PI bit on stealing bases

DMZ · April 21, 2005 at 11:08 am · Filed Under Mariners 

My weekly bit’s right here. Nothing you haven’t read here and elsewhere: depending on the situation, you need to steal bases with a high rate of success to make it worthwhile.

Feel free to email with article suggestions you’d like to see in the PI. Trying to find things to hit in a short form for a different audience while still trying to be informative and worth reading… whew. I’m thinking about writing about Bloomquist, generate some emails.

And as always, I’d like to mention that if you want these columns to keep going, or if you’d like to see this stuff in print, a totally random passer-by offered places you can write to express your opinions.


27 Responses to “PI bit on stealing bases”

  1. IgnatiusReilly on April 21st, 2005 11:17 am

    Thanks; in its appropriate place then:

    Regarding your article on the PI today DMZ, I have a question. You say that a man on second with no outs will typically lead to 1.2 runs being score by the offense. Does it matter how that man on second got there? For instance, a batter that hits a double might lend you to believe that the pitcher was grooving some pitches, whereas a bloop single and a steal might be more of a factor of random chance / weak catcher’s arm, etc.

    Does it make any difference? Does it statistically matter even if there is? Is there anything to show that stealing a base “messes with a pitcher’s head?”

  2. Baltimore M's Fan on April 21st, 2005 11:19 am

    Nice article Derek. You have gotten better and better with each one. People need to see more of this kind of stuff, instead of the ex-players who love to talk about themselves and think sabermetrics is for geeks.

  3. Milorad V on April 21st, 2005 11:44 am

    I enjoyed the article…I always do. Sabermetrics are an extremely interesting and valuable way to look at baseball. I would offer however that the ‘wincing’ I feel when someone says “Numbers are bull***t.” is identical to when someone says “Numbers are God.”
    In the case of the stolen base, there is no meaningful way to quantify the psychological effect on a pitcher and the other infielders when a notorious base-stealer is on first, throwing the minute, personal details of their game into crisis…errors are made, pitches get away, curve-balls hang, or fast-balls are thrown too frequently as a measure against the runner, infielders break for the bags…furthermore, the tiny mental-derailments may not show up in during the inning in question.
    Pressure in whatever form has its effects…whether they can be charted or not.
    Sabermetrics yes…but not at the exclusion of all other thinking!
    “When the only tool you have is a hammer, every problem looks like a nail” goes the old saw; ergo: If all you believe in are numbers…

  4. Jeff on April 21st, 2005 12:03 pm

    What about the psychological effect of having a 40-hr hitter at the plate? Or it’s the bottom of the ninth, you’re up by one, but runners are on first and second? There’s certainly pressure from all of these situations, but if we want to credit stolen base threats for added pressure, we need to do it all over the place. And other factors in a situation are more important than who’s on first base. I don’t think the difference between Ichiro at first and Boone at first is even close to the difference between Ibanez at the plate and Sexson at the plate. Or the difference between no outs and one out.

    What I’m saying is, if you have a guy who can steal bases at a high rate, use him. But don’t waste outs by sending a low percentage player just to create the threat of a stolen base.

  5. djw on April 21st, 2005 12:13 pm

    errors are made, pitches get away, curve-balls hang, or fast-balls are thrown too frequently as a measure against the runner, infielders break for the bags…

    These things, except maybe the last one, are perfectly quantifiable and countable (I’m assuming the success of the batter can serve as a proxy for 2-4; if it doesn’t, than I don’t much care). Your theory is prima facie plausible, but that’s no reason to believe it absent actual evidence. Your theory may contain a pyschological element which defies quantification, but you’re suggesting an impact on actual game events that can easily be counted.

  6. Jeff on April 21st, 2005 12:14 pm

    Addendum: Numbers are God, but the use of numbers are not. If an intangible helps a team win, it will show up in wins and losses – but we might not be able to measure it or identify it right now.

    While the stolen base percentage cut off might be 73%, or 72% if you correctly factor in added pressure or whatever, the theory is sound and the threshold is pretty accurate.

