Stealing 3B with 2 outs.

[AztecBill]

The Padres scored their 6th run in SF when Cabrera was given home via an error by the Giants on an attempted steal of 3B with two outs.

I always agreed with those that said attempting to steal 3B with 2 outs is really stupid. But the numbers show that it may not be as stupid as it seems.

Percentage chance of Scoring with 2 outs
From 2B 23.29%
From 3B 29.16%

Results of attempting to steal 3B (1993-2011)
25.8% Out
70.3% Safe
3.8% Score on Error

Total chance of Scoring after steal attempt of 3B with 2 outs
0% = 25.8% * 0%
20.5% = 70.3% * 29.16%
3.8% = 3.8% * 100%
------------------------------
24.30% <-- Percent chance of scoring if steal is attempted.
23.29% <-- Percent chance of scoring without an attempted steal

---

Batters have a different success rate of scoring runner via a single. If the batter has little power, hits line drives, and gets a lot of infield singles (Eric Owens), his chance of scoring a runner from 2B is much less than a batter that has enough power that the outfield plays deep and gets few infield singles (Ruben Rivera). I use those two because they prove an interesting point. Eric Owens hit .378 with RISP but only plated half the runners from 2B with a hit. Ruben Rivera only hit .198 with RISP but never failed to plate a runner from 2B when he got a hit. Not all RISP hitting is equal. Anyway back to the point. If Eric Owens is up stealing 3B with 2 outs becomes even more of a good idea but If it is Ruben Rivera it is a worse one. Likewise, who is catching and who plays 3B makes a difference too.

So does making the throw to 3B make sense? You make an out 25.8% and he scores 3.8% so he scores 12.8% of the time that one of those happens which is much better for the defense than allowing the 29.16% chance of him being on 3B.

[Aztec89]

Wow.

I am not sure what to do with this new found knowledge. But the numbers are certainly way more favorable to the runner than I would have thought.

Of course, get thrown out at third with 2 out, you kill the inning. Which then means you need to extrapolate, the number of potential runs lost, following the 3rd out at 3rd.

[nctyaztec]

Bill would be great as a press secretary for the White House ;-)

[AztecBill]

Sept 25, 2012 16:33:32 GMT -8 Aztec89 said:

Wow.

I am not sure what to do with this new found knowledge. But the numbers are certainly way more favorable to the runner than I would have thought.

Of course, get thrown out at third with 2 out, you kill the inning. Which then means you need to extrapolate, the number of potential runs lost, following the 3rd out at 3rd.

Quite true and very good point. If you have to have more than one run, it is really stupid. If you add in the chance of getting a second run (8.62%), than attempting a steal of 3B is stupid. The numbers above should be considered only if that one run is all you really need.

Bottom of the 9th in a tie game, the value of that run is the game and there is no value in the second run.

Top to the 9th in a tie game that run changes the chance of winning from below 50% (because without it the other team is half an inning ahead) to around 90% depending on your closer. A second run would only marginally increase your chance of winning above that.

[AztecBill]

Sept 25, 2012 20:13:15 GMT -8 nctyaztec said:

Bill would be great as a press secretary for the White House ;-)

That takes none of the skill I have shown. They just have to lie and the press will nod their heads and report it as fact.

[AztecBill]

Using the same methodology, the increased percentage chance of scoring one run is shown below for different steal attempts.

2B with no outs - 4.97%
2B with 1 out - 4.43%
2B with 2 outs - 3.69%

3B with no outs - 10.34%
3B with 1 out - 10.29%
3B with 2 outs - 1.01%.

---

Facts and Calculations

Facts
Table of chance for scoring 1 run.

Outs - Emp - 1B - 2B - 3B
0 27.54 41.54 57.29 84.67
1 16.70 27.82 40.68 63.94
2 08.62 12.97 23.39 29.16

Results of attempt to steal 2B
Safe 65.8%
Out 29.5%
Go to 3B on error 4.6%

Results of attempt to steal 3B
Safe 70.3%
Out 25.8%
Score on error 3.8%

Calculation
Steal 2B with no outs = .5729*.658+.167*.295+.8467*.046 = 46.52%
Steal 2B with 1 out = .4068*.658+.0862*.295+.6394*.046 = 32.25%
Steal 2B with 2 outs = .2329*.658+.00*.295+.2916*.046 = 16.67%

Steal 3B with no outs = .8467*.703+.167*.2580+1*.038 = 67.63%
Steal 3B with 1 out = .6394*.703+.0862*.2580+1*.038 = 50.97%
Steal 3B with 2 outs = .2916*.703+.0*.2580+1*.038 = 24.30%

---

One of the bigger surprises is stealing 3B with no outs increases your chance of scoring more than any other steal attempt. Traditional baseball says to never try stealing 3B with no outs. The big difference between the chance of scoring from 2B with not outs (57.29%) and scoring from 3B with no outs (84.67%) makes it a good try, given normal chances for success. If you wait until there is one out and then attempt a steal, you get to combine some chances. To calculate how that works we would have to know:

1. The percentage chance of a hit scoring the runner.
2. The percentage chance of a hit moving the runner to 3B.
3. The percentage chance of the runner being moved to 3B on an out.
4. The percentage chance of the batter not moving the runner but making an out.

I have none of those numbers.

[AztecBill]

Cabrera stole 3B with 2 outs twice yesterday. The first one unsettled the pitcher so much he gave up a home run (stretch) and the other steal resulted in a run when the 3.8% kicked in and he scored on an error.