The End of the Terrible Number-Two Hitter

If you’ve recently spent time with other humans, it’s likely that you noticed that they tend to be overconfident about how well they understand the world around them. Think of all of the people you know who have tried to weasel their way out of admitting they were wrong even when presented with strong evidence that they had misinterpreted a situation. Humans are bold and unapologetic in their declarations and do not like it when you point out that they’ve made a serious error.

It’s hard to criticize people for that when it seems to be a pretty fundamental aspect of the species. It’s not good or bad, it simply is. But that doesn’t mean we can’t enjoy little moments when someone makes a compelling argument and then the world totally destroys their hard work by changing around them.

For example, two political scientists once wrote a book called Congress’ Permanent Minority? Republicans in the U.S. House which was the first major scholarly account of how a minority party operates when it expects to be in the minority for the foreseeable future. It’s a well-researched book and was well reviewed when it came out. Unfortunately for the authors, it came out in January of 1994, just 11 months before the Republicans would win control of the House for the first time in 40 years. It was a perfectly fine analysis, it was just totally detached from the reality of American politics almost immediately.

Those political scientists did good work at a very inopportune time, but the boldness of the title makes the book a bit of punch line. Nothing they said was particularly untrue at the time they wrote it, but by the time most people read it, it was a relic of a bygone era.

Neil Paine of FiveThirtyEight may have stumbled into a similar situation in 2014 when he wrote the following:

Sabermetrics has come a long way since the first analysts began tinkering with mathematical models, and there are certainly places where statistical thinking has made its way onto the field (for example, the explosion of defensive shifts in today’s game is rooted in probability theory regarding where a batter is most likely to hit the ball). But when it comes to the two-hole, baseball’s decision-makers still have a bit of a climb ahead of them.

That paragraph came at the end of a piece in which Paine shows that number-two hitters have typically been the fifth- or sixth-best hitters in their lineups on average in recent years. The post contains a couple of graphs that make his point quite well, and I have absolutely no qualms with Paine’s argument or method. The post went live on May 2, 2014 and seemed to be an entirely accurate reflection of reality.

Until it wasn’t.

In 2015, number-two hitters produced their highest wRC+ in a generation. Second-place hitters moved decisively ahead of leadoff men and sixth hitters, passed fifth-place hitters, and trailed only number-three hitters and cleanup hitters in production in 2015.

second hitters

Now, this could still be a one-year trend. We can’t say for certain that 2015 demonstrates a true culture change in the way teams assemble their lineups, but the evidence does lean in that direction. Our own August Fagerstrom had an inkling this was coming last Opening Day, and the full season supported his observation. We’ll have to wait another year or two for confirmation.

Weighted Runs Created Plus, Second Spot
Team 2014 wRC+ 2015 wRC+ Difference
Padres 72 125 53
Royals 74 124 50
Rangers 71 114 43
Blue Jays 119 159 40
Cubs 98 130 32
Mariners 74 105 31
Indians 82 112 30
Reds 104 129 25
Cardinals 94 109 15
Red Sox 84 95 11
Marlins 86 96 10
Yankees 87 97 10
Phillies 98 107 9
Diamondbacks 97 104 7
Braves 79 82 3
Tigers 97 97 0
Rays 112 111 -1
Athletics 81 79 -2
Mets 107 105 -2
Giants 121 116 -5
White Sox 76 67 -9
Brewers 100 90 -10
Dodgers 142 129 -13
Pirates 117 103 -14
Twins 113 97 -16
Astros 129 112 -17
Nationals 116 91 -25
Rockies 110 83 -27
Angels 165 135 -30
Orioles 139 93 -40
League 102 107 5

The big driver of the overall trend seems to be that teams are learning not to put terrible hitters in the two hole. In 2014, six teams got less than 80 wRC+ from the second spot in their lineup, but only two were that bad in 2015. There were 11 teams with less than 90 wRC+ in 2014, but just four such teams produced that little from the two slot in 2015.

