Monday, October 19, 2009

Geeking Out: The #2 Batter

As the Voice of Reason™ and I are kicking back and watching her Phils tonight, I notice #2 Shane Victorino's stats: .292/.358/.445. And I thought - now there is a guy Parker would like.

Because in the TwinsCentric Offseason GM Handbook, Parker wrote a great essay on the importance of the Twins finding a solid #2 hitter, called Never Break the Chain. It starts like this:

In 1986, Bill James constructed a poignant analysis on lineup composition and revealed that the total runs scored and second spot in the batting order had the strongest correlation among any player in the lineup — more than leadoff, third or cleanup. Mr. James noted in his 1986 Baseball Abstract that “many managers tend to waste the second spot in the order by putting somebody there who isn’t one of the better hitters on the team...Too many managers will say ‘bat control’ as if these words were a magic wand, and place some .260 hitter with a secondary average of .150 batting second…”

Someday I want to compose a poignant analysis. Anyway, Parker's essay then covers the Twins woes at the second spot in the order and options they might have this offseason. To get a better sense of the essay, you can download the free 1/3 book at

I wondered what trait of the second batter was most important? On-base percentage makes sense, but so does batting average. Or maybe it's a patience-power mix, like OPS? So I ran a quick correlation between runs scored by teams and the stats their #2 hitter put up. The higher the number, the higher the correlation. Here are the results:

OBP - 0.526
OPS - 0.523
BA - 0.469
SLG - 0.446

It looks like on-base percentage or OPS are a little more important than batting average. That's not exactly ground-breaking, but it's good to know. Because I had also copied all the other stats for #2 hitters, I ran correlations on them too. Here they are:

R - 0.680
H - 0.532
RBI - 0.525
TB - 0.508
AB - 0.451
HR - 0.255
2B - 0.247
GP - 0.158
3B - -0.037

Heh. Shocking. Runs is the most important. So a team that scores a lot tends to have a #2 hitter that racks up a lot of runs. Whoulda thunk?

Still, it's nice to clarify a little what we're looking for from the #2 hitter. As we might suspect, we want them to set the table.

If anyone has any other correlations that they would like to see, make sure to enter them into the comments section. Thanks.


Anonymous said...

Joe Mauer is good at baseball. Maybe he should bat second. Span, Mauer, Morneau, Cuddyer, Kubel gives your 5 best hitters the most at bats and splits Kubel off from the lefties. Not that hard.

TT said...

Isn't this where some stat geek points out there is a difference between correlation and causation?

I suspect there is a strong correlation between teams whose players are better hitters and how many runs they score. I think its safe to say one causes the other. And the Twins will likely score more runs if they add better hitters at any position or any point in their batting order.

TT said...

"Not that hard."

Well yes it is. Because having Mauer bat second means that he and Morneau will have runners on base about half as often when they bat. Mauer hit almost 30 home runs, you would like to have more guys on base when he comes to the plate, not fewer.

It will also mean that the number nine hitter effectively becomes the leadoff hitter after the first inning, with Span batting second. You want Span to bat ahead of someone who can move him over into scoring position for Mauer and Morneau even when they don't hit home runs.

As for more at bats, one at bat per game is changed by moving Mauer up. Over half the time that extra at bat will go to someone other than the four hitters you moved up in the order.

In short, its only not that hard if you don't think about it.

rghrbek said...

Great post. Maybe this needs to be sent to PA (who is in love with Orlando) and the beat writers, who seem to think that OC is our best option going into next year.

Drives me nuts.

Anonymous said...

Runners on base divided by plate appearances for our regulars. Cabrera statitics are only after the trade.

Span 0.58
Cabrerra 0.69
Mauer 0.58
Morneau 0.70
Kubel 0.72
Cuddy 0.70
Young 0.63
Punto 0.65
Tolbert 0.63

Gardy's insitence on putting a low OBP middle infielder in front of our best hitter led to Mauer seeing the lowest number of men on base per plate appearance. Mauer converted those opportunities at the highest rate on the team.

Mauer also had fewer baserunners on 2nd per plate apperance than Kubel, Cuddy, Morneau, Cabrera. Seems like walks and hits are also a good way of moving up the runner.

I don't really care if Mauer bats second as long as Gardy places a good hitter in the two hole.

MC said...

The Runners on Base per PA is eye opening. If you lay it out in that format for the GM and manager and they refuse to make a change, then they should be dismissed.

TT said...

"The Runners on Base per PA is eye opening."

No it isn't. Its just misleading.

For one thing Mauer got almost 1/4 of his plate appearances as the number two hitter. Which means the guys at the end of the order, like Punto or Gomez, were responsible for getting on base ahead of him.

For another, the list left out a couple players with more plate appearances (Crede and Harris) and included Cabrera who only played during the teams two hottest months and Tolbert who had most of his at bats during that period.

"I don't really care if Mauer bats second as long as Gardy places a good hitter in the two hole."

You mean he should have had Joe Mauer batting in front of Joe Mauer in the number two spot? That certainly would be ideal. Because Mauer was on base a lot when Morneau, Cuddyer and Kubel came to the plate. But, as your numbers show, that also meant Mauer was batting without anyone on base a lot.

I think the obvious reason there wasn't a better number two hitter is that there wasn't a better hitter in the lineup who fit that profile.

Anonymous said...

Crede 0.66
Harris 0.56
Gomez 0.56
Casilla 0.72
Morales 0.59

Here are the numbers for the rest of the team. Joe is in a tie for 3rd lowest runners on base per PA. Casilla is tied for the team lead, YUCK! Everyone who spent significant time in the two hole had more runners on base for them.

I would love to clone Joe, but since that is not possible I would like to see him hit behind someone who doesn't make outs 68% or more of the time. Not making outs should be the "profile" of a 2 hitter.

TT said...

"Not making outs should be the "profile" of a 2 hitter."

On the Twins this year, that would be Nick Punto. Of the twins regulars, he had the next highest OBP after Span, Mauer, Morneau, Kubel and Cuddyer. Morales and Buscher did slightly better in limited duty.

"Everyone who spent significant time in the two hole had more runners on base for them.

You must mean "everyone else". But Brendan Harris had more plate appearances in the number two spot than Mauer and he had fewer batters on base for him according to your numbers. And Harris had many more at bats in the 6th and 7th spots where Mauer was probably still on base for his at bats.

The only other player besides Harris who had more plate appearances in the number two spot than Mauer was Cabrera.

Of course, another reason Mauer didn't have as many runners on base is that Span drove in 68 runs and Cabrera another 36. With 13 home runs between them, that is about 90 runners they inherited that weren't there for Mauer.

The real problems in the number two spot were Casilla and Harris. They had a combined OBP of .230 in the number two spot with more plate appearances combined than Cabrera.

Cabrera is not an ideal number two hitter. But he is almost certainly the best of the bunch this year.