Batter's Box Interactive Magazine - Striking out, Less than 2 outs

Over at BTF, there was a little disagreement, at which point the author of the linked piece said: You love the DIPS, the PECOTA, the whatever. I don't. I am concerned with wins, losses, what guys hit with RISP, and the amount of times guys strike out with men on third and less than two out.

I'm excerpting those comments from the author/poster, heals9, as a way to make a point, and not to attack anyone.

I'm concerned about the same things as heals9, as well as loving DIPS, PECOTA, whatever.

Let's tackle the "strikeouts with men on third and less than 2 outs". Is there a way to quantify this, rather than leave it to our memory and anecdotes? You betcha.

If you go to my site, let's go down to the man on 3b and 0,1,2 out rows. Look for the "out" and "K" columns.

With 2 outs, the run value of an out and a K are virtually the same, according to the chart. They are in fact exactly the same, but, because of sample size issues, it comes out only as close.

With 0 outs, the run value of an out is a loss of .28 runs, but a K is a loss of .42 runs. That's a whopping .14 run difference here. That's the difference between the best hitter in the league and an average one. So, a K is a killer (to the hitter) here, and it's certainly something that you want to know about.

But, it could be worse. That K could have come with 1 out, where the run value of the out is a loss of .25 runs, while the K is .63 runs, or a difference of .38 runs! If .14 runs is a whopper, this is downright inexcusable. K-ing with 0 outs still leaves you with the runner on 3b and less than 2 outs. But a K with 1 out leaves you in a much worse position, and hence the truly incredible turn of events that a K gives you in that position (which is also why we should love relievers with high K rates and men on base).

To me, that link I show is what baseball is almost all about. It quantifies the impact of every event in various states. Including the inning, score and count would make it almost perfect.

I am concerned with wins, losses, and K's with men on 3b and less than 2 outs. I supplement that by quantifying that concern.

There's nothing wrong with doing that, and there's nothing wrong with not doing that. To each his own, and don't let one person's perspective supplant another's.

(Note, above post also made at BTF.)

Posted by TangoTiger on Wednesday, June 15 2005 @ 11:24 AM EDT.

Striking out, Less than 2 outs | 12 comments | Create New Account

The following comments are owned by whomever posted them. This site is not responsible for what they say.

Mike Green - Wednesday, June 15 2005 @ 12:18 PM EDT (#119669) #

You're very kind, Tango. I think the poster was speaking of starting pitcher's wins and losses, and that's a whole 'nother story from King with less than 2 outs. There is some correlation between starting pitcher's wins and losses and their real effectiveness, but it is quite weak, as you know.

What the poster doesn't seem to understand is that starting pitcher's "wins" and "losses", unlike team wins and losses, is a constructed stat, in the same way that PECOTA and DIPS are. It's an old unreliable constructed stat. Branch Rickey would have understood that, and hired Allan Roth to get him more meaningful information.

There is nothing wrong with searching for qualitative descriptions of things, to supplement and in some cases to supplant quantitative ones. But, to presuppose that old quantitative constructions are better than new ones is dumb.

Mike D - Wednesday, June 15 2005 @ 12:31 PM EDT (#119671) #

I know I don't have a lot of supporters for this point, but here goes...

Note that this demonstrated difference is significant even when strikeouts are compared to outs. If you take a "reverse-DIPS" perspective, like I am tragically wont to do, then this chart actually underestimates the negative effect with no-out or one-out K's with men on third, because striking out keeps the ball out of play where all kinds of good things can happen. In other words, you lose the opportunity for a ball-in-play hit plus you make a demonstrably worse out when you K with a man on third and 0 or 1 out.

I agree with Tango's conclusion exactly: Long live the high-K reliever.

Mike Green - Wednesday, June 15 2005 @ 01:23 PM EDT (#119675) #

Mike D,

As you know, I agree with you. However, Tango's chart accounts for all the factors. The run expectation chart includes the effect of all possible outcomes.

High K/High opposition OBP is not a good profile for a reliever despite the advantage obtained in the runner on third/less than 2 out situations. This advantage is more than offset by the disadvantage in other situations.

TangoTiger - Wednesday, June 15 2005 @ 01:39 PM EDT (#119678) #

Right, the correct way to do it is to do the LWTS by each of the 24 base-out states. Then, you figure out the frequency of each of those 24 states, and then, voila, you get your custom LWTS chart.

For hitters, it won't make that much of a difference, which is why there is limited gain in optimizing a batting order.

For relievers, there's some wiggle room, since the manager can select the entry and exit point of a reliever.

Mike Green - Wednesday, June 15 2005 @ 01:55 PM EDT (#119680) #

For relievers, there's some wiggle room, since the manager can select the entry and exit point of a reliever.

There is some wiggle room, but it's a fairly heavily constrained choice- lack of options due to prior usage, changes in effectiveness due to handedness of upcoming order, and so on. I doubt that there is a reliever in the big leagues who faced more than 10% of his batters with a runner on third and less than 2 outs. I'll bet Tango knows the answer to that one too!

TangoTiger - Wednesday, June 15 2005 @ 02:02 PM EDT (#119684) #

I don't know the answer, but I'd bet there are. The average pitcher faces about 11% of his batters with a runner on 3B (regardless of outs). So, I think it's easy to see a situation where a pitcher would face more than 10% of his batters with runner on 3B and less than 2 outs.

The larger point would be that this doesn't come into play too much, that it probably is not a major concern... until it happens!

Jonny German - Wednesday, June 15 2005 @ 02:12 PM EDT (#119686) #

I am concerned with wins, losses, what guys hit with RISP, and the amount of times guys strike out with men on third and less than two out.

And over here in left field, I'm concerned that the word "amount" is now mis-used so frequently that soon we'll all forget that usage as seen above is incorrect. You can't have an amount of times. They're integers.

Jonny German - Wednesday, June 15 2005 @ 02:17 PM EDT (#119687) #

For hitters, it won't make that much of a difference, which is why there is limited gain in optimizing a batting order.

Is there significantly more gain in making a Left-Right alternating lineup? Assuming it's practical, of course... obviously you're not going to bench superior hitters simply based on their handedness.

Mike Green - Wednesday, June 15 2005 @ 02:23 PM EDT (#119688) #

Absolutely, Tango, and you'd certainly hope and expect that K rate would be a huge factor in a manager's decision on who relieves with a runner on 3rd and less than 2 out.

TangoTiger - Wednesday, June 15 2005 @ 02:41 PM EDT (#119693) #

Johnny, yes, definitely, on a per-PA basis. However, that issue will only come up, at most, once or twice a game, and not in every game. I'd go lefty-righty, especially when going 5-6-7-8.

Craig B - Wednesday, June 15 2005 @ 08:23 PM EDT (#119716) #

How many situations per year are we talking about here? Just to give me a rough idea...

TangoTiger - Thursday, June 16 2005 @ 12:25 PM EDT (#119772) #

I never looked at how often the lefty-righty thing comes up, if that's the question.

Striking out, Less than 2 outs | 12 comments | Create New Account

The following comments are owned by whomever posted them. This site is not responsible for what they say.