Correlation between stats and batting order

[keystone12]

I've been wondering for a while what stats correlate to batting holes. What's the optimal stats for each batting hole.

1. OBP of course but what else?
2. OBP and?
3. OPS and?
4. OPS and?
5. ?
6. ?
7. ?
8. ?
9. ?

[HHReloaded]

There really isn't one, because it's been proven that batting order has nearly no impact on a team's success.

[Mariano_Rivera]

I've been wondering for a while what stats correlate to batting holes. What's the optimal stats for each batting hole.

1. OBP of course but what else?
2. OBP and?
3. OPS and?
4. OPS and?
5. ?
6. ?
7. ?
8. ?
9. ?
EQA in every one. For a batting order what teams should do is order their best hitters in order to give them the most PA's

[Ubiquitous]

There really isn't one, because it's been proven that batting order has nearly no impact on a team's success.


Has that been proved?

I think what has been proved is that there isn't much difference between optimum and what is being done now. Since baseball through trial and error and years of play have settled upon a lineup that is pretty close to optimal.

[Tango Tiger]

EQA in every one. For a batting order what teams should do is order their best hitters in order to give them the most PA's

Wrong. The 1st, 2nd, and 4th slots should typically have your best hitter, but ordering from best to worst is not the right thing to do.

[Xeifrank]

what the two of you suggested are both pretty similiar. You are suggesting putting your best hitters near the top of the order, the same as he did. You are just ommitting one of your top 3 hitters from the number three hole. I assume this is because the person batting third is least likely to lead off an inning over the course of a game/season. For every spot in the batting order a player is dropped he loses 18 plate appearances (if he were to play every day) over the course of the season. 36 to 54 plate appearances add up when you consider some of the scrubs some managers like to put at the top of their batting lineup. But I agree with the main premise in this thread that batting orders have a minimal effect considering that nobody is batting their pitcher at the top of the order. It's probably more effective to just bat lefty, righty, lefty, righty etc... to create havoc with LOOGY and right specialist pitchers. vr, Xei

[Tango Tiger]

I wouldn't say they are pretty similar. I wouldn't put Pujols as leadoff hitter, or Howard, and any big bopper, and on most teams, the big boppers are the best hitters. Ichiro, if he's the best hitting on the Mariners, would be the leadoff hitter.

Here's a "how to" to setting up your batting order:
http://www.insidethebook.com/ee/index.php/site/comments/optimizing_the_batting_order/

The impact is runs created is 5-15 runs per season. That's either a little or alot, depending on your point of view. The adjective is of course irrelevant, since we can ascertain the number.

[SABR Matt]

5-15 high leverage runs can be 5 wins in a season. Give the Reds 5 wins and they're inj the post-season...not the Cardinals.

[Tango Tiger]

In my case, I mean 5-15 random runs, and therefore, would have no additional leverage.

The lefty/righty, which would impact leverage, but would have no gain or loss in random runs, other than a smidge, could generate say an extra 0.5 wins. Philly batting three lefties in a row is inexcusable.

[SABR Matt]

I'm not sure it's true that the runs will be unleveraged, Tango. If you are clumping your best hitters together high in the batting order led by a high OBP threat, I believe it will be more often true that that part of the order will hit in higher leverage situations than average...even if the difference is minor.

[brett]

I do think that hitters may HIT better if put in a certain spot in the order. I think it is psychological. A slow footed power hitter does not want to bat in front of faster guys. Also, you may want somebody up to take pitches in the first 2 spots just to see what the pitcher is doing.

If the percentage stats are not affected by the positional move then I don't think order matters that much.

[Xeifrank]

I'm not sure it's true that the runs will be unleveraged, Tango. If you are clumping your best hitters together high in the batting order led by a high OBP threat, I believe it will be more often true that that part of the order will hit in higher leverage situations than average...even if the difference is minor.

[Not wanting to put words into Tango's mouth] I believe that by high leveraged runs he means runs that were scored that drastically increased that team's win expectancy. Like a run in the bottom of the 8th to break a 3-3 tie would be. By random I would think he meant, what inning those 5-15 runs occurred in. To see what an additional 15 random runs (take the high limit) would do for a team add 15 runs to a team and run the pythagorean (or a similiar algorithm) on it. A team that scored 750 runs and gave up 750 runs would have a pythagorean of .500, add 15 runs to their tally and they would have a pythagorean of .510 which is good for another 1.6 wins. The extra 15 runs would benefit a team with a low pythagorean more than a team with a high pythagorean but not by much. 1 to 1.5 wins could've made a difference in a couple of division races this year, so I guess every little bit helps.
vr, Xei

[SABR Matt]

Um...that's also what I was talking about. I was talking about leverage index. I believe that a good line-up contstruction will guarantee that your best hitters are hitting in high leverage situations more often than average because of the way those hitters function toegether as a group and becasue of the way you utilize your pinch hitters and L/R match-ups to gain advantage if you're not Mike Hargrove and therefor incapable of using your bench correctly.

[rockiesfan4ever]

1=OBP
2=OBP & AVG.
3. AVG & HR
4. HR
5. AVG & OPB & HR
6. AVG.
7. AVG.
8. AVG.
9. Pitcher

[plask_stirlac]

So the third hitter will lead off an inning less often than a second hitter?

I play sim baseball and... well I like to have two great hitters at 3 and 4, but it's also pretty effective to get a leadoff guy who can get on and a ferocious hitter at no. 2 like Darryl Strawberry or Jason Giambi, who might come in with .550 slugging and then pace the team in that category. But having someone more traditional like Alan Trammell (not '87) or Mark Loretta still works.

[Tango Tiger]

I do think that hitters may HIT better if put in a certain spot in the order. I think it is psychological. A slow footed power hitter does not want to bat in front of faster guys. Also, you may want somebody up to take pitches in the first 2 spots just to see what the pitcher is doing.

If the percentage stats are not affected by the positional move then I don't think order matters that much.

I like that you question it, rather than asserting it.

In THE BOOK, I do find some evidence that hitters hit best in the 5th slot and worst in the 9th slot (excluding pitchers), likely for the reason that the pitcher might be worn it by the the time he gets through #3/4, and is cruising through the #7/8. I didn't look too hard into it, but is definitely a study worth pursuing on its own.

Also in THE BOOK is that basestealers disrupt the batter.

It's not a stretch to suggest that there are all different kinds of profiles of runners or batters ahead of you that will help/hurt the batter at the plate.

We don't know the extent of that, and that's what we should look for.

[Tango Tiger]

I'm not sure it's true that the runs will be unleveraged, Tango. If you are clumping your best hitters together high in the batting order led by a high OBP threat, I believe it will be more often true that that part of the order will hit in higher leverage situations than average...even if the difference is minor.

My guess is that the LI will be similar, regardless of the ordering of the hitters. Even if they are different, I don't know whether it would be higher if the good hitters are clumped or if they are spread out.

In LI, every little bit helps. After all, an LI of 1.00 and 1.01 seems pretty minor. But, that's taking 700 runs scored and converting them, with leverage, to the equivalent of 707 runs scored. Those 7 extra runs, in optimization, are huge.

I'm not dismissing the idea, but I just don't know the effect one way or the other, or the extent of the impact.