Cleaning Up Becomes a Chore

There’s been discussion here and elsewhere about the decline in run scoring across the majors this season and in recent seasons. Yes, across the majors the average runs scored per game is currently at 4.14 so far in 2014, down slightly from 4.17 for the full 2013 season, and from 4.32 for the full 2012 season. OPS (On Base Percentage plus Slugging Percentage) across the majors as a whole is down from .724 in 2012, to .714 last season, to .707 so far this season. But the decline in hitting performance has not been uniform across the batting order, and in this post I want to focus on the particularly dramatic drop in the average performance of clean-up hitters in the majors.

The number that baseball reference calls tOPS+ provides a guide to the relationship between OPS of a particular “split” during a particular season, and overall OPS that season, with a split that exactly matches overall OPS being set at 100. Prior to this year, over the previous century of major league baseball (1914 through 2013), the fourth spot in the batting order has averaged 128 tOPS+ over a full season (for comparison, the third spot in the order has averaged 123.8, the fifth spot has averaged 113.9, the eighth spot has averaged 84). Since the adoption of the DH in the AL in 1973, the average tOPS+ for cleanup hitters over a season has been 123.8 (down from 130.8 over the 1914-1972 period). It’s been a little lower over the past ten full seasons (2004-2013) at 122.3.

The best OPS years for cleanup hitters, relative to overall OPS, have been 1961 and 1936; in both of those seasons fourth-spot batters across the majors compiled a collective 140 tOPS+. The lowest tOPS+ for cleanup hitters over a full season has been 117, which has occurred twice, in 1976 and 1985. But this season, with the 2014 All-Star Break having arrived, and with close to 60% of the regular season complete in terms of games played, the collective tOPS+ for cleanup hitters across the majors this season is an astonishingly low 112, easily the lowest of the last 100 seasons. The drop is dramatic just in the last two years — from a 124 tOPS+ for 2012 to 120 for 2013 and now down to the 112 level thus far in 2014.

Let’s look at some raw numbers to add some additional detail to the relative abstraction of tOPS+. OPS for cleanup hitters across the majors averaged .812 in 2012, fell quite a bit to .789 last season, but this season the bottom has dropped out as clean-up hitter OPS in the majors this year has averaged just .752. The only full seasons over the past 100 that fourth-spot hitters have had a lower raw OPS have been 1968 (.745 clean-up spot OPS), 1976 (.739) and the deadball era years before 1919.

As recently as 2008, fourth-spot batters had the highest OPS of any place in the batting order:
4th spot: .840
3rd spot: .825
5th spot: .783

In 2012, fourth spot and third spot batters were very close in terms of average OPS:
3rd spot: .813
4th spot: .812
5th spot: .758

In 2013, the third spot has opened up more of a lead in OPS:
3rd spot: .798
4th spot: .789
5th spot: .741

Now, thus far in 2014, cleanup hitters have lost, at least for the moment, even their claim to having the second-best hitting position in the batting order:
3rd spot: .803
5th spot: .753
4th spot: .752

What’s going on with cleanup hitters in 2014? Are we seeing a mostly random effect that will soon revert toward the mean? Are we seeing the kinds of hitters who tend to bat fourth suffering disproportionately from current trends in the game — an increase in strikeouts without a commensurate increase in walks or power? Are we seeing a change in strategy by managers in who they tend to bat in the clean-up spot in the order? Is the term “clean-up spot” itself becoming obsolete? I offer no answers at the moment, just questions to ponder here at All-Star break time.