Looking at the top seven teams, the Padres had the biggest jump, giving plate appearances to Yangervis Solarte, Derek Norris, and Cory Spangenberg instead of Everth Cabrera, Chris Denorfia, Seth Smith, and Will Venable. Their tale is one of slightly better players performing better in a particular spot rather that a baseball revolution.

The Royals are a different story. Omar Infante made 97 starts in the second spot in 2014 and the team gave 32 to Eric Hosmer, 15 to Nori Aoki, and 13 to Alcides Escobar. In 2015, Mike Moustakas started 93 games in the two spot and Ben Zobrist started 47. That appears to be a pretty clear philosophical adjustment.

The Rangers followed suit and went with Shin-Soo Choo (86 starts), Elvis Andrus (34), and Rougned Odor (21) in 2015 after Andrus got 151 starts all to himself in 2014.

The Blue Jays relied on Melky Cabrera for 112 starts in 2014 along with 21 from Jose Bautista. That worked out really well for them, but they managed to snag Josh Donaldson from the Athletics over the winter and put him in the second spot for 136 games in 2015 to make their club even better.

The Cubs shared the wealth in 2015, going with Kyle Schwarber (51), Anthony Rizzo (43), Kris Bryant (28), and Jorge Soler (16), but it was a major upgrade over guys like Javier Baez (52), Junior Lake (30), and Justin Ruggiano (28) who got the call in 2014. It’s hard to say how much of this change was a shift in thinking and how much was simply based on having much better options.

The Mariners tried Dustin Ackley (48), James Jones (37), and Brad Miller (21) in 2014, but leaned hard on the vastly superior Kyle Seager (75) in 2015 along with Austin Jackson (24), and Seth Smith (22). This too seems like a change of pace.

Finally in Cleveland, Terry Francona went with a strong mix of Francisco Lindor (85) and Carlos Santana (30) instead of Asdrubel Cabrera (51), Jose Ramirez (42), and Nick Swisher (39). It’s hard to say if this is a new direction given that Lindor dramatically outperformed the industry’s first year offensive expectations. It’s entirely possible that the Indians thought of Lindor as a Cabrera-type hitter for 2015 and lucked into the top of this table. If you’re interested in the other 23 clubs, Baseball Reference has a batting order table for each team.

I’ve had discussions with plenty of people who argue against deploying your best hitter in the second spot because they believe the third and fourth spots are better for maximizing total lineup production. That’s a reasonable hypothesis as far as I’m concerned and am open to more sophisticated lineup analysis going forward. Yet, despite that potential uncertainty, there does appear to be a clear shift away from low-production, contact guys who can bunt in the two hole. Teams aren’t necessarily embracing the specific advice from The Book as much as they are learning its broader lesson: good hitters at the top, bad hitters at the bottom.

You shouldn’t make too much out of a single data point, but it does appear as if 2015 will be remembered, in part, as the year major league managers let the idea of the old fashioned number-two hitter wither on the vine.



Print This Post



Neil Weinberg is the Site Educator at FanGraphs. He is also the Managing Editor at Beyond The Box Score and can be found writing enthusiastically about the Detroit Tigers at New English D. Follow and interact with him on Twitter @NeilWeinberg44.


Sort by:   newest | oldest | most voted
jdbolick
Member
Member
3 months 18 days ago

The value of on-base percentage is greatly influenced by the quality of the following hitters while the value of slugging percentage is greatly influenced by the on-base percentage of the preceding hitters. So it makes sense that you want your highest OBP, relatively low SLG hitter at the top of the lineup followed by your next highest OBP, mediocre SLG hitter, and then your very best hitter in the third spot. I understand why teams don’t want someone slow-footed in the two spot clogging up the basepaths and risking double plays, yet they were previously placing far too much value on that loss compared to the gain of simply being on-base much more frequently.

