Power Shift: Homers and the Batting Order

So far in 2012, more major league homers have been hit from the third spot in the order (346 homers) than the fourth spot (337 homers).  It is quite rare over a full major league season for the cleanup spot not to be the place in the batting order with the most dingers.  Over the past seventy full seasons, only in 1955 and 2001 has any spot in the order other than cleanup been the source of the most homers in the majors.   But the percentage of homers that are hit from the fourth spot, as compared to the rest of lineup, has been headed generally downward for decades.    Some details after the jump.

During the 1930s, 22.2% o all major league homers were hit from the fourth spot in the lineup.  That percentage has never been so high in any decade since:

1940-1949: 21.9%
1950-1959: 19.1%
1960-1969: 19.8%
1970-1979: 19.5%
1980-1989: 19.2%
1990-1999: 19.1%
2000-2009: 17.5%
2010-2011: 16.8%
2012 through June 22: 16.2%

Except for the one jump up from the ’50s to the 60’s, the trend has been generally downward since the 1930s.

If fourth-spot hitters have been hitting a smaller percentage of homers, some other spot or spots must be gaining percentage.  In large part, the gainers seem to be hitters at the bottom of the order:

Percentage of Major League homers hit from the 7th, 8th and 9th spots in the lineup:
1930-1939: 17.6%
1940-1949: 18.2%
1950-1959: 20.2%
1960-1969: 19.4%
1970-1979: 20.1%
1980-1989: 21.6%
1990-1999: 21.1%
2000-2009: 22.5%
2010-2011: 23.3%
2012 through June 22: 24.2%

Before 1950, fewer than 1 in 5 homers were hit from the bottom three spots in the order. This season, MLB is close to 1 in 4 homers being hit from those bottom three spots. Although major league home run hitting generally has declined from its peak in the 1999-2001 period, the trend of home runs coming increasingly from all parts of the lineup continues. Equality in income and wealth in America may be declining, but egalitarianism continues to be the trend in major league home run hitting.

Leave a Reply

81 Comments on "Power Shift: Homers and the Batting Order"

Notify of
avatar
Sort by:   newest | oldest | most voted
e pluribus munu
Guest

Interesting stuff. With the coming of the DH, it would be natural to expect an increase at the tail of the order, since something approaching half of pitcher PAs would have been replaced. What surprises me most here is that the effect isn’t more profound.

Hartvig
Guest

I would guess that there’s a correlation between this and the increased number of strikeouts. There are always a few guys around in the Ichiro mold- Brett Gardner & Juan Pierre also come to mind but it does seem that there are more of the Jose Valentin/ Alex Gonzalez #1 middle infield types around than there used to be- low BA & OBP and a ton of K’s but generally in the 20+/- home run range. I can’t think of any Walt Weiss types- no power but decent OBP- playing now.

Timmy Pea
Guest

Juan Pierre hit a home run yesterday, a three run bomb. He also stole 2 bases and has 16 for the season. He’s only been thrown out twice. I don’t know what else one man can do. Juan has only one error this year, he deserves his first all-star appearance. I’m going to send an email to the NL all-star manager requesting he pick JP for the team.

John Autin
Editor
The first-ever 3-run jack by Juan Pierre should be occasion for a national holiday! Maybe, because of his name, Congress isn’t sure that he’s “one of us.” Or maybe they’re waiting for when he finally gets that grand slam. ‘Course, I didn’t work today, anyway. 🙂 There’s still a few spots to fill on his Bingo card: Out of 17 career HRs, there are only 3 scenarios in which he has ever gone deep: — Bases empty (13) — Man on 2nd only (3); and — Men on 2nd & 3rd (1). For the other situations — [1–], [12-], [1-3]… Read more »
Paul E
Guest

Have to agree with you, Hartvig, on the concept of middle infielders providing more punch. In the 1960’s most of them slapped the ball around, hit and run, take a pitch, sacrifice bunt, etc….

Middle infielders eventually evolved into larger athletes and seem to be able to drive the ball much better nowadays

MikeD
Guest

I’m guessing it’s because more players hit for power, thus reducing the overall percentage of HRs from the traditional slugger’s spot. It’s not surprising nowadays to see players hitting 20+ HRs down in bottom third of the order, where that would have been more rare in decades past.

