Yadier Molina, hit-and-run plays, and ideas for a study

What you get here is a little like an asteroid collision in progress, such as that recently observed by the Hubble Space Telescope. A couple ideas from different places crashed together in my head, a few parts stuck, and others went shooting off in random directions. I haven’t corralled them into nice orbits, yet. Instead you get to observe the messy collision and its immediate aftermath in all its glory, or lack thereof.

The first idea came from a FanPost at Beyond the Box Score that asked why teams even bother to attempt steals against the best-throwing catchers when the net result is below break even. Trip Somers suggested in the comments that some of the plays recorded as caught stealing may be failed hit-and-run plays.

I decided to look at video of all the runners caught stealing by Yadier Molina in 2009, using Retrosheet to identify all the plays. Retrosheet and Baseball-Reference list Molina as allowing 32 stolen bases against 22 caught stealing in 138 games played as a catcher in 2009. However, the play-by-play accounts indicate that six of the runners caught stealing were caught by the pitcher, either as the result of a pickoff throw, or in one fascinating case, when the pitcher Chris Carpenter simply held onto the ball while the runner Michael Bourn broke for second. Carpenter threw to shortstop Julio Lugo, who applied the tag, and umpire Paul Emmel called Bourn out and immediately ejected Bourn for arguing. Television replays appeared to show that Bourn had beaten the tag.

Accuracy of umpire calls notwithstanding, that leaves us to account for 16 baserunners caught stealing by Yadier Molina. How many of them were caught on broken hit-and-run plays? That turns out to be a little bit harder to answer than I expected, even with the excellent resource of MLB.tv video archives available for every game. My best interpretation is that at least nine runners were caught stealing by Molina on straight steal attempts, and at least five and possibly as many as seven runners were caught stealing on failed hit-and-run plays.

Date Opp Runner     Pitcher     Score Inn Out Count Pitch Type                Batter Action     Comments
4/11 Hou K.Matsui   A.Wainwright 0-0   1   0  1-1  slider low and outside     take for a ball   straight steal attempt
4/16 Chi K.Fukudome A.Wainwright 1-1   3   1  3-2  slider down the middle     swing and miss    possible hit and run, no look back to home
5/9  Cin A.Rosales  K.Lohse      2-2   4   2  2-2  fastball high-middle       take for a ball   straight steal attempt after 0-2 pitchout
5/25 Mil C.Counsell C.Carpenter  0-0   7   0  1-1  cutter low-outside corner  swing and miss    hit and run
5/31 SF  F.Lewis    D.Reyes      3-5   8   1  0-0  pitchout                   take for a ball   straight steal attempt
6/23 NY  L.Castillo J.Pineiro    1-0   6   0  0-0  sinker low and outside     swing and miss    hit and run
7/1  SF  T.Ishikawa A.Wainwright 1-0   5   1  0-0  slider low-outside corner  swing and miss    hit and run, squared to bunt but swung
7/5  Cin B.Phillips C.Carpenter  4-0   4   1  3-2  cutter high-outside corner foul tip          possible hit and run, also running on 2-2
7/30 LA  C.Blake    B.Thompson   3-3   8   1  1-0  fastball outside           foul tip          hit and run, weak swing
8/4  NY  A.Cora     J.Pineiro    4-7   5   2  0-0  curve low-outside corner   take for a ball   straight steal attempt
8/4  NY  A.Pagan    D.Reyes      4-7   6   2  1-0  fastball low-inside corner take for a strike straight steal attempt
8/20 SD  T.Gwynn    J.Pineiro    0-0   1   2  0-0  fastball low-middle        take for a ball   straight steal attempt
8/20 SD  E.Cabrera  J.Pineiro    4-1   3   1  2-0  fastball low-inside corner take for a ball   straight steal attempt
8/28 Was W.Harris   J.Smoltz     1-1   3   2  1-1  slider outside             take for a ball   straight steal attempt
9/15 Fla J.Cantu    A.Wainwright 0-0   2   0  3-2  fastball low-inside corner swing and miss    hit and run
9/25 Col E.Young    C.Carpenter  0-1   5   1  1-0  cutter low and inside      take for a ball   straight steal attempt after 0-0 pitchout

image
Yadier Molina shows off his arm, September 22, 2009. (Icon/SMI)

I don’t see any particular patterns, but I present all the detailed data in case anyone wanted to see it. However, this got me thinking about how one could tell from the available data whether a particular play was a hit and run without having to consult the video for each play. This was the other asteroid that collided with my investigation of catcher defensive recordkeeping, ending my hopes of arriving at a useful and tidy conclusion.

In the spirit of this thread at the Book Blog, I’m simply going to present my ideas for how one might go about identifying hit-and-run plays.

The idea is to have a section where anyone can post a suggestion for a study that they can’t do themselves for whatever reasons. Then maybe someone else can take these ideas and run with them, perhaps writing an article on one of the various sabermetric web sites. (Of course, if there is already some research in that area, someone can point that out too.)

When evaluating a play using the data available from MLB Gameday, here’s my ranking of how confident we can be that the play was a hit and run. This list is not exhaustive. For example, I left out some uncommon plays like the pitchout that is swung at by the batter. All of these scenarios assume a runner on first base. I’m not considering the situation with runners on first and second base, although I suppose a hit-and-run play might also be executed occasionally in that situation. What I’m looking for here is a basic framework and a reasonable approximation. The data we currently have, short of getting real-time baserunner information from a system like FIELDf/x, doesn’t allow us to get anywhere close to perfect information on when hit-and-run plays occur.