Richie
Member
Richie
3 months 18 days ago

NOBODY! clogs up the basepaths. Probably hasn’t happened once in the history of professional baseball that Fast Player has passed Slow Player on the basepaths other than one or the other was running to the wrong base. Which itself I don’t recall ever actually seeing happen.

jdbolick
Member
Member
3 months 18 days ago

There are plenty of times where a slow runner prevents a faster trailing runner from taking an extra base.

williams .482
Member
Member
williams .482
3 months 18 days ago

Sure, but in order to actually “clog up the bases” the fast runner would have to get a full base *ahead* of where the slow runner would have been. That practically never happens.

jdbolick
Member
Member
3 months 18 days ago

Huh? No, it doesn’t. “Clogging up the basepaths” would simply mean preventing a runner from achieving a base that he otherwise could have.

Bip
Member
Member
Bip
3 months 17 days ago

Yeah, you guys realize a faster runner will intentionally hold up if he sees the runner ahead of him cannot advance and free up the next base, right? He’s not going to just keep blindly running.

bostnboy3
Member
bostnboy3
3 months 17 days ago

This happens pretty regularly.

Slow runner on 3rd, fast runner on 2rd, Somewhat shallow fly ball, runner on 3rd has to hold because he isn’t fast enough to tag up. But a runner with average speed would have scored, allowing the fast runner to take 3rd on the throw home.

So instead of a run home and a runner on 3rd. It’s still runners on 2nd and 3rd because the slow runner didn’t allow anybody to advance a base.

Anon
Member
Anon
3 months 18 days ago

My first thought was that perhaps Mike Trout was single-handedly driving at least some of that (you know, best or nearly best hitter in baseball and all that), however he was only in the #2 spot for about half the season last year after being the #2 hitter for all of 2014 and half of 2013. It does explain the Angels’ huge drop-off from 2014 to 2015 – Calhoun got most of the rest of the #2 PA last year and put up a 760 OPS. However the overall trend does track at least somewhat with Trout spending a lot of time in the 2 hole – Trout by year (all # are PA):
2012: 639 at #1
2013: 88 at 1, 405 at 2, 223 at 3
2014: 696 at 2, 9 at 3
2015: 346 at 2, 336 at 3
(Interesting note – he has not batted lower than 3rd since his cup of coffee in 2011)

Bip
Member
Member
Bip
3 months 18 days ago

I’m not sure that I find there is a convincing signal in there given this data. I am perhaps more inclined to believe in it just because I know in general that teams, including on-field management, are moving towards conforming with sabermetric principles. However, given that 1st spot is still the 2nd-weakest of the top 6, and that the 3rd is still clearly the strongest, despite the Book saying it is the 5th most important, tempers my belief that any of this is happening by very very slowly.

Bip
Member
Member
Bip
3 months 18 days ago

is happening *but* very very slowly.

scooter262
Member
Member
scooter262
3 months 18 days ago

I don’t want a complete rehashing of the Book (I’ve not read it), but can you give a bit of insight into why the #3 spot in the lineup is only the fifth most important one? From most of the analysis that I’ve read and heard, I was under the impression that the 3rd spot is where you *should* put your best overall hitter.

grandbranyan
Member
grandbranyan
3 months 18 days ago

The #3 spot in the order comes up significantly more often than any other spot in the order with bases empty & two outs, which is of course the least desirable base/out state from a run expectancy point of view.

bostnboy3
Member
bostnboy3
3 months 17 days ago

This seems counter-intuitive that the guy hitting after the #1 and #2 hitters would be up more often with 2 out and the bases empty. Could this be *because* team’s have historically been bad at batting order optimization?

If you put your 2 best hitters #1 and #2, wouldn’t the #3 hitter bat with 2 out and nobody on the least? Since nobody else has 2 better hitters directly in front of him?

Jorge Fabregas
Member
Member
Jorge Fabregas
3 months 18 days ago

The short (and probably over-simplified) reason is that the third hitter often comes up with two outs and no one on, which is not a particularly important situation.