Lawrence Azrin
Guest

In 1977 Butch Hobson, the 3Bman for the Red Sox, hit 30 HRs batting mostly 8th (89 G) or 7th (47 G). He batted higher than 6th only 5 games that season.

Jimbo
Guest

Teams want their best hitter getting more PA’s. I like it.

I also like having my slugger bat in the 1st inning, everytime. As opposed to when he bats 4th, and often ends up leading off the 2nd inning.

bstar
Guest

While that’s true, the problem, as stated in The Book, of batting your slugger third is he’s going to come to the plate too often with no one on base and two out in the first inning. The Book actually advocates batting your best hitters 1st, 2nd, and 4th for this reason.

Jimbo
Guest
I have an unconventional theory I guess that the batter will be aided by knowing he is hitting in the first inning, as opposed to preparing to hit, then going out to field (often in the outfield), then jogging in and quickly preparing to hit again. I just think that batting third gives the batter a better rhythm/timing thing. I guess the point is moot though, because if one player benefits from this, another player suffers from it. I also like the idea of increasing your chances of scoring in the 1st inning, as it sort of sets the tone… Read more »
Jimbo
Guest

This is the same as why I dislike the way closers never know which games they will pitch, and have all this uncertainty as to when or if they will come into a game.

no statistician but
Guest

Question: other than the hypothetical “winning run,” which is the most important run of the game?

Answer: The first run. It doesn’t just set a tone, it settles all shutouts and serves as a basis to keep the lead or shortens the gap when you fall behind. I don’t have any statistics to prove this, although I’ve seen them at various times. I think I’ve also seen stats that say the first score is more important in baseball than football.

Can anyone substantiate or shoot down these assertions?

e pluribus munu
Guest

Like, you, nsb, I’m ns but it seems logical that since a certain number of games end 1-0, any aggregate measure of this question would have to wind up with the answer you propose. (But if this reasoning is wrong, I’ll learn something about statistics when a statistician points out why.)

bstar
Guest

RE: increasing your chances of scoring in the 1st inning

Batting your best hitters in the 1-2-4 slots is all about the first inning. If you bat your best hitter third, you waste him too many times with two out and no one on in the first inning. Your run expectancy over the course of a season will be higher if you bat him 4th instead of 3rd.

Richard Chester
Guest

Perhaps it makes sense to have the batter with the higher OBP hit third rather than fourth. Then when the two-out-no-one on situation arises there would be a lesser chance for both the number 3 and 4 hitters to bat with no one on.

Jim Bouldin
Guest
Interesting question, that I’ll bet Bill James must have flogged to death. Seems to me that there’s some clear tradeoffs when it comes to the 3 vs 4 spot. While it’s true that sometimes neither of the first two guys will get on (an argument for batting fourth, to allow a chance for the 3rd hitter to get on), this is offset to an unknown degree by the fact that sometimes none of the first 3 will get on, which guarantees that he bats with the bases empty the next inning. The other argument against the 4 spot is that… Read more »
Jim Bouldin
Guest

that should be:
“…and those PAs that are sacrificed are *always* the last one”

Richard Chester
Guest

Just for the record in 2011 the number 3 batter had an average of 16.6 more PA than the number 4 hitter.

bstar
Guest
That number seems about right where it should be. Batter #3 should get one more plate appearance than #4 once every nine games if you assume that games end at each batting position in an even distribution over time. But games are a little less likely to end at the 3 or 4 spot than 8 or 9 since your 3-4 hitters are better than 8-9 and less likely to end the game in their final at-bat. But if you do assume it’s evenly distributed, games should end at each position once every (162/9) = 18 games. So 16.6 for… Read more »
Richard Chester
Guest

bstar:
That stat in my post 28 is for 2010.
Here are the stats for the ML in 2011 for the various positions in the line-up (differences in PA over the course of a season):

Between 1st and 2nd: 17.6
Between 2nd and 3rd: 17.6
Between 3rd and 4th: 16.4
Between 4th and 5th: 17.3
Between 5th and 6th: 17.6
Between 6th and 7th: 18.3
Between 7th and 8th: 19.9
Between 8th and 9th: 18.6

Between 1st and 9th: 17.9 (not 18 because there are extra innings which serve to increase that number and home team wins with no bottom of the ninth which serve to decrease that number).

