Pitch Calling

by Jeff on Jan 20, 2009 8:33 PM PST in The More You Know Comment 25 comments

Sometimes when I get to playing with spreadsheets, my interest is piqued by a particular question and I take on a project without really knowing where it's going. This is one such project. A few days ago, when I was working with the Felix data, I found myself wondering "how did he pitch with different guys behind the plate?" Not so much in terms of results, but in terms of pitch sequencing. Who was calling what pitches in what amounts? To satiate my curiosity, I went to Baseball-Reference, figured out who was behind the plate for each of Felix's starts, put all the numbers together, and one thing led to another and now I'm here with a whole bunch of numbers for six different starters and at this point I'm not really sure what they mean. So I'm going to dump them on you.

I think the tables are pretty easy to understand. For each pitcher, I've broken the numbers up by catcher and by batter handedness. I'm sorry for what this post may do to your eyes.

			n	FA%	CH%	CU%	SL%
FELIX	Righties	Burke	138	71%	7%	5%	17%
		Clement	177	64%	11%	6%	19%
		Johjima	834	70%	8%	7%	16%
		Johnson	108	70%	5%	4%	21%
	Lefties	Burke	155	60%	21%	8%	11%
		Clement	264	66%	17%	9%	7%
		Johjima	1318	62%	20%	10%	8%
		Johnson	81	65%	10%	16%	9%

			n	FA%	CH%	SL%
SILVA	Righties	Burke	118	74%	14%	12%
		Clement	356	66%	15%	19%
		Johjima	583	73%	11%	15%
	Lefties	Burke	109	64%	31%	5%
		Clement	480	70%	24%	6%
		Johjima	781	68%	29%	4%

			n	FA%	CH%	CU%
BEDARD	Righties	Burke	731	63%	4%	34%
		Clement	94	70%	0%	30%
		Johjima	162	60%	1%	39%
	Lefties	Burke	219	73%	0%	26%
		Clement	35	69%	0%	31%
		Johjima	102	64%	4%	32%

			n	FA%	CH%	CU%	SL%
WASHBURN	Righties	Burke	227	74%	8%	17%	0%
		Clement	375	66%	25%	9%	1%
		Johjima	1282	73%	19%	7%	1%
	Lefties	Burke	38	63%	3%	3%	32%
		Clement	112	66%	1%	4%	29%
		Johjima	539	58%	3%	7%	32%

			n	FA%	FC%	SL%	CU%	CH/SF%
BATISTA	Righties	Burke	19	58%	32%	5%	5%	0%
		Clement	107	64%	27%	1%	7%	1%
		Johjima	570	52%	36%	4%	8%	0%
	Lefties	Burke	76	64%	33%	0%	1%	1%
		Clement	145	58%	34%	0%	1%	8%
		Johjima	760	54%	36%	1%	2%	8%

			n	FA%	CH%	CU%	SL%
RRS	Righties	Clement	180	63%	18%	13%	7%
		Johjima	518	52%	27%	17%	3%
		Johnson	151	56%	19%	23%	2%
	Lefties	Clement	75	68%	9%	15%	8%
		Johjima	199	43%	2%	19%	36%
		Johnson	54	57%	4%	17%	22%

I only looked at starts, so none of RRS, Washburn, or Batista's relief appearances are taken into consideration. Also understand that there's a bit of an error whenever you're dealing with PITCHf/x in this sort of way, as nobody can look at the numbers and identify pitches with 100% accuracy. Washburn and Batista in particular posed a challenge, as Washburn apparently has more than one pitch after all and Batista was all over the place. Bedard and Washburn each threw cutters, but I didn't bother trying to isolate them like I did for Batista because (A) that's hard, and (B) Batista used his cutter far more than anybody else. Finally, note that I grouped Batista's changeup and splitter together because sweet Jesus you try to tell them apart

So with that out of the way, bam, there you go. Data. I don't even know where to begin, and with some of the sample sizes, I'm not sure it's worth beginning. It's interesting to see how each entry is a little different from the others in each group - that catchers may indeed have their own similar but distinct game plans for every arm on the staff - but honestly, this isn't a project that I thought would lead anywhere amazing on its own. I just did it to do it, because as far as I know, nobody else has done this before, and if you want to gain an understanding of how pitch-calling differs from one catcher to the next - I mean a real understanding - I figure this is where you have to start. Any other approach would be inadequate.

0 recs | Comment 25 comments | Share

Story-email Email Printer Print

Comments

Pretty interesting

Given that we had to deal with the whole “personal catcher” phase earlier in 2008. Only thing that stands out in that regard (so far) is how the Bedard/Johjima and the Bedard/Burke battery handle lefties (with the former having a noticable emphasis on offspeed pitches compated to the latter).

by ThundaPC on Jan 20, 2009 8:47 PM PST reply actions 0 recs

what would be really interesting to see is what the era/baa is for each of these pitch/catcher combos. That would probably take some work though.

by pumpkin on Jan 20, 2009 9:22 PM PST reply actions 0 recs

CERA and CBAA aren't very useful statistics

I could look at, say, strike% and sws% tomorrow, but I’m not sure those would tell us very much, either.

by Jeff on Jan 20, 2009 9:25 PM PST up reply actions 0 recs

I'm confused about n

Is ‘n’ representing the number of pitches you were able to isolate out of the whole season? Felix’ add up to 2009, while fan graphs has him throwing 3199 overall.

by batura on Jan 20, 2009 10:17 PM PST reply actions 0 recs

Felix is fixed

I copied the wrong column into the table.

