entire relief staff as a group and their FB/HR ratios as individuals. FB/HR ratio, just as BABIP, is about suppression. Betancourt may be one of only two on our list whose numbers are a victim of luck. We don’t know whether most of these guys are deteriorating, bad in the first place or experiencing bad luck. Their base sample size is small prior to this year and their sample size for this year is small. Suggesting that there’s a regression to some meaningless mean of small sample size, is self serving (homerfanwise).

by elsandito on Jun 20, 2008 6:29 PM EDT up 0 recs

Judging by what research has shown about nearly all pitchers – namely, that their BABIP should be around .300, with some difference based on the quality of the defense behind them – as well as from Betancourt’s career stats, I think it’s fair to say that his BABIP is abnormally high and is unlikely to continue to be that high for the rest of the season.

I think the same thing is true for Betancourt’s homer rate, as well as the homer rate for the other Indians pitchers I mentioned.

I do believe, of course, that pitchers can get worse, but this often shows up in what I consider to be “skills” (such as walk rate, strikeout rate, velocity, amount of swinging strikes, etc), rather than what I consider to be “outcomes” (amount of hits given up, amount of homers given up, heck even amount of runs given up). If we have reason to believe that a pitcher’s skills are unchanged, then we have to reason to be skeptical of a big change in outcome.

by Peter Bendix on Jun 23, 2008 12:20 PM EDT up 0 recs

From a recent gamethread:

well, raffy. just be good.
by Brick. on Jun 11, 2008 9:17 PM

Whooaaa slow down. I’ll regress to my career numbers if you do too!
by Fire Slider on Jun 11, 2008 9:18 PM

I think he’ll be good again once the homeritis wears off. Really, I feel fine about Raffy R
by jhon on Jun 11, 2008 9:18 PM

Thank you for making the strong analytic case here. We were watching him pitch, and stuff/location wise he looked every bit as good as the Raffy we’ve known. Minutes later Betancourt had pitched an inning, struck out two and not allowed any runs. It confirms what I’ve also noticed. He actually still is very good; that he hasn’t collapsed in the way that some of our other stars have.

by jhon on Jun 20, 2008 6:19 PM EDT up 0 recs