Ichiro, High Leverage Base Stealer

Dave · April 9, 2008 at 9:51 am · Filed Under Mariners 

During one of the off-season threads, a commenter and I had a discussion about Ichiro’s base stealing. Among those who aren’t Ichiro fans, it’s a fairly often repeated claim that Ichiro doesn’t run enough in situations that matter, and he pads his SB totals by taking bases early in games when the impact on wins and losses isn’t that large.

Well, thanks to the amazing Baseball-Reference.com, we can now disprove that myth once and for all. Here’s Ichiro’s career splits, now with Leverage Index added. Look at the breakdown:

High Leverage: 826 PA, 349 times on base, 67 steals, 19 caught, 25% Steal Per On Base
Med Leverage: 2606 PA, 1006 times on base, 147 steals, 37 caught, 18% Steal Per On Base
Low Leverage: 1784 PA, 684 times on base, 58 steals, 10 caught, 10% Steal Per On Base

About a 1/4 of the time that Ichiro reaches base in a high leverage situation, he takes off, compared to 1/5 of the time in normal situations and 1/10 of the time in low leverage situations.

I think we can safely put that false critique to bed.


64 Responses to “Ichiro, High Leverage Base Stealer”

  1. Grizz on April 9th, 2008 3:02 pm

    – J. Walker relieved M. Albers
    – I. Suzuki singled to left center

    You do realize that Jamie Walker is left-handed, right?

  2. Kunkoh on April 9th, 2008 3:02 pm

    I wonder how the people that think Ichiro should steal more, particularly in high leverage situations, would react if he started getting picked off more.

    It’s easy to say that if he tried more high leverage, late inning steals when the M’s are within 2 runs that he would maintain a 78% success rate; but I don’t think that we can accept that because of a few factors that are unknowable to us.

    How is he choosing? Is it when he believes he has a 75% chance of success? Whatever his method, attempting to steal more then would mean he would have to be less choosy. The more he steals the more risk there is involved. I think it’s a safe bet that his success rate may drop; even as fast as he is.

    All it takes then is for a few inning ending/shortening outs by him attempting to steal in close games, and the fans that are already saying he’s selfish because he only steals to pad his stats would be saying, “he’s selfish and only stealing to pad his stats and it’s costing the M’s games. He’s not a team player.”

    It’s a double edged sword, and with the people that aren’t happy with his production right now, will not be happy no matter what he does. Personally, I’m happy with the results.

  3. Oolon on April 9th, 2008 3:09 pm

    Grizz Says:

    – J. Walker relieved M. Albers
    – I. Suzuki singled to left center

    You do realize that Jamie Walker is left-handed, right?

    Yes, but I believe there are several reliably documented cases of bases being stolen on left handed pitchers – especially by the better base stealers.

    This is where Ichiro’s skills come into play. He’s a great base stealer and hopefully can steal on a left handed pitcher when the situation requires it.

  4. Jay R. on April 9th, 2008 3:22 pm

    “That was a number I pulled out of a hat.””

    Interesting. Most other people wear their hats on their heads.

  5. Grizz on April 9th, 2008 3:27 pm

    Yes, but I believe there are several reliably documented cases of bases being stolen on left handed pitchers – especially by the better base stealers.

    And there are reliably documented cases of more runners getting thrown out against left-handed pitchers. The chances for success go demonstrably down. Ichiro is well aware of this.

  6. MKT on April 9th, 2008 3:30 pm


    I recently took a look at stolen base attempts in the ninth inning, using WPA, to assess the value of various stolen base attempts in the ninth. It turns out that in the 9th inning of a close game, you don’t need to be nearly as good a base stealer as you do typically in order for get an advantage from attempting to steal

    Really nice stuff. I like how it takes a more detailed look at specific score/out/men on base situations, rather than doing what most analysis of stolen bases that I’ve seen does, namely treat all base stealing situations as if they’re the same.

    It’s similar to the overly blanket criticism of sacrifices. Somewhere I read something similar to your analysis except applied to bunts, exploring what are the situations when it does make sense to sacrifice — this might’ve been in Tango/Dolphin/Lichtman’s excellent “The Book”.

  7. NateDawgUS on April 9th, 2008 3:33 pm

    Whoa! I only recently started reading this blog but I’m just amazed at the fact that there are people who aren’t Ichiro fans. Is that even legal?

  8. msb on April 9th, 2008 3:44 pm

    looks like pretty much the same line-up as yesterday, with Burke tossed in at catcher.

  9. Jeff Nye on April 9th, 2008 3:45 pm

    There might be two or three people on the planet who know more about stealing bases than Ichiro.

    I’m pretty sure none of them post here.

  10. doededoe on April 9th, 2008 4:12 pm

    59- One of those people is Ricky Henderson, but if he posted here we would know because of the whole talking in the third person thing.