Richard Chester
Guest

In my post 32 that 17.9 figure between 1st and 9th is an overall average figure between adjacent positions in the line-up.

bstar
Guest

Richard, thanks for the research. I am traveling back home tomorrow where my copy of The Book is and I will try and find some passages and numbers that detail why they came to the conclusion that 1-2-4 is where you want your best hitters in an optimal lineup.

Tristram12
Guest

Per The Book (and Bill James before it), the ideal type of batter for the 3-spot is a lower OBP, higher SLG type (think Kingman/Canseco). Since this batter comes up a significantly disproportionate number of times with 2 outs and no base runners, you really don’t want to waste a walk or a single in the situation. You want a batter who, if they do something, either hit a HR, or get themselves into scoring position. Two outs is not a good time to start a rally that needs 3 (or more) consecutive baserunners.

Tristram12
Guest

Looking at JAs favorite team, the ideal lineup for the Mets (just using their slash line for this year) would be; (1) Tejada (363/404) (2) Duda (353/442) (3) Hairston (307/542) (4) Wright (455/565) (5) Nieuwenhuis (339/424) (6) Murphy (316/343) (7) Thole (325/315) (8) Davis (275/375). This ignores speed, R/L/S, ego, history, etc. FWIW, the #8 spot is similar to #3 in that it also has a disproportionate amount of two out ABs, and therefore should utilize a low-OBP slugger (like Ike Davis).

Tristram12
Guest

As for the question about putting your second best hitter #2 or #5; While 16-18 ABs are not enough to overcome the number of two-out situations of the 3-slot (thereby justifying having your best hitter 4th instead of 3rd), the 50-55 ABs between #2 and #5 are significant enough that you want the better hitter #2 instead of #5.

Richard Chester
Guest

I think that the statement in parentheses in my post 32 should be reversed or perhaps even ignored.

bstar
Guest

Great stuff, Tristram, thanks!

Jim Bouldin
Guest
Great topic, and clearly one lending itself to some combined simulation and real data analyses and general geekery and possibly even some mild name calling. bstar, I see your point now. If we take a hypothetical OBP of .333 for each of three hitters who potentially bat 1-2-3 in the order, assume no dependencies of those OBPs on the situation at hand, and all other things equal, then a power hitter: 1. Put into the 3rd spot comes up with nobody on .444 of the time: 1 – [p(one guy on) + p(2 guys on)] = 1 – [2*(2/3*(1/3)^1) +… Read more »
Richard Chester
Guest

Reply to #58:
Jim: would you run those numbers again for these situations: Batters 1 and 2 have OBPs of .333. There are two sluggers being considered for the nos. 3 and 4 slots. One has an OBP of .333, as in step 1 of your post, and the other has an OBP of .400. Which slugger should hit 3rd?

Jim Bouldin
Guest
Richard, I get the following: 1. If the .333 obp slugger bats 3rd then the results are as before, that is, nobody gets on in the first inning .296 of the time. 2 If the .400 obp slugger bats 3rd, then nobody gets on 0.267 of the time: a = 1/3 = obp of first two hitters b = 2/5 = obp of third hitter (slugger) then, p[at least one gets on] = p[one guy gets on] + p[two guys get on] + p[3 guys get on] = [2*(a*(1-a)*(1-b)) + (b*(1-a)^2)] + [2*(a*(1-a)*b) + ((1-b)*a^2)] + [(b*a^2)] = .733 and… Read more »
Jim Bouldin
Guest
I also forgot to mention in my 12:18 post, the offsetting fact I mentioned before. That is, although your slugger is coming up with the bases empty 15% less often if he bats 4th (.444 – .296 = .15), this only applies to the first AB of the game, so over the full season that’s 162 * .15 = about 24 ABs. So the gain of 24 ABs hitting with runners on in the first AB must be weighed against the loss of (162 * 1/9) = 18 total ABs over the full season, all of which are lost in… Read more »
Jim Bouldin
Guest

Also, I’d have to disagree strongly with the idea from “The Book” of putting a poor OBP slugger in your #3 spot. I can’t see that at all. In fact, I’m pretty sure the opposite would be true: you want to bat your highest OBP guy right in front of your slugger at the #3 spot, because that guy is the one most likely to bat in the same inning as the slugger after the first time through the lineup.