Lets look at some raw

0 0 votes
Article Rating
Subscribe
Notify of
guest

31 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Voomo Zanzibar
Voomo Zanzibar
9 years ago

Thank you:

San Diego
.187 .266 .305 .571

Houston
.199 .272 .318 .590

Doug
Editor
9 years ago

That .752 OPS seems to ring true. Looking only at the players to bat fourth most often, the median OPS of the 18 players with 200 PA batting cleanup is only .779. Starlin Castro, of all people, sits above that median with .788 OPS. Below him are such luminaries as Allen Craig, Ryan Howard, Evan Longoria, Curtis Granderson, Billy Butler, Mark Teixeira and Adrian Gonzalez. Applying that same methodology to 2010-2013, these are the number of players with 350 PA at cleanup and their average OPS (sorry, Excel doesn’t provide median with its pivot tables). 2010 – 20/.846 2011 –… Read more »

Wine Curmudgeon
9 years ago
Reply to  Doug

I think you’re on to something with the Castro bit. Why is Starlin Castro batting cleanup? Has there been a chance in the way managers approach the fourth spot?

Jim Bouldin
Jim Bouldin
9 years ago

Has to be due to higher tOPS+ values by the other 8 slots in the lineup, by definition, since the mean over all lineup slots has to be ~ 1.0. A look at how each slot is trending over the last few years should give the answer.

Power hitters’ OPS values suffer the most if there’s an across-the-board drop in making contact, which is what an increase in the K rate rate implies.

John Autin
Editor
9 years ago
Reply to  Jim Bouldin

Jim, good point about the other 8 slots. There’s definitely a long-term trend towards distributing power throughout the lineup.

And a belated welcome back! BTW, did you buy your baseball alarm clock from the Red Sox gift shop?

Jim Bouldin
Jim Bouldin
9 years ago
Reply to  John Autin

Ha ha, yep. The “All-or-Nothing” clock I believe it was labeled.

John Autin
Editor
9 years ago

Re: cleanup hitters — I don’t know if this is cause, effect, or nothing important. But the years cited for high 4th-place performance — 1936, 1961, 2008 — were also years of great consistency at the position. For N = number of teams in MLB, here’s the median of the top N players in 4th-place PAs for those years: 1936 — 470 PAs 1961 — 425 PAs 2008 — 436 PAs Now, the same median for the last three years: 2012 — 375 PAs 2013 — 348 PAs 2014 — 390 PAs (projected) I’m sure the recent PA stats partly… Read more »

Artie Z
Artie Z
9 years ago

Combining thoughts from the above posts: We’re looking at OPS splits right? What are the breakdowns by SLG and OBP? It might be that managers are moving power hitters who walk further up the lineup. What did I hear last night – the 2-spot in the batting order is the big thing right now (I think Reynolds said that when Trout came up to bat in the 2-spot)? Basically, this might mean more players like Joe Carter batting cleanup than Mike Schmidt, as the Schmidts get moved to the 3rd spot and the Votto-types to the 2-spot. It might be… Read more »

Artie Z
Artie Z
9 years ago
Reply to  Artie Z

Replying to my own thought … Mike Trout is .310/.400/.606 this year. He has 22 HRs, and he’s leading the AL in OPS, OPS+, and TBs. He’s hit 2nd in 88 games, and 3rd in 2 games. Pujols is the Angels 3rd hitter (92 starts, Trout has the other 2), and Hamilton has hit 4th the most (41 starts, with Ibanez, Kendrick, and Freese all at 15+ games), with Kendrick, Aybar, and Freese all having 15+ games in the 5 spot. 10 years ago, wouldn’t Kendrick or Aybar hit 2nd, with Trout, Pujols, and Hamilton in the 3, 4, and… Read more »

Richard Chester
Richard Chester
9 years ago
Reply to  Artie Z

A-Rod in 1996 and Robin Yount in 1982 led their leagues in TB while batting second. Each had about 90% of his TB while batting second. There have been 47 players who had an OPS of .900+ in a season with a minimum of 400 PA in the number 2 position. The OPS is only for games in which they batted second. That includes Mike Trout for this year. There have been 1400+ players overall with a .900+ OPS and 400 PA minimum, regardless of batting order position. The top 5 number 2 batters by OPS are Lefty O’Doul, 1.096… Read more »

John Autin
Editor
9 years ago
Reply to  Artie Z

Mike Trout aside, I see no change in the tendency to waste the #2 spot. Both the OBP and OPS for MLB 2nd hitters this year are not quite 3% above the overall marks. That’s not a significant change over the last 25 years; 1990-92 had better percentages.