  11. skyking162 on April 9th, 2008 4:18 pm

    Also, maybe they were criticizing Ichiro last year, when his leverage splits were:

    High — 5 SB, 4 CS
    Medium — 20 SB, 3 CS
    Low — 12 SB, 1 CS

    For reference, the LI cutoffs at B-Ref are:
    High >1.5
    Medium .7 to 1.5
    Low <.7

    20% of PAs are in the High category and 40% are in each of Medium and Low. Since the categories aren’t equal, you need to use Dave’s approach of dividing stolen base attempts by opportunities. Here’s a shortcut, because i don’t have opportunities: just double the high-leverage SB attempts:

    High – 18 (9*2)
    Medium – 23
    Low – 13

    Ichiro didn’t steal when it didn’t matter. And it did hurt that he was only 5/9 in High LI situations.


    There are many high-leverage situations when the break-even SB-rate is very low, maybe 60%. In low-leverage situations, it might be more like 90%.

  12. walkie83 on April 9th, 2008 4:40 pm

    When I say he doesn’t steal “enough” … I guess I’m just unfairly comparing him to Ricky Henderson .. or .. well, any guy with speed back in the day. It just seems like with his speed, he should be able to swipe bases more often. I guess I miss those days where people were worried about fast guys on the base paths.

    You can make the claim that he does affect the game when he is on base … but he sure doesn’t the way some of the top base stealers do. I just think with all the singles he hits, he should be stealing more. Henderson used to turn a walk into a triple on a regular basis.. now I know Ichiro isn’t that fast.. but he’s fast enough.

  13. xeifrank on April 9th, 2008 4:57 pm

    Success rates SB/(SB+CS)
    HighLev: 77.9%
    Medlev: 79.9%
    LowLev: 85.3%

    vr, Xei

  14. Oolon on April 10th, 2008 9:02 am

    OK, as I’ve said repeatedly, my impression is that while Ichiro is a great player he doesn’t try to steal bases late in close games.

    Using the Baseball-Reference.com link Dave gave in the first posting on this discussion you can go up a few lines from the “Leverage Index” numbers and see Ichiro’s stats in “Late and Close” plate appearances. “Late and Close” is defined as: “Late and Close are PA in the 7th or later with the batting team tied, ahead by one, or the tying run at least on deck.” We’re ahead by one, or tied, or behind by one or two runs.

    These are the exact situations that I was looking for.

    Here’s what the stats show. I took the total hits and subtracted triples and home runs (not much chance of a stolen base with either of those) and added in the base on balls, intentional base on balls, the hit by pitches, and the reached on errors to get an on-base total. Then I took the total number of stolen base attempts and divided by this number.

    Ichiro tries to steal a base 14% of the time in these “Late and Close” situations. That’s about midway between his “Medium Leverage” and “Low Leverage” averages.

    I wanted to compare him to other base stealers, so this morning (they don’t consider baseball statistics “work related” here, so I can’t see them at work…) I grabbed the stats for some of the leaders in the stolen base category from last year. I only had a few minutes before heading to work so I was only able to print out the stats for the 6 leaders and Ichiro. The names listed as last year’s stolen base leaders were Reyes, Ramirez (Hanley), Pierre, Roberts (Brian), Crawford, and Byrnes.

    I ran the same percentages for these players and here are their stolen base attempts per time on base in these “Late and Close” situations:

    Crawford – 35%
    Reyes – 26%
    Ramirez – 24%
    Pierre – 23%
    Roberts – 18%
    Ichiro – 14%
    Byrnes – 12%

    These are lifetime stats – not a single season.

    I think the numbers show that Ichiro is very conservative late in games. This very well might be a very good thing. I don’t know if he and Byrnes are the smart ones and the others are costing their teams runs instead of gaining runs by their impetuosity.

    Here are the success percentages in these “Late and Close” situations for the various players:

    Crawford – 87%
    Reyes – 89%
    Ramirez – 81%
    Pierre – 72%
    Roberts – 79%
    Ichiro – 81%
    Byrnes – 79%

    And combining the two we get the percentage of time they successfully steal a base when they get on in a “Late and Close” situation:

    Crawford – 30%
    Reyes – 23%
    Ramirez – 19%
    Pierre – 17%
    Roberts – 15%
    Ichiro – 12%
    Byrnes – 9%

    So the bottom line is that if Ichiro gets on base in the 7th through 9th innings during a “Late and Close” game – he’ll steal a base about 1 time in 8 whereas Hanley Ramirez will steal the base 1 time in 5, Reyes 1 time in 4 and Crawford almost 1 time in 3.

    I think these numbers support my observation that Ichiro is very conservative in close games in the late innings. Whether this is a good thing or a bad thing would require additional research.

Leave a Reply

You must be logged in to post a comment.