  7. Christopher Michael on April 21st, 2005 12:17 pm

    The effect on having a base-stealer on first base has more to do with the idea of speed. If Ichiro is on first base he doesn’t have to try to steal the base to cause the pitcher to be edgy.

    Same thing goes for when a player is thrown out. The pitcher was just given an out even though he originally had been beaten by the batter. Its much easier for a pitcher to get through an inning with one out and nobody on than no outs with one person on. So getting caught stealing helps the mood of the pitcher.

    Everything breaks down to averages at some point.

  8. chris w on April 21st, 2005 12:26 pm

    James Click at Baseball Prospectus wrote an article recently that quanitifies quite nicely the affect of having a fast runner on first (dubbed a “major threat” in the article) http://www.baseballprospectus.com/article.php?articleid=3913:

    His conclusion is that having a Major Threat on base does improve a batter’s performance by a little:

    The “‘major threat’ group increases the following batter’s expected performance by .015/.009/.023 over other baserunners.”

  9. chris w on April 21st, 2005 12:27 pm
  10. Todd on April 21st, 2005 12:31 pm

    Are there any thoughts on the wear and tear that a player endures constantly trying to steal? Ankles, feet, wrists, and fingers all get bumped, bruised and banged stealing bases and diving back to first. Am I wrong in thinking that Griffey Jr. stopped attempting to steal many bases because of injury concerns, or did he just outgrow them as he developed as a player.

  11. Milorad V on April 21st, 2005 12:39 pm

    #6 ‘than I don’t much care.’
    …and if it isn’t a nail, you won’t want anything to do with it…or it will cease to exsist.

    I in no way mean to ‘dis’ the perpetration/use/application of sabermetrics. I believe this is good for baseball. And I personally enjoy the insights into the game they provide. But for me, they seem a fascinating, exciting ‘window’, not the new omnipotent vista.
    It is quite possibly a failing of mine that I withdraw from ideologies which have at their cornerstone ‘We Are The Answer’.

  12. Jeff on April 21st, 2005 12:56 pm

    My point is that if something affects the wins and losses of a team, then it can be distilled into a number. You could say “a team’s chance of winning increases if the 2B picks his nose in the third inning.” Ok, this is something we could test. We’d need to know in which games the 2B picked his nose in the third inning. We don’t have this information now, but if we did, we could figure out if this improved the team’s chances of winning.

    How about “Willie B does the little things to help a team win.” If this is true, then one would expect a team with Willie B to win more games than the same team without Willie B, but with someone roughly equivalent in the “big things to help a team win.” This is harder to get a definite number for, but there is a number out there. And given enough evidence, we could get an estimate. (And given alternate universes and time lines, we could get an exact number.)

    This isn’t to say that everything that happens on the field does affect a team’s wins and losses. If the 2B picks his nose, it likely doesn’t affect whether the team wins, but it is an added part of watching the ball game. There’s a difference between watching baseball and analyzing baseball.

    [Note: I am in no way implying that Bret Boone picks his nose in the third inning, or any inning for that matter.]

  13. sodo on April 21st, 2005 1:13 pm

    DMZ, is the random passer-by that you refer Mike Thompson? Or is it some other person that you are talking about? Because Mike is the guy that runs the small Seattle-PI blog on the sidebar, and I just don’t think that he is exactly a random person.

  14. DMZ on April 21st, 2005 1:14 pm

    Uhhh… yes. That’s sort of the joke.

  15. sodo on April 21st, 2005 1:25 pm

    Sorry I didn’t see your sarcasm laced in between your post there. Now I feel pretty stupid.

  16. Nick on April 21st, 2005 1:30 pm

    I’m not big on stats, so I’m not going to do the research, but I’d be interested to know wher ethe teams that led the league in A. Stolen Bases & B. Stolen base success percentage ended up over th past 10 years. Or, more accurately, did teams who had lower percentages of successfull steals lose?

  17. Adam S on April 21st, 2005 1:50 pm

    Nick, good question. Here’s an article that’s some of what you’re asking — http://baseballprospectus.com/article.php?articleid=2607. It looks at runs scored since wins are depend highly on how good the team’s pitching was.