Yes, n = total number of pitches.

by Jeff on Jan 20, 2009 10:53 PM PST up reply actions 0 recs

Is there a way

to get Bedard’s spreadsheet?

Thanks.

by ele on Jan 20, 2009 10:59 PM PST reply actions 0 recs

You betcha

http://assets.sbnation.com/assets/60232/bedard_catchers.xls

by Jeff on Jan 20, 2009 11:03 PM PST up reply actions 0 recs

Awesome

Thanks again!

by ele on Jan 20, 2009 11:08 PM PST up reply actions 0 recs

The difficulty is it doesn't really have situation figured in.

Japanese catchers tend to mix up on different factors, for instance. Joh’s inclination to clean the plate area is to track a batter’s tendencies in the box and also subtle changes, like grip, ‘lean’, etc. Which apparently messes up the defense too, but eh. Whatever.

Favor the Bold

by IcebreakerX on Jan 21, 2009 5:23 AM PST reply actions 0 recs

Jeff Clement seems to have Jarrod Washburn and Cesar Jimenez confused

by seattlebruin on Jan 21, 2009 8:42 AM PST reply actions 0 recs

Zdurienciks new video advance scouting, I wonder if this is a piece of what they do.

The direction this going, maybe it’s in line with their work. He wouldn’t have replaced advance scouts unless he’s got something significant, a new approach or methodology.

I’d really like to see how they go about what they do, there has to be more to it than analyzing pitcher film to see if they are tipping their pitches, maybe this type of analysis to pick apart pitching sequences. Find some trends, habits, that type of thing. Not just for the pitcher, but for the catcher.

by Kermit. on Jan 21, 2009 10:56 AM PST reply actions 0 recs

Yes

by Jeff on Jan 21, 2009 11:04 AM PST up reply actions 0 recs

Do you know any specifics about what they are doing?

If you have info, are you permitted to talk about it?

by Kermit. on Jan 21, 2009 4:06 PM PST up reply actions 0 recs

They are taking all the advance scouting responsibilities in-house because some of the information they purchased from third parties in the past was unsatisfactory

by Jeff on Jan 21, 2009 4:17 PM PST up reply actions 0 recs

Oh. I meant specifics as in what they do as advance scouts.

I imagine they look for recent trends, new pitches for a pitcher, a guy that suddenly can’t turn on a fastball, things like that. But that’s all conjecture on my part, I figure there’s a few things they do that would make for a ‘gee whiz’ kind of moment.

by Kermit. on Jan 21, 2009 4:37 PM PST up reply actions 0 recs

I did my best

http://www.lookoutlanding.com/2008/7/8/567091/advanced-scouting

>:(

by Matthew on Jan 21, 2009 4:44 PM PST up reply actions 0 recs

Clement slightly more reliant on offspeed pitches against righties as a general trend.

I DRINK YOUR MILKSHAKE!!! I DRINK IT UP!!

by abender20 on Jan 21, 2009 11:38 AM PST reply actions 0 recs

I thought it was interesting

Clement trends to be fastball “light” on righties (out of the all the catchers), and fastball “heavy” against lefties but with Batista and Bedard, he’s the exact opposite, and with RRS, he seems to be fastball “heavy” vs. RHB and LHB. Maybe I’m restating what ThundaPC and abender20 are seeing, but at a first glance, that’s what immediately drew my eye. The fact that the trend didn’t care if the pitcher was LH (Bedard) or RH (Batista) was interesting as well.

"Relax, don't worry, have a homebrew." - Charlie Papazian

by kzuudj on Jan 21, 2009 12:57 PM PST up reply actions 0 recs

Two things immediately crossed my mind when I looked over the data:

-First, I tried to see if this led me to conclude that the Mariners as an organization had a consistent plan they expect their catchers to stick to. I don’t know if there’s enough data to confirm or deny this, which led me to…

-This doesn’t take into account any of the situations in which the pitches were thrown – who the hitters were, early or late in the game, runners on or not, and so on. I would be interested to see a breakdown of how different catchers pitched to the same hitters. If Clement calls for Hernandez to throw a certain sequence to Grady Sizemore, and Johjima calls for a totally different sequence, that would be worth nothing. This would drastically reduce the amount of data to compare, but I think that might generate more of a story.

Interesting data – but to me, lots of things that can affect the numbers and not provide context.

by PatLagreid on Jan 21, 2009 3:16 PM PST reply actions 0 recs

If you have enough pitches context drops out

because all catcher/pitcher combos see similar situations. They won’t be exact but it should be close.

by Edgar for Pres on Jan 22, 2009 5:19 AM PST up reply actions 0 recs

I think if you start getting that specific the sample sizes are going to be too small

But I would be interested in seeing 2-strike pitches by battery.

by Vatinius on Jan 22, 2009 12:27 PM PST up reply actions 0 recs

I wonder if some catchers stick to a gameplan

and others go with what is working well.

Do you have standard deviation data for these pitcher/catcher combos for pitch percentage between different starts?

Example:
Day 1 Bedard/Clement FB% = 60 (FB not working well, pitch abandoned)
Day 2 Bedard/Clement FB% = 80 (FB working well so used higher )
Day 3 Bedard/Burke FB = 70 (FB not working well, Bedard told to man up)
Day 3 Bedard/Burke FB% = 70 (FB working well but Burke thinks curve command will improve throughout game if continue with it)

Not sure which catching style works better or if its really visible to this degree.

by Edgar for Pres on Jan 22, 2009 5:25 AM PST reply actions 0 recs