Jorge Fabregas
Member
Member
Bip
Member
Member
Bip
3 months 18 days ago

Yeah the idea of the #3 hitter as the most important definitely falls in the “conventional wisdom category.” As linked below, here is the actual order of lineup importance, based on the study:

1, 4, 2, 5, 3, 6, 7, 8, 9

I think what is also very clear is that that conventional notions of the ideal lineup are clearly designed as the “best case scenario.” In other words, people are designing the lineup for what they think should happen, without considering the consequences when that doesn’t happen. The idea goes something like this:

1: Speedy leadoff guy gets on and maybe steals. Doesn’t consider that a lot of speedy players have OBP around .300, meaning this doesn’t happen 70% of the time
2: Good contact hitter comes up, gives the #1 guy a chance to steal, or executes a hit-and-run, advances the runner in some way, without necessarily getting on himself. Relies on poor assumption at #1 and also overlooks that it is better to just have a guy who gets on, rather than just advancing the runner.
3: Best hitter comes up, to drive in the leadoff guy. Fails to consider that the most common individual outcome is that 1&2 both fail to reach, and that while having him up with runners on is good, you are minimizing the impact of his OBP, which the best overall hitter also generally does well. The 4th spot, conversely, either comes up with guys on, or he leads off the next inning, both very valuable.

joecb91
Member
joecb91
3 months 18 days ago

Don’t forget “#2 hitter intentionally tries to hit the ball to the right side to advance the leadoff hitter instead of trying to get a hit”

domxbomb
Member
domxbomb
3 months 17 days ago

what? all of this is trying to predict sequencing which we simply can never do with any confidence. in fact the only time you can ever correctly predict lineup sequencing is your 1-2-3 hitters in the first inning.

I thought basic sabermetric principle held that lineup order doesn’t matter much, other than getting your best hitters the most at bats (top of the lineup). The idea that the 6 spot is more important than the 3 spot seems ludicrous to me. In that study the 6 spot saw more important base/out situations because the best hitters (with the best OBP) bat 3/4. If you put the best hitters 6th, all of a sudden the 7 hitter would become the most “important”.

Antonio Bananas
Member
Member
Antonio Bananas
3 months 17 days ago

domxbomb

It doesn’t say that the 6th is more important than the 3.

“#1, #4, #2, #5, #3, #6, #7, #8, #9” is the order it says they are important. With 1 needing to be the highest OBP. I agree with your general logic tough, that the rules above are based off the situations caused by traditional sequencing.

bostnboy3
Member
bostnboy3
3 months 17 days ago

Antonio Bananas

>the rules above are based off the situations caused by traditional sequencing

If this is true, which logic tells me it is, then the rules are absolutely meaningless because you can’t actually use them to optimize your roster. They just tell you where the most important positions in the batting order were based on how all teams currently configure their lineup.

If you change your lineup, it changes all the calculations about where the most important spots in the order are. The most important spot will always be the batter right *after* your two best hitters.

AJS
Member
AJS
3 months 17 days ago

I don’t think having 1 and 2 fail to reach is the most common individual outcome.

Even with 2 guys with a .300 OBP ahead of the #3 hitter, you would expect at least 1 to reach more than >50% of the time (admittedly this is if you assume the chances of them reaching are independent, which they may well not be). If the OBPs are higher than .300, the chance only increases.

Bip
Member
Member
Bip
3 months 17 days ago

(#1 reaches and #2 doesn’t), and (#2 reaches and #1 doesn’t) are two different individual outcomes.

Brian Reinhart
Member
Member
3 months 18 days ago

aaaaaaaaand my team just hired Dusty Baker.

Richie
Member
Richie
3 months 18 days ago

:-) Actually, I half-recall that by his final year in Cincy Dusty had finally gotten over that ‘bat my very worst hitter 2nd’ thing?? Or am I misremembering?