e pluribus munu
Guest

Given the scale of absolute numbers in play here, GIDP should probably be taken into account. If you’re putting Aaron or Yaz into the order at #3, it’s worthwhile thinking about how many times they’ll end a man-on-first one-out first inning before Mathews gets up or before . . . oh, I was going to say before Rice gets up . . .

dannyc
Guest

Most do not want to admit it but it was LaRussa who popularized
moving the beat hitter from the 4 hole to third

AlbaNate
Guest

Didn’t Babe Ruth bat third? 🙂

no statistician but
Guest

Williams, Mathews, Musial, Mantle until 1960, Aaron after Mathews was gone, Schmidt quite frequently, Brett almost always—in the third spot.

Lawrence Azrin
Guest
As several others have pointed out downthread, the team’s best hitter has traditionally batted #3 going back 100+ years, at least to the early days of Cobb and Speaker. I recently read a history of the 1912 Red Sox where early in his career as a regular, Speaker threw a major fit because he had a bit of a slump and was removed from his customary #3 BOP slot. So batting #3 was kind of a “prestige thing” even back then. I don’t think the cleanup BOP was as clearly defined in early MLB history, since HRs were relatively scarce… Read more »
John Autin
Editor

Apologies for the hijack, but nobody’s on chat and I have to get this off my chest.

After the Yanks loaded the bases against Dickey for the second time this game, I checked — nobody had loaded the bases against Dickey all season.

But now, the bubble has burst … 4-0 Yanks.

bstar
Guest

JA, if you’re still lurking Evan and I are actually forming a conversation in chat. It’s 10:10 eastern.

Tmckelv
Guest
I was intrigued by the 1950’s on this list and why that decade was the only one against the trend. At first I wondered if it could be 1 player that hit a lot of home runs and batted 3rd (like aaron or mantle – williams and musial played too much in the 1940’s to have an effect, theoretically). So I checked the home run leaders for 1950 thru 1959, and I noticed something interesting (to me). The top 2 were Duke Snider and Gil Hodges. I thought I remembered that neither of them were the primary cleanup hitter for… Read more »
Jim Bouldin
Guest

In thinking about this topic, I was surprised to discover that the minimum number of guaranteed PAs for the #8 and 9 hitting slots, over a full nine inning game is…
two!

I’ll bet it’s never happened though.

Richard Chester
Guest

On 6-22-06 at Chicago the White Sox beat the Cardinals 1-0. Two Sox reached base, Jermaine Dye on an error and Jim Thome on a HR. There was no bottom of the ninth and SS Juan Uribe, the #9 hitter, had only 2 PA.

Richard Chester
Guest

And on 7-25-1992 the Braves beat the Pirates at home 1-0 and #8 and #9 hitters Mark Lemke and Charlie Leibrandt had just 2 PA each. I should have bet you before I submitted these two posts.:-)

Jim Bouldin
Guest

Ha, ha, never say never, as they say. I should have known…thanks for digging that up!

Richard Chester
Guest

There have been more games with players having 2 PA.

Doug
Guest

Richard’s Braves game in 1992 is the only one with both the #8 and #9 hitters only batting twice.

The #9 hitter has batted only twice 47 other times. These are the games.

http://bbref.com/pi/shareit/7TKq2

Doug
Guest

One of those games was from earlier this year – Atlanta faced only 26 Cubs on May 9 despite allowing 5 hits.

The White Sox have twice faced only 26 batters while allowing 6 hits, on 04/25/93 and 06/23/67.

John Autin
Editor

Good post, by the way, birtelcom. An interesting topic. Have you thought of doing a similar look at, say, stolen bases?

John Autin
Editor

Also, I’m hoping there’s a way to collect this kind of data other than slogging through the MLB splits pages, one year at a time. Any tips?

birtelcom
Guest

The easiest way to do batting order studies is if you can do them with the Event Finder. If you can grab 10 years of home runs, say, in one search, there in the results summary is one tidy little table that shows how many times the event occurred at each spot in the batting order. If the Event Finder won’t do for a particular search, the Game Finder will often work for batting order studies. But then you have to do one batting order position at a time.