Yes, a hit and run is highly likely to have occurred.
An air ball is caught and the runner is doubled off.

Yes, a hit and run is fairly likely to have occurred, but we can’t be sure.
The batter swings and misses and the runner gets a stolen base or is caught stealing.
The batter fouls the pitch off and the runner was going.
The runner goes to third on an outfield single (we may be able to assess the probability in a more fine-grained manner by looking at the location where the ball was fielded.)

Yes, a hit and run may have occurred, but the likelihood is questionable.
Ground ball double play turned 6-3 or 4-3.
Ground ball out with runner advancing to second.
The runner thrown out at third on an outfield single (we may be able to assess the probability in a more fine-grained manner by looking at the location where the ball was fielded.)

We can’t tell with any accuracy either way.
There are three balls and the batter takes ball four or is hit by pitch.
An air ball is caught in the outfield or a popup is caught on the infield.
Batter hits an infield single, runner to second.
Batter doubles, runner scores or is thrown out at home.
Batter triples or hits a home run.
Batter is safe on error.

No, a hit and run is unlikely to have occurred.
3-2 count, 2 outs.
The batter takes the pitch with less than three balls.
A line drive is caught on the infield and there is no double play.
Ground ball double play unless it is turned 6-3 or 4-3.
Batter hits an outfield single, runner to second.
Batter doubles, runner to third.

This isn’t a terribly satisfactory article for me in terms of the lack of conclusions. However, rather than letting this work molder on the substantial pile of research I have started but never finished, I decided to share my results anyway in the hopes that my ideas will spur further analysis.

By the way, from among the trivia I learned from Retrosheet during this inquiry, did you know that all three Molina brothers have the name Benjamin? Benjamin Jose Molina, Jose Benjamin Molina, Yadier Benjamin Molina. Really creative with the naming there, Mom and Dad. (It should come as no surprise the Molina brothers’ late father was named Benjamin.)


7 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Peter Jensen
14 years ago

Of course, you do have the pitch sequence “*>X” to tell you when the runner was running on a hit ball.  I have speculated, as have others, that the conventional hit and run, where a batter is required to swing at the pitch, has largely been replaced by what used to be called a “run and hit”, where the batter knows that the runner is attempting to steal but only swings at pitches at which he would normally swing.  For those of you who are in a position to interview managers and players that would be an interesting question to have answered.  The “run and hit” is almost impossible to differentiate from a green light steal where a runner steals without a signal to the batter.

Nick Steiner
14 years ago

Very cool Mike. 

Another thing I think that you could use to identify hit and runs would be to look at when a batter swung at a pitch well outside of the strike zone on a 1-0 or 2-1 count (which is usually when hit and runs occur, I think).  If the batter is safe or caught stealing on a swing and miss on a pitch a half a foot off of the plate, you can bet it was a hit and run.

Mike Fast
14 years ago

Peter, thanks, and good questions.

I tend to piddle with Retrosheet data rather than using it to its full extent, and I sometimes forget what data is available there.  Do you know where Retrosheet gets its data on when the runner was running with the pitch?

Peter Jensen
14 years ago

I believe they get it from MLBAM.  I asked Ross about this when I got my tour of watching the MLBAM stringer at work, and he said that they collect the information but that it was not part of the information that goes into the XML files.

Mike Fast
14 years ago

Ah, that explains a lot.  I thought they got their data from MLBAM, so I didn’t even think of looking in the Retrosheet event files for data that wasn’t in the XML files.

I know that MLBAM is driven primarily (and understandably so) by revenue goals from increasing the entertainment value of Gameday, and recordkeeping for analysis is a secondary concern for them, if at all.  But it sure would be nice if all the data they collected were made available in the XML files.

Peter Jensen
14 years ago

My impression was that some of the data was being collected for paying clients and at their request the results were not being made public.  But I could be totally wrong about that.

Todd
14 years ago

I know that it wasn’t really the main point of your article, but I have two other thoughts on why teams might attempt to steal on Molina. First, from a game theoretic standpoint, the optimal number of attempts really has to be above 0, because otherwise the Cardinals would gain advantages from the knowledge that other teams would never steal (things like never having to pitchout). It’s entirely possible the number should be lower then it is, but 0 seems unlikely. There was a short series of articles on pitch choice and sequencing either here on on Fangraphs around November that talked about the game theory that goes into this.

The other thing is pitch selection. If opposing managers or runners are good, or at least think they are good, at guessing when Cardinal pitchers will, e.g., throw breaking balls, then their odds of stealing successfully (presumably) go up. And perhaps if they guess right, their odds go above break-even (granted, they ought to factor in the probability that they guess right, but people tend to be overconfident in their abilities… which is no doubt another factor here, in the more straightforward sense of thinking that YOU can steal on Molina, even if others can’t…).

Of course, speaking as a Cardinals’ fan who watches most of their games, none of the above keeps me from reacting with glee at the foolishness of attempts to steal on Molina. Or even better, when they take big leads off first and then get picked off.