Jason B
Member
Jason B
3 months 18 days ago

Somewhere, somewhere, Pokey Reese and Royce Clayton are sitting by the phone, just waiting for ol’ Dusty to call…

MonkeyMan
Member
Member
MonkeyMan
3 months 18 days ago

Don’t forget Neifi.

JDX
Member
JDX
3 months 18 days ago

He was still batting Zack Cozart #2 through most of the first part of the season, with a smattering of Chris Heisey, Derrick Robinson, and Xavier Paul.

Starting in August of 2013, we saw Todd Frazier and Brandon Phillips manning the #2 hole, with Ryan Ludwick (!) hitting 2nd the least 3 games of the year.

Joey Votto batted 2nd zero times.

Recall this was also a team with Shin-Soo Choo and his .423 OBP. Baker had the opportunity to bat Choo and Votto 1-2 and chose to do it zero times because, of course, CANT BAT TWO LEFTIES BACK-TO-BACK! They might bring in a lefty in the 7th inning! Oh no!

JDX
Member
JDX
3 months 18 days ago

Forgot to add this.. Votto is a career 146 wRC+ hitter against lefties. Unfortunately for us Reds fans back then… and Nats fans now… I don’t think Dusty has ever heard of wRC+.

MonkeyMan
Member
Member
MonkeyMan
3 months 18 days ago

And his introductory press conference hardly inspired confidence that his thinking evolves with the times…

Immanuel Kamment
Member
Immanuel Kamment
3 months 18 days ago

The competition between Neils has never been fiercer!

TKDC
Member
Member
TKDC
3 months 18 days ago

I think a large factor is just the run environment now. For the most part the same types of players are hitting in the 2-hole. They are not power hitters, and if they are, they are in their first 1-2 years. There are just not that many established good power hitters, and often times nowadays you can be a light hitting on-base type guy and be one of the better offensive players on your team.

I’m not saying there is zero shift in thinking, but when you see the same team go from one of the top hitters in the 2-hole one day to one of the worst the next day, it’s hard to put much stock in this. Very few teams go the whole year without giving some significant amount of 2-hole starts to one of the worst hitters in the lineup. This doesn’t happen with the 3-4 spots.

Fernando
Member
Fernando
3 months 18 days ago

One thing that’s always confused me about lineup optimization is that most of the numbers seem to be based on teams using inefficient orders. For example, the 3 hitter is statistically less likely to come up to bat with runners on than the 4 or 5 spot, so the sabermetric argument is that the 3 hole is less important. However, isn’t that data derived from years of the 1 & 2 hitters being misused? So if those top spots in the lineup were fixed to optimize efficiency then wouldn’t the 3 spot become more valuable? Or am I misunderstanding the data?

Richie
Member
Richie
3 months 18 days ago

Been awhile since I read ‘The Book’, but was that actually data-based in the first place? Or just theoretical? (good theoretical, mind you)

Fernando
Member
Fernando
3 months 18 days ago

I’ve never read it myself, just articles that discuss the conclusions. So if you’re right in that the conclusion is theoretical then I would assume that takes into account the very issue I’m curious about. I just assumed it was numbers driven.

Sinnycal
Member
Sinnycal
3 months 17 days ago

I’ve wondered about that myself. Three hole hitters statistically bat with no men on and two outs more than any other lineup spot, but like you said, that study is based on decades of three hole hitters slotting in after a speedy leadoff man and a lousy “bat control” hitter. I have to imagine the situation would improve with better lineup optimization at the top of the order, but it would probably remain true just because the three hole hitter still comes up third in an inning most often out of any lineup spot. Without looking into any data, I would presume the most common outcome for any given inning is three up, three down, simply because even the best hitters fail to reach base more often than they succeed. The player with many opportunities to bat third in an inning will face the no men on, two outs gamestate most often pretty much by default.

AJS
Member
AJS
3 months 17 days ago

While 3-up, 3-down might be the single most common outcome of an inning (I’m not actually sure if this is true), it’s definitely not the majority of innings. Most innings involve at least 1 batter reaching.

