A Little More On The Phillies Batting Order
Earlier in the week, we focused on the top two spots in the Phillies order. Specifically, the revelation that using Jimmy Rollins in the leadoff spot over Placido Polanco would only cost the team about 1.18 baserunners over the course of a season. The commenters pointed out, rightly, that this analysis ignored the slugging aspect of the two batters’ repertoires. Questions were also raised about the lefty-lefty tandem at the heart of the order. So, let’s dive right in and try to get this right by the numbers: A little moron, the Phillies’ batting order.
Remembering back to the LOOGY wonder that was Javier Lopez in the playoffs last year, you might over-rate the “problem” of batting lefties Chase Utley and Ryan Howard back-to-back. Lopez only pitched 4 1/3 innings, but his four strikeouts were memorable, and it seemed that he shut the duo down. But this wasn’t really a year-long problem. Yes, Howard has a nasty platoon split (.424 wOBA vs righties, .324 vs lefties), but Utley doesn’t. His work versus righties (.382 wOBA) is basically indistinguishable from his work against southpaws (.390 wOBA). Opposing managers can bring out their LOOGY for both guys at their own peril.
So, the heart of the order is set. We know that some combination of Rollins, Polanco and Shane Victorino will end up manning the top two spots. The lefty Raul Ibanez has just enough power to be a good fifth hitter – or sixth hitter if you don’t want to put three lefties in a row. Carlos Ruiz has a great OBP, but doesn’t have great footspeed and may have gotten an OBP boost from batting in front of the pitcher all year. Perhaps he can cede that spot to rookie Dominic Brown most days, so that he can work on improving in a low-stress spot.
That seems to handle most of the spots – are we over-thinking this? Especially when we found the difference between Rollins and Polanco to be worth so few baserunners, perhaps this is just one of those things that will work itself out. There isn’t any baseball happening right now, though, so let’s persevere with a remaining tool in our bag.
Baseball Musings has a lineup optimizer tool that we will employ here to ‘check our work.’ Here is the best lineup based on inputting Bill James’ projections into that tool:
Chase Utley
Ryan Howard
Carlos Ruiz
Domonic Brown
Shane Victorino
Jimmy Rollins
Raul Ibanez
Cliff Lee
Placido Polanco
Well. So much for our own musings. In fact, of the top thirty lineups by this methodology, only Utley and Ruiz should be leading off – and Ruiz only appears there in the 14th-best lineup. Still, this is the lineup the numbers like.
Just because this lineup won’t ever appear on a lineup card doesn’t mean we can’t take some things away from the optimizer. For one, it agrees with us that Rollins is a good middle-of-the-order guy. Two, perhaps Placido Polanco’s lack of pop should relegate him to the bottom of the order. Three, Bill James’ projections for Domonic Brown are crazy nice (.288/.346/.505 with 26 home runs and 11 stolen bases).
Four, the difference in (likely) lineup construction is not huge. The best hypothetical lineup here averaged 5.213 runs per game and the thirtieth-best lineup 5.205 runs per game. The worst lineup put forth 4.686 runs per game – but that was with Cliff Lee batting leadoff. Our more likely lineup – featuring Rollins at leadoff and Victorino second – was worth 4.942 runs per game. Lead off with Victorino, put Polanco second and Rollins sixth, and you get 4.961 runs per game. So the difference between the two “most likely” lineups is only .019 runs per game, or 3.078 runs per season.
Given Rollins’ demeanor and personality, he may lead off again this year, with Victorino bridging the gap to the two lefty sluggers. But perhaps it would be worth selling him on being a ‘run-producer’ and ‘protecting’ Howard? It might be the best use of his skills. Once again, though, we should be careful not to over-rate the importance of batting order, as evidenced by the small differences made by the most likely different batting orders. If Rollins is the best Rollins, he can be at the top of the order, that might just undo the negative effects of having him there.