This year, 14 teams have a lower OBP from #2 than overall, and 13 teams have a lower OPS from #2 than overall. For every Trout, there’s an Elvis Andrus, a B.J. Upton and a Gordon Beckham.

bstar
9 years ago
Reply to  John Autin

John, I’ve got to sleep at night. Please don’t suggest there’s another Upton out there for the Braves to acquire. 🙂

mosc
mosc
9 years ago

I feel like we just don’t have the variance in players in general we had before. I looked at player OPS+, going back to the #60 guy (average cleanup hitter ability approximately?) 2004: Mark Kotsay 116 2005: Emil Brown 113 2006: Hanley Ramirez 116 2007: Carl Crawford 117 2008: Brian Roberts 118 2009: Johnny Damon 118 2010: Stephen Drew 113 2011: Eric Hosmer 118 2012: Cody Ross 115 2013: Nick Swisher 116 2014: Josh Donaldson 115 So I guess that’s a failed experiment. There are a few cleanup type guys on there and there are also some productive obp light… Read more »

mosc
mosc
9 years ago
Reply to  mosc

Average DH in 2014 AL belongs to Oakland headlined in PA’s by Alberto Callaspo. Same player criteria in 2004? David Newhan slashing .311/.361/.453/.814.

Jim Bouldin
Jim Bouldin
9 years ago

This topic is more interesting than I’d first realized, as it gets at an aspect of offensive strategy that perhaps flies under the radar. I, and I think a lot of people, tend to dichotomize about that issue (i.e. as either primarily power- or contact/speed-oriented). But how hitters are arranged/dispersed through the lineup, for either strategy, is also potentially quite important. I think it’s useful to remember that each slot in the batting order gets, on average, about 162 x (1/9) = 18 more PAs over a season than does the next lower slot. When playoff races are tight, as… Read more »

Brendan Bingham
Brendan Bingham
9 years ago

I agree wholeheartedly with Jim (#15) that optimizing batting order is a topic rich for further discussion and research. Most of what I have read on the subject points to the importance of putting productive hitters in the 2 and 4 position and NOT batting your least productive hitter 9th (7th or 8th is preferred). Perhaps some small part of the recent downturn in run production can be tied to a recent step away from employment of optimal batting orders (specifically poor productivity from the 4th position). If so… here’s a crazy idea. What else has happened in recent years?… Read more »

David P
David P
9 years ago

There’s a huge flaw with the whole concept of lineup optimization. Performance varies too much between and within seasons for it to have practical application. This is particularly true for non-superstars, which is the vast majority of the league. Cleveland makes a good example since they’re a team bereft of superstars and their lineup was very stable between 2013 and 2014 (David Murphy being the only significant addition). Here’s each players OPS+ with 2013 first, then 2014: Rayburn 152, 47 Santana 136, 110 Kipnis 132, 95 Gomes 132, 109 Swisher 116, 80 Brantley 106, 154 Cabrera 97, 99 Bourn 92,… Read more »

Jim Bouldin
Jim Bouldin
9 years ago
Reply to  David P

Good point David; you can’t optimize a lineup without confidence in how it’s going to respond over time. Having said that, I do think a lot of managers could do a better job of optimizing on shorter time frames based on who does, and does not, have the recent hot bat. This strategy will of course fail at times, perhaps even frequently, but so does sticking with a lineup that’s not working well. A topic perfect for some simulation analysis.

David P
David P
9 years ago
Reply to  Jim Bouldin

Jim – I think the best opportunities for lineup improvement lie in managers making more intelligent decisions. For example. Tonight, Terry Francona has Carlos Santana batting cleanup against Max Scherzer. That’s just plain stupid. Sherzer has completely owned Santana (.343 OPS in 40 PAs). Certainly Francona must know this and yet he decided to completely ignore the evidence. He also has Chris Dickerson in the lineup even though Dickerson is 0-9 with 7 Ks against Scherzer. At least he has Dickerson batting 9th but again, what’s the point of even having him in the ballgame? Meanwhile, Asdrubal Cabrera and his… Read more »

David P
David P
9 years ago
Reply to  David P

And of course Dickerson promptly makes me look like an idiot and Francona like a genius by hitting a home run off Scherzer.