  18. Tim on April 21st, 2005 2:14 pm

    Jeff, that’s all true, but often there are many lurking variables that cannot be accounted for. I think the saber community often uses correlation and other statistics as being cause and effect, when in reality it is just measuring a relationship between two factors. Too many times people use some statistic and make a definitive statement on a particular strategy. I don’t think that’s the case with this stealing article, but you get my point.

    The numbers tell a very powerful story, but there is so much noise in baseball that it can makes the numbers less meaningful or conclusive. I would say you can analyze most of the results with these numbers, but not everything.

  19. Steve Thornton on April 21st, 2005 2:21 pm

    You could just as easily make the argument that aggressive basestealing, or the threat of it, messes with the BATTER as much as the pitcher. Especially if you’re asking him to swing at garbage to help delay the throw.

    Hitting takes concentration; paying attention to whether Fleetfoot is going to run or not, whether the next ball is coming towards the batter’s box or over to first base, the pitcher bobbing and swiveling and coming as close as he can to balking without actually doing so — these things hurt concentration.

    Just as anecdotal and suppositional as “messing with the pitcher’s head” and just as likely to show up in the evidence — i.e., not very.

    As for psychological factors not being measurable: if they are effects, they will show up in hits, walks, errors, runs. If they don’t show up in those things, then BY DEFINITION THEY DON’T MATTER. No one would argue that there are no psychological factors; but if they don’t change what happens on the field, who cares?

  20. Carl on April 21st, 2005 2:46 pm

    #17 Adam, Thanks for posting the Prospectus piece. It definitely adds to DMZ’s PI piece.

    #19 Steve, Your last paragraph is a beauty. Compact, pointed and, IMHO, correct.

  21. robbbbbb on April 21st, 2005 2:59 pm

    One point that I think bears on the discussion is the nature of the threat. To create that psychological effect on the pitcher (if it exists), then the pitcher has to fear the runner.

    Fifty-percent base stealers don’t create fear. Quite the contrary. They create anticipation. “Oh, yeah. Go. I need that extra out. Please?”

    If you’re going to advocate for putting pressure on the defense, then you’d better execute well in the first place. Good execution will show up as high-percentage basestealing. To create that pressure, that fear, then perform well. The psychological effect will take care of itself.

  22. roger tang on April 21st, 2005 4:05 pm

    re 21

    Yeah….I think good execution is a potent force, in and of itself, and in the psychological sense. Bit of a feedback issue (which I think DOES show up in the numbers; i.e., they’re already being accounted for).

  23. RL on April 21st, 2005 4:21 pm

    Question for DMZ (or anyone else). Whenever I see these articles about when it makes sense to steal a base, they are based on run expectation tables. So we can compare, like Derek did, how many runs a team should score with a runner on 2nd and no outs v. no one on with 1 out, etc. But this doesn’t always comport with what happens. For example, Winn could attempt to steal 2nd, the ball could get thrown into centerfield, and Winn gets to 3rd. I think that happens more than an insignificant amount of time. Are those times taken into account?

    Same things with sac bunts. Generally not a good idea using the run expectation table – but what about the times the infielders screw it up and everyone is safe? How often does that happen?

  24. Christopher Michael on April 21st, 2005 5:37 pm

    Well the only time I ever like sac bunts is when there is a good chance that the hitter will get an out anyway. Having our catcher sacrafice bunt to try and manufacture a run is a good idea. Having Winn sacrafice bunt is stupid.

  25. Ralph Malph on April 21st, 2005 6:09 pm

    Re 23: and what about the times the runner jams a finger sliding and misses the next couple games (or worse).

    Obviously a good statistical model would take all these things into account.

  26. Tim in Japan on April 21st, 2005 8:17 pm


    In the situation described in the article, besides Winn’s success rate at stealing bases, don’t you also have to consider the success rate of the pitcher/catcher at throwing runners out? Should these be weighted equally?

  27. Trev on April 21st, 2005 11:19 pm

    Someone at BP (maybe even you, DMZ) noted in a Triple Play that Lou Pinella had run nearly every one of his teams since he started managing the M’s above the break even point.

    Granted a lot of this has to do with player speed, but when it comes to grading managers, SB% should be considered something quantitative to measure.