Binomial distribution says that if you have two hitters each wiith a .350 OBP (and therefore a 35% chance of reaching base) that the odds at least one of them reaches base is 58%. Compare that with a 51% chance if both batters have a .300 OBP.

This obviously assumes each at bat is independent, which it isn’t (i.e. the 2nd batter is more likely to reach if the 1st batter has also reached). But nonetheless, we would actually expect the 3rd hitter to come up with a guy on base more often that he comes up with two out nobody on, and we’d have an even greater expectation of him coming up with someone on if teams do a better job of optimizing their lineup.

All of which to say, I’m not sure it’s still true that the three spot is less valuable than other spots in the order.

bostnboy3
Member
bostnboy3
3 months 17 days ago

This is correct. The #5 and #6 spots in the order are only considered more valuable spots because they are directly after where teams traditionally put their best hitters – #3 and #4.

Basically the most valuable spot in the batting order will always be the spots directly after the best hitters.

xeifrank
Member
3 months 18 days ago

Be careful of any study or any “book” that gives a template that states where each “kind” of hitter should be hitting in the batting order. There is no one size fits all template. Each team is different and has different types of hitters and should have their own unique template per the 9 players (and opposing pitcher) in their starting lineup.

Only glove, no love
Member
Only glove, no love
3 months 18 days ago

I think I understand your argument and you have done some cool work with lineup optimization and the value or lack thereof…

But I can’t imagine a specific scenario where a team would want to sig deviate from the 1, 4, 2 ,5, 3 etc paradigm…

Do you have one? Thanks.

xeifrank
Member
3 months 18 days ago

OGNLS,
Thanks for the reply. The 1,4,2,5,3,etc paradigm probably gets you into the top 25th percentile more than half the time. I think we should strive higher.

How about this. You give me the 9 batters and tell me which projections to use and what the most efficient lineup (vs LHP, RHP or both) using the paradigm listed above or anywhere else and I will see how many wins/162 I can beat it by.

vr, Xei

Only glove, no love
Member
Only glove, no love
3 months 17 days ago

Hmmm… well I agree that the 1,4,2,5,3 etc paradigm is not always or even most often the best.

I was just trying to think of a type of hitter or a group of hitters that would require “blowing up” the paradigm…

Not saying they don’t exist, just that I am not sophisticated enough to come up with them.

xeifrank
Member
3 months 17 days ago

Those are great thoughts by the way. I am not sure any human being is smart enough or able to compute fast enough to come up with a top 1% lineup consistently. Would be a great competition to have between people. Here are 10 starting lineups, put them each in most efficient order. See who does best.

Only glove, no love
Member
Only glove, no love
3 months 18 days ago

As a cards fan it drives me crazy that there is such resistance to batting Carpenter first. In spite of the fact that he is the perfect case study as to why your best hitter should hit there!

But it goes to show the power half sophisticated concepts can hold over the human mind.

I always reckoned it was the combination of easy sophistication and lack of obvious and instant negative feedback that did it…

Paul22
Member
Paul22
3 months 18 days ago

Maybe Carpenter does not feel comfortable hitting first. One must consider the human mind of those who are expected to execute these grand strategies

Only glove, no love
Member
Only glove, no love
3 months 17 days ago

Carp is, at the very least, fine with batting leadoff. And he has gotten very good results doing it. And much better results than batting elsewhere…

Paul22
Member
Paul22
3 months 18 days ago

I have to question the Yankees wRC+ of 87 in 2014. Jeter batted 2nd 85 percent of the teams #2 spot PA, and had a wRC+ of 74. Half of his replacements did worse in terms of OPS at the 2nd spot. Can’t find team splits to check wRC+ on this site though

Paul22
Member
Paul22
3 months 18 days ago

Ok, found the splits here. Jeter had a 76 Rc+ and was the primary number 2 hitter. The boost was due to a handful of productive AB on days he rested by his replacement and some pinch hitters that produced 14 RC in only 52 PA compared to Jeters 53 in 619.