David P
David P
9 years ago
Reply to  David P

And how a second home run by Dickerson off of Scherzer! I was hoping my #23 might inspire a bit of “reverse jinx” but that was a bit more than I expected.

bstar
9 years ago
Reply to  David P

David, I think all Francona is doing here is trusting the true talent of the player over small-sample batter v. pitcher splits. The Book looked at this in depth. They grabbed the 30 matchups which looked the most lob-sided, one way or the other (pitcher seems to own batter or batter owns pitcher). They then tested the next year’s PAs for these particular match-ups and found that the hitters, on average, hit almost exactly as they do overall vs. this pitcher, not how they had hit against that pitcher in the past. Same thing for the pitchers: the ones who… Read more »

David P
David P
9 years ago
Reply to  David P

Bstar – Well Santana made the evening complete by driving in the winning runs with a bases clearing double (not off of Scherzer however). Anyway, I’m not familiar with the study from The Book but it sounds fairly limited in scope and would need to be expanded considerable to draw conclusions (maybe that’s been done already?). As for “true talent”, that gets back to my #23. How do you know what someone’s true talent is and how much data is necessary to detect a change in talent level? BTW, I’ll be back tomorrow to provide more criticisms of Francona’s lineup… Read more »

bstar
9 years ago
Reply to  David P

David, basically in #23 you’re saying to trust the true talent of the hitters over in-season performance.

I (and The Book) totally agree with that!

But I also agree with Jim that sometimes simply making a change when a lineup isn’t working can produce positive result. (similar to changing a manager just for the sake of introducing some new energy to the team.)

bstar
9 years ago
Reply to  David P

OOOps, I meant #17, not #23.

Jim Bouldin
Jim Bouldin
9 years ago
Reply to  David P

David, the analysis bstar is referring to is just one example among many of a much broader concept that basically states that deviations from a central tendency tend to cancel out over time (“regress to the mean”), such that at high enough sample sizes, those deviations are seen as just that, deviations, not central tendencies (i.e. means or medians). But your question is good too–clearly not all matchups are the same, so how big of a sample do you need to determine the expected value in any particular one, which is further complicated by the fact that it’s not a… Read more »

Artie Z.
Artie Z.
9 years ago

Based on JA’s comments about the early 1990s, I started looking at those seasons. What I saw was that some time in the mid-1980s (it was either 1985 or 1986), the major league wide OPS numbers for the 1 and 2 hitters were higher than that of the 6th place hitter. Prior to the mid-1980s, it didn’t seem like this was a very common occurrence – typically the 6th place OPS was better than both the 1st and 2nd spots, and if it was below it was only below one of the two and not very far below (2, maybe… Read more »

Jim Bouldin
Jim Bouldin
9 years ago

Note that the issue of the drop in tOPS+ this year from the clean-up spot is largely an NL phenomenon. The AL numbers are low also, but not historically so as they are in the NL. Might be possible to look at a few key NL teams/players and detect the cause.

Richard Chester
Richard Chester
9 years ago

Jim:
I checked the differences in tOPS+ for the NL teams from 2013 to 2014. Here they are.

Cubs, +3
Dbacks, +1
Braves , 0
Reds, -5
Brewers, -10
Mets, -12
Giants, -12
Nats, -12
Marlins, -13
Rockies, -23
Phils, -27
Cards, -28
Dodgers, -29
Pirates, -32
Padres, -32

NL overall, -14

Jim Bouldin
Jim Bouldin
9 years ago

Hmmm, interesting, thanks Richard. So basically, the Rockies through Padres part of that list are largely responsible for the big drop.

trackback

[…] “just like human hair&#8221. With certain inbuilt characteristics such as being able to use high heat up to 180 degrees. Hence one can curl or tong them and then straighten them again. One can […]