Fun With Numbers

by Jeff Sullivan on Oct 8, 2009 3:46 PM PDT in Seattle Mariners Stats

The six relievers who got the most playing time for the Mariners this year were David Aardsma, Sean White, Shawn Kelley, Miguel Batista, Chris Jakubauskas, and Mark Lowe. Here's how they did pitching to Rob Johnson and Kenji Johjima (Jak's starter stats are included because I didn't feel like separating them, but he conveniently threw to each catcher about equally often as a starter).

	Johnson				Johjima
	uBB	K	HR	PA	uBB	K	HR	PA
Aardsma	16	48	2	151	11	21	1	104
White	6	8	1	125	11	14	2	100
Lowe	9	32	3	153	12	26	3	130
Batista	15	17	1	106	21	30	6	175
Kelley	1	9	3	61	6	26	5	110
Jak	11	20	9	177	12	19	5	180
TOTAL	58	134	19	773	73	136	22	799
RATE	7.5%	17.3%	2.5%		9.1%	17.0%	2.8%

You'll see that, over remarkably similar sample sizes, the relievers walked fewer batters when throwing to Johnson than when throwing to Johjima. The other two stats aren't meaningfully different.

But wait! That's not a fair comparison, because the playing time distributions aren't the same for each catcher. For example, Johnson got to catch 50% more Aardsma, while Johjima had to catch 60% more Batista. What happens when you weight Johjima's numbers to equal Johnson's PT distribution?

Johnson: 7.5%, 17.3%, 2.5%
Johjima: 9.3%, 16.8%, 2.4%

There you go. Neither the strikeout nor home run rates are meaningfully different, and the walk rates are within two standard deviations of each other, but something to think about.

Tweet Comment 21 comments | 0 recs |

Do you like this story?

Comments

Are the numbers for each individual relief pitcher from too small of sample sizes to mean anything significant?

by Dewey N on Oct 8, 2009 5:00 PM PDT reply actions

Probably

Let’s take the biggest sample in there – Joh catching Batista. You’ve got a 17.1% strikeout rate with a standard deviation of +/- 2.8%, or 5 strikeouts.

by Jeff Sullivan on Oct 8, 2009 5:22 PM PDT up reply actions

Doesn't factor in bad umpiring, pitcher's ability to hold runners on, catcher's ability to throw guys out

Umpiring: What if one of them just happened to be in games with a home plate umpire that called low strikes / outside corners correctly? What if one of them was in games with CB Bucknor more often than the other? That might be masking one catcher’s deficiencies.

Holding guys on / throwing them out: Some of those pitchers were pitching more often with baserunners they’re responsible for, or did better in situations with inherited baserunners. Wouldn’t that have a ripple effect on pitch selection, and thus walk or strikeout rate?

by Spoomeister on Oct 8, 2009 5:18 PM PDT reply actions

In conclusion this table doesn't cure cancer

by Jeff Sullivan on Oct 8, 2009 5:20 PM PDT up reply actions 2 recs

Do you put together tables like this, looking at parts of the team that you don't publish?

I want to say doodling around to satisfy curiosity, but that phrase sounds really trite and while it’s an apt description for the image in my head this is a serious question. Not important, but perhaps you could indulge my curiosity.

by Kermit. on Oct 8, 2009 5:51 PM PDT up reply actions

Probably half of the things I spend at least half an hour on with Excel don't get published anywhere

Mainly just indulging my curiosity.

by Jeff Sullivan on Oct 8, 2009 6:19 PM PDT up reply actions

Perhaps I should've used a different expression

by Jeff Sullivan on Oct 8, 2009 6:23 PM PDT up reply actions

After three re-writes I gave up on my comment. I hated it but every other version sounded like a demanding little shit.

I’ve wanted to ask that question quite a few times, thanks for the answer.

by Kermit. on Oct 9, 2009 12:24 PM PDT up reply actions

Even right now I'm playing around with something that isn't going anywhere!

by Jeff Sullivan on Oct 9, 2009 12:59 PM PDT up reply actions 1 recs

so many jokes, so little desire to be permabanned

Nice Guys Finish Third - My semantics are a waste of time.

by pdb on Oct 9, 2009 1:37 PM PDT up reply actions

I know!

by Jeff Sullivan on Oct 9, 2009 2:31 PM PDT up reply actions

I want a chart that illustrates that Nolan Ryan's supposed "no pitch count" edict had nothing to do with Texas' success this year.

by msb on Oct 9, 2009 8:00 AM PDT up reply actions

And you want us to take it seriously?

by Dewey N on Oct 8, 2009 6:09 PM PDT up reply actions 1 recs

Blah, no slight intended - just the first things I wondered about

Especially umpiring. I’d love to see some standardized stats that let you figure out UIP (umpire independent pitching).

by Spoomeister on Oct 8, 2009 11:52 PM PDT up reply actions

Ideally those things would average out but over sample sizes this small they probably don't

That’s why I caution that, while this is interesting, it’s not nearly conclusive.

by Jeff Sullivan on Oct 9, 2009 12:42 AM PDT up reply actions

Can you post one with WP and PB for Johjima and Johnson by pitcher?

I can’t find a decent source for that kind of data. I’m curious of Johnson’s “catching” is as bad as we all think it is.

From B-R.com (Wild pitches for 2009 aren’t listed yet.):

Johjima: 26 past balls in 3692.2 innings (2006-2009). That’s 142 innings per past ball. Johjima: 102 WP in 3112.2 innings(2006-2008). That’s a WP every 30.5 innings.

Johnson: 9 PB in 754.1 innings (2007-2009). That’s one every 84 innings caught. 4 PB in 68 innings. That’s one every 17 innings with a hugely limited sample size.

by CMC_Stags on Oct 8, 2009 10:49 PM PDT reply actions

Unfortunately I don't have that info

Limited by what B-R has in the splits.

by Jeff Sullivan on Oct 9, 2009 12:45 AM PDT up reply actions

And no one wants to have to delve through each box score

I can go through each game’s box score and try to find it, but don’t have the time. I was just hoping you had a DB you could query against to find it.

I’ve got to think that the path to measuring both catcher defense and game calling will come back to that level of data through Tango’s WOWY process.

The walk rates you posted are a great start for the game calling portion. But without being able to have a good source for throwing, catching, and defensive splits, doing the defensive portion would be difficult to impossible. This is probably why people much better at this than me haven’t figured out a good way to measure it yet.

by CMC_Stags on Oct 9, 2009 9:30 AM PDT up reply actions