I think its probably best to see who the managers pencil in as the every day number 2 hitter to see if they have learned this lesson. Giradie obviously has not since he had Brett Gardner who had a 66 wRC+ in the 2nd half get most of the number 2 spot PA in a pennant race

LHPSU
Member
LHPSU
3 months 17 days ago

Funny, before that we were complaining that Gardner wasn’t hitting at the top of the order.

Tangotiger
Editor
Member
3 months 18 days ago

1. If you haven’t read The Book, you can read a good portion of it for free, using Amazon’s Look Inside feature. Just do a search for “lineup” and that’ll get you to where you need to be.

2. The Book is giving us guidelines, not hard and fast rules. Clearly there is dependence among all the hitters, so it’s a very nuanced process to try to get an optimal order.

3. As it relates to the #2 hitter, The Book is suggesting that it should be one of the three best hitters. Sometimes it might be your best, in odd cases it might be your 4th or 5th best.

4. It becomes more complicated when you think of the handedness issue, and that the hitting talent might be so tight on a particular team. Someone who is .360 wOBA against LHP but .340 against RHP compared to someone who is .350 against LHP and .370 against RHP. What might be a close call against LHP can be a big difference against RHP.

5. And I can say two dozen more things about it.

Jon C
Member
Jon C
3 months 18 days ago

The Book. What an egotistical name. Anyway I read it years ago, noticed the flaws in methodology, and moved on. There are more factors to consider than obp and slg. It’s not necessarily a bad thing to have a better hitter lower in the order. I could explain why but not now as I may publish it.

Tangotiger
Editor
Member
3 months 17 days ago

I noticed the flaws in your post and moved on. I could explain why, but not now as I may publish it.

Jon C
Member
Jon C
3 months 17 days ago

Tom you turn me on with your snarky copycat attempts at humour.

Jon C
Member
Jon C
3 months 17 days ago

You didn’t have an original thought when you co wrote your book either.

fredsbank
Member
fredsbank
3 months 16 days ago

*Tango drops the mic*

Serbian to Vietnamese to French and back
Member
Serbian to Vietnamese to French and back
3 months 17 days ago

Book. Name selfish. After reading a long time, found a fault in a review of methods and move on. There are many factors to consider that some access and GIS. which is not necessarily a bad thing to be a better fighter, low in the order. It may explain why, but now that you can publish it.

MGL
Member
3 months 18 days ago

Neil you should have looked at pre-season projections for those number 2 hitters rather than what they actually did. And maybe mid-season projections too. At the very least, look at both. Too much noise in how they actually do. We are interested in whether managers put good or bad TRUE TALENT hitters in the 2 hole, not whether they may have put a poor hitter in the 2 hole who had a really lucky season or a good one who had a really unlucky season. Obviously for 30 teams combined and especially in several seasons combined those things will equal out and actual performance will very likely mimic projected performance, but in one season, even for 30 teams, we could see some substantial differences.

Maybe someone can compare pre-season projections in 2014 with those in 2015.

Jon C
Member
Jon C
3 months 17 days ago

You’re assuming the managers weigh the batters the same as a fangraphs or pecota projection. Some teams don’t have analytics departments and those that do probably use their own projections.

Noah Baron
Member
Noah Baron
3 months 17 days ago

Teams aren’t looking at Steamer and ZiPS projections…

MGL
Member
3 months 17 days ago

That is not my point. Has nothing to do with team decisions. Let me give you an (extreme) example of what I mean.

Let’s say that teams in 2015 were just like the Royals and had players like Infante batting second, so that nothing really changed.

And let’s say that half of those players just had a lucky season and beat their pre-season projections by a lot.

Should we conclude that teams were changing their approach? No!

The only way to test whether team approaches are any different (at least for one season) is to factor our the luck by looking at either pre-season or post-season projections for those #2 hitters, not their actual performance.

