THE BOOK--Playing The Percentages In Baseball

Okay, he threw 55 pitches, 36 of which were fastballs.  Here is the breakdown of average velocity for every four fastballs:

1-4: 89.5 mph
5-8: 90.6 mph
9-12: 90.7 mph
13-16: 91.2 mph
17-20: 91.1 mph
21-24: 89.7 mph
25-28: 90.4 mph
29-32: 89.5 mph
33-36: 89.7 mph

Good job.

01-12: 90.3
13-24: 90.7
25-36: 89.9

Presuming that 1 SD per fastball = 1mph, then 1 SD per 12 pitches must be 0.3mph.

***

Two questions:
1. did his % of fastballs thrown go up as the inning went on (i.e., he had only the strength to throw fastballs, and not to throw breaking pitches)?

2. how much movement did his fastballs have?  as he went on, did the movement on his fastballs lessen?  (like say, Pedro in the Grady Little game?)

Do you mean with regards to the same first inning, or his distribution/selection in innings 2-4 compared with 1 (he only went 4 innings)?

If just the first inning, he faced 11 batters, one full go around and then Yunel+Renteria once more.

Batters 1-3: 10/16 fastballs (63%)
Batters 4-6: 14/20 fastballs (70%)
Batters 7-9: 7/10 fastballs (70%)
Batters 10-11: 5/9 fastballs (56%)

If fastball usage for the whole game:

First Inning: 36/55 (65.5%)
Second Inning: 11/17 (64.7%)
Third Inning: 6/13 (46.2%)
Fourth Inning: 14/19 (73.7%)

In terms of movement it is probably easiest to look at in a graph.. is there a spam blocker here with linking to a file?

Well, before I have to run out for a little, without a graph, here is the average movement in numbers:

1-12: 7.7 horiz, 10.3 vert
13-24: 6.3 horiz, 11.2 vert
25-36: 8.0 horiz, 10.4 vert

And looking at his combined fastball usage from the four innings:

Inning 1: 36/55, 65.5%
Innings 2-4: 31/49, 63.3%

The XML data file for Wandy Rodriguez on August 1, 2007, can be found at http://gd2.mlb.com/components/game/mlb/year_2007/month_08/day_01/gid_2007_08_01_houmlb_atlmlb_1/pbp/pitchers/434643.xml

The parameters of interest are start_speed, pfx_x, and pfx_z.

Well, well, well… that is some fantastic research right there.  I would not have believed that someone could maintain that level of performance over 50 pitches.

I think I’ll have to withdraw my “40 pitch threshhold”.  It doesn’t seem like he was being overworked, if his fastball had very similar speed, and the same amount of “bite”.

The velocity for fastballs by inning:

1: 90.3
2: 87.5
3: 89.1
4: 88.1

So, from the loooong first inning to the second inning he decreased somewhat significantly and then worked himself back up, throwing an average of 16 pitches per inning for innings 2-4.

Great stuff, Eric.  Hopefully, you can work this into an extended article, and look at the 20 or 30 or so longest outing cases since 2005 and report on the results.

Yeah that’s something I’m looking into now.  I could probably get my hands on BIS data somehow for the non-pitch fx games.

You needn’t withdraw your proposal of the 40 pitch rule.  I thought it was to prevent players from getting injured… this just proves that throughout the inning, they can still pitch well.  We don’t know how they did in the next start, or next 3 starts.

Anyways… I’m not a fan of the idea.  Being inefficient with pitches in a start is a quality that bad pitchers are more likely to have than good pitchers.  You’re giving a talent rebate to the crappier pitchers.  As far as injury concerns go, it’s the manager’s job to pull a pitcher when he could be facing injury, and it’s the player’s job to speak up if he thinks he might be getting hurt.  In my opinion.  I’d like to think if it was my multi-million dollar arm on the line, I’d be pretty vocal if I thought I was risking injury.

Did you see the names on that list?  Beckett?  Lowe?  Bedard?  The manager’s job, like leaving him in for 150+ pitches?  Thanks, but no thanks.  Managers are the last people to treat pitchers like human beings.

As I recall, in game 7 of the ALCS, when Pedro Martinez went beyond his 100 pitch comfort zone, he threw his fastest pitches of the game (according to the radar gun readings reported on TV). But he started missing his spots by a foot or two, going to a deep count batter after batter. So fastball speed need not be a clear indicator of fatigue.

Also to properly explore the in-game aftereffects of a high pitch inning, we should look as well at rest between innings, which we could model primarily from opposing pitchers’ pitch counts in their half-innings.

Rodriguez’s 55 pitch inning also seems to be the highest of the 2000-2007 period, with Victor Zambrano next, throwing 54 in 2nd inning of a game 4/18/2003.

whoops; that’s Martinez in ALCS of 2003

Perhaps the better questions should be, over the years, how many arms have bit the dust (surgery or extended DL time) after being exposed to a 40+ pitch inning and then required to come back out and throw at least one more inning?

I’d guess if you broke that question down into two categories, MLB service time =< 4 years and those above 4 years you might find out why so many young arms visit Doc Andrews. And I’d even further guess many of those visits happened within 18 months of their 40+ pitch inning.

The flip-side is, if they didn’t have surgery or go on the DL in 18 months, how many of them saw a significant climb in ERA (and xERA for that matter) indicating potential arm fatigue?