Ichiro thinks in a unique fashion. Check his thinking here. [fixed it for you]

BTW, what Harold Reynolds et al. fail to mention is that while the running game can put pressure on the defense, you don’t actually have to run for that pressure to be there, and after a successful steal the pressure is sometimes actually removed (runners on first and third, the pitcher holds the runner at first, throws over, perhaps even lets his pitch selection be affected; runners at second and third, a lot of that goes away).

No, it is not feasible.

Then I must not be understanding– are you contending that there is no opportunity where Ichrio believes with a greater than 80% chance he could steal that he does not take the opportunity to? What about with a 15-0 lead in the 9th?

Also, the assumption you’re arguing against is not my point at all.

I know, and I agree with the point you’re trying to make. I’m arguing that this part of your post:

If he picked the next best ten times to run, and they were 70% chances

…implies that he has no additional 80% chances from which to choose as one of the next best ten. I think that might not be right, which in turn means the way you illustrate the point may be flawed.

If the M’s wanted him to get his 80 steals by any means necessary, then he could run in that 15-0 situation. At some point Ichiro is making a decision about whether he should or shouldn’t run, and chance of success is only one factor in that equation. If he’s persuaded to change the other inputs to that decision, he could increase his total without impacting his success rate.

I digress. It’s a good post and I’m probably just nit-picking this minor point to death for no reason.

It’s certainly possible that in a game situation that Ichiro thinks he’s only got a 70% chance of success, but the breakeven is 60%, will present a better opportunity than a game situation where he thinks he has a 75% chanceof success, but the break-even point is 74%.

In any case, the basic win gain per SB is .02 wins and the basic loss per CS is -.04 wins. An 80/20 SB/CS would give you +0.8 wins, and 40/10 would give you +0.4 wins.

It’s not clear that Ichiro would be able to do an 80/20 set, but the gain here is +0.4 wins than what he otherwise does. On top of which his baserunning can make up for some of that as well.

When it comes to this kind of gain, that requires alot of gut feelings by the runner, the *only* thing you should be telling Ichiro is the breakeven point each time he gets to 1B (1B coach to Ichiro: “71% right now… it’s 3-0, 84%… 0-2, 62%”).

You let Ichiro figure out if he can beat that level.

The reason I would suspect this … I have to assume that at least some of the time, Ichiro is given the sign to steal (either a straight stel or a hit and run). I also assume that Ichiro does do (even if he thinks it a poor decision) what he is instructed to.

Based on what McLaren sais, and my limited knowledge of baseball, I would assume that, with a runner with Ichiro’s talent, it is rare that a straight steal is called for, so I would have to guess that most of the time these steal attempts are actually part of a hit-and-run.

I don’t have any statistical analysis to back this up (so, it could be total hogwash), and I know that announcers are not always … umm … aware of the truth of certain situations, but it appears to me that failed (meaning, when the batter swings and misses) hit and runs much more often produce a caught stealing than a successful steal. On the flip side, successful hit and runs never produce a stolen base (or a caught stealing).

With only 8 CS last year, even if only two or three are a result of a failed hit and run, unless MANY of his successful attempts are either the rsult of a called steal or a successful steal on a failed hit and run, Ichiro’s SB% would go up even higher.

Also, the assumption you’re arguing against is not my point at all.

It doesn’t necessarily justify the attempt, because his judgment in valuing those situations differently may also be spot on, but it would mean that the theory put forth in this post that each additional attempt reduces his chance of success would not be true.

The average of these 50 attempts will be .8275.

So, on the 51st attempt, he might be at .700. And on and on. At that point, it becomes a break-even play. If he thinks, based on the game situation, the batter, the pitcher, the catcher, the turf, that he’ll be as well off running as not, he might as well not run.

Now, clearly the model is not necessarily a drop of .005 like I’m showing. At the same time, his 50th easiest stolen base attempt *must* be less than the overall average of all 50, since there’s a bunch of steal attempts in there that are at the 90% level and higher, in terms of what he expects his success rate to be.

Ichiro, Raines, Beltran. These are smart and fast baserunners. I’d let them decide how often they should steal.