Post-season projections are actually better for this because they reflect player true talent during that season a little better than pre-season projections.

The technically correct and rigorous way to do this analysis is to take every game that a batter batted second and use his to-date projection on that day.

Again, for multiple seasons it is not worth the trouble as the luck will even out for those #2 batters as far as their observed performance versus their true talent (it will be about the same).

jim fetterolf
Member
jim fetterolf
3 months 17 days ago

Depth may have something to do with, Moustakas was bad in ’14 and Royals hadn’t traded yet for Zobrist, so probably less philosophy than dumb fool luck and having players improve.

Antonio Bananas
Member
Member
Antonio Bananas
3 months 17 days ago

On the thoughts of sequencing, has anyone done a study on pitching? Surely the current model isn’t the best. Lineup seems pretty random but what we know about the number of times a pitcher goes through the lineup has always intrigued me.

Here’s what I think would work better from a pitching perspective.

2 relief aces alternate starts. 3rd best reliever starts on the 5th day. The normal “closer” starts and only pitches one inning unless he got through the inning in something like 10 pitches.

My logic is that, as starters go through the lineup multiple times, they are more likely to get hammered. You would assume that the further down the lineup, the less likely that damage is. So the 2nd inning he’s facing somewhere between the 4th and 7th hitters. His 3rd time through the order he’s not seeing the best hitters first and can cruise (hypothetically) through the 6-9 hitters before seeing the 1-5 hitters for the 3rd time. From there you just manage lefty/righty.

Each “closer” gets about 65 starts/innings a year, which is about their current workload. Each “starter” comes in for the 2nd inning until he’s hit the 3rd time through the order, then you use your remaining 4 relievers to get through the game.

Jon C
Member
Jon C
3 months 17 days ago

I’ve thought about this as well but you likely get hammered in higher leverage situations.

bodly
Member
bodly
3 months 17 days ago

“If you’ve recently spent time with other humans, it’s likely that you noticed that they tend to be overconfident about how well they understand the world around them.”

This is possibly the best opening line of an article that I have ever read.

Jon C
Member
Jon C
3 months 17 days ago

Appears the author is guilty of this very sin.

Jon C
Member
Jon C
3 months 17 days ago

I’m not sure if using a simple linear of wrc+ from year two and year one is the right methodology. I’m also not sure assuming The Book is correct methodology.

It would probably make more sense to isolate the teams with low wrc hitters from 2014 and see what happened the next year. The noise from a team that went from 110 or so to 102 probably doesn’t reflect a change in their ideology.

Pinstripe Wizard
Member
Pinstripe Wizard
3 months 17 days ago

So having a defined three year trend that the wRC+ of the 2nd hitters in baseball in increasing doesn’t show you a general change in ideology?

Jon C
Member
Jon C
3 months 17 days ago

No. Why would data from teams already using a non-terrible hitter in the two slot help? You should only look at the teams that had terrible hitting in the two slot.

TrevorCap
Member
TrevorCap
3 months 17 days ago

No, because then regression to the mean would cloud the results. So when you include teams that had good numbers to begin with, you balance the numbers between teams that regresses in a positive way and teams that regressed negatively. That’s not even mentioning the benefits from increased sample size.

Jon C
Member
Jon C
3 months 17 days ago

You can’t have increased benefits of a sample size when the the increase is due to adding invalid data. The thesis is teams aren’t using bad hitters in the two slot. You would therefore remove those that already use decent hitters in the two slot.

Really using wrc at all is sloppy because it doesn’t speak to the talent available to choose from. Also a team going from high 80s to low 90s isn’t proof positive of a change in ideology.

Jon C
Member
Jon C
3 months 17 days ago

Granted, the author does hit on the fact less less teams are under 90 wrc, but the improvement seems to be consistent with the overall league improvement, which could be explained away with the scoring environment is rewarding hitters that put the ball in play.

wpDiscuz