Jim Bouldin
Guest
My optimal lineup when I get the managerial offer I know is coming, perhaps from the Red Sox: 1 Base stealer, generally smart and aggressive baserunner and just flat out fast. Above average OBP. 2 Good bat handler. Excellent bunter, hits to all fields, moves runners over. High OBP and pretty good speed. 3 Best hitter on team, and with highest OBP. Great eye, draws walks; good hitter to all fields, switch hitter if possible. 4 Your best power hitter, conditional on not striking out excessively (for a power hitter, i.e. no Adam Dunns) 5 A poor man’s version of… Read more »
Lawrence Azrin
Guest
I’d rather have the best four hitters on the team bat 1 to 4, and slot them as best I can to lead-off, clean-up etc… As Richard Chester breaks down in #32/#33 above, the average decrease in PA/year going down the batting order is about 18 PA, so I’d rather not give put my 6th/7th best hitter in the #2 slot and give him a bunch more PAs just because he’s a good bunter and can run well. Some power isn’t a bad thing in the #2 BOP (see Robin Yount, 1982). I do agree that speed in the #9… Read more »
Jim Bouldin
Guest
My #1 and #2 guys are designed to harass the pitcher basically, the #1 on the basepaths, and the #2 at the plate, by fouling off lots of pitches, obstructing the catcher’s view during an SB attempt, successfully hitting and running, and then creating some of his own havoc on the basepaths when he gets on. But I don’t mind a little power either and would take a Robin Yount any day that’s for darn sure. Though I might bat him 3rd, not 2nd. And actually, that’s not really my ideal lineup, because I wouldn’t have an Adam Dunn type… Read more »
Phil Gaskill
Guest

> Some power isn’t a bad thing in the #2 BOP (see Robin Yount, 1982).

Or see Curtis Granderson right now.

John Autin
Editor

I’m just guessing, but this might be the best season ever by a full-time #2 guy:
http://www.baseball-reference.com/players/split.cgi?id=matheed01&year=1959&t=b#lineu::none

168 OPS+, .306 BA, 46 HRs, 114 RBI.

e pluribus munu
Guest
But a lot of Mathews punch came when leading off an inning: 116 PAs: .400/.457/.743 (10 HRs, but somehow only 10 RBIs . . . oh, right) – his truer #2 stats would be .289/36HR/104 in 552 PA. In ’59 Haney batted Mathews/Aaron/Adcock in #2-4 (presumably because he lost Schoendienst suddenly at the start of the season). The previous two years, Haney generally batted Mathews and Aaron #3/#4 and won pennants (although in ’57 he had Aaron in #2 one-third of the season, and Aaron’s record was super there, 1.000+ OPS). In ’59, Haney wound up losing the pennant with… Read more »
bstar
Guest

birtelcom, you mentioned the other day using the Event Finder to find batting order stats. Is this what you used to find Sandberg also? For the uninitiated, could you go into a little more detail about exactly how one would search for #2 batting order stats?

Jim Bouldin
Guest

“they were, however, 8% grittier and had 13% more intangibles.”

🙂

no statistician but
Guest

Red Rolfe’s 1939 season is also outstanding: 139 R, 213 H, 46 2B—all leading the league, with .329 BA, .404 OBP, .495 SLG. Rolfe wasn’t a power hitter, so he didn’t get the HR boost to his slugging.

Ed
Guest

Here’s an old post by Raphy showing most games from the #2 spot. Not sure how he did the search.

http://www.baseball-reference.com/blog/archives/1299

Richard Chester
Guest

Reply to post 77:

My search shows Eddie Yost with 1728 games from the lead off spot which does not show up on Raphy’s list.

Richard Chester
Guest

Reply to post 77: Here’s a current list of most games for a #2 batter.

1 Nellie Fox 1711 Ind. Games

2 Omar Vizquel 1569 Ind. Games

3 Ozzie Smith 1529 Ind. Games

4 Jay Bell 1319 Ind. Games

5 Derek Jeter 1304 Ind. Games

bstar
Guest

@75 bitelcom, thanks for the details of that. I am going to bookmark this page to refer to some of those tricks you suggested. Thanks.

bstar
Guest

make that “birtelcom” above

wpDiscuz