With a phone full of pictures of pitchers' fingers, strange beers, and his two toddler sons, Eno Sarris can be found at the ballpark or a brewery most days. Read him here, writing about the A's or Giants at The Athletic, or about beer at October. Follow him on Twitter @enosarris if you can handle the sandwiches and inanity.
Ibanez is a lefty, which would make your lineup have 4 lefties in a row if Utley/Howard/Ibanez/Brown are 3/4/5/6.
Pretty sure Raul Ibanez is a LH hitter.
He’s a team player. He said he’ll bat from the right side if it helps the team.
This is funny.
“So the difference between the two “most likely” lineups is only .019 runs per game, or 3.078 runs per season.”
but the difference between the most likely lineups and the best lineup is ~0.25 runs per game or ~40 runs per season. that’s huge!
That’s true, but there’s just no way that batting order will happen.
yea i agree it won’t happen. but i think it’s ridiculous how managers refuse to use a more optimal lineup, especially when the difference is on the order of 4 wins over the course of the season.
also i think you’re unerestimated just how bad having utley/howard hit back to back is. sure, utley’s true talent platoon split might actually be neglibible, but you’re ignoring the pitcher’s platoon split there. even if utley isn’t any different against a lefty than a righty, lopez (or any other loogy) is much better against lefties than righties, and having, say rollins or ruiz or another righty between utley and howard woudl either force them to use the loogy against a righty, or force them to only use him for a single at bat against howard, both of which are a much better outcome for the phillies than having him face utley and howard back to back.
Jono, maybe I’m confused, but isn’t a pitcher’s platoon split a direct consequence of the hitters platoon split? How can you view those as separate stats to be investigated against each other?
Big Jgke: hitters and pitchers both have their own true platoon split. the estimate for the outcome of a pitcher-batter matchup is a function of four things: the pitcher’s overall skill, the batter’s overall skill, the pitcher’s platoon split, and the batter’s platoon split.
so say you have a .330 wOBA true talent lefty hitter with a typical lefty platoon split, so maybe .345 vs. righties, .315 vs. lefties.
let’s say we also have Pitcher A who is a true talent .330 wOBA against with no platoon split, and Pitcher B who is also a true talent .330 wOBA against but with a split of .345 vs. righties, .315 vs. lefties.
so the batter against pitcher A would be a .345 if Pitcher A is a righty, or .315 if Pitcher B is a lefty (since the pitcher’s .330 number is around league average). but the batter against pitcher B would be a .330 hitter if pitcher B were a righty, or a .300 hitter if pitcher B were a lefty.
(i don’t know how the exact numbers work out but it’s something like that)
does that make any sense? i’m not sure if i explained it very well.
If it helps, think of it like this. A pitcher with a sick slider is going to be better against same handed hitters than a pitcher with a sick change up. Even if they had the same true talent level the slider guy is going to be better against same handed hitters. Just in case you needed a foundation, I know I have trouble accepting things without a logical foundation for why they happen.
Thanks, to both of you. I was having a conceptual issue with that, and I feel like I have a much better handle on what you’re saying now.
Good thing Charlie Manuel is a fan of advanced statistics and isn’t afraid of a little outside the box thinking or Phillies fans might actually have to worry about that.
No, but seriously, it’s not .019, it’s .19. Which amounts to 30 runs/seasons. BFD.
4.961 – 4.942 is .019, that much I’m sure of.
Yep, yep, after 6 years of math and constant yelling at my teachers that I can just use a calculator to do the work I showed them. 4.961 – 4.942 = .019
The right handed Raul Ibanez was murdered and replaced by his evil left-handed twin, which is funny because you’d think playing baseball someone would’ve noticed the different handedness.
I watch baseball. I also make mistakes.
I don’t know why you’re calling the Phillies batting order, ‘a little moron.’
Damnit!
Was calling myself a moron, which was prescient when it came to the Ibanez gaffe.
Here’s something I thought I’d never want to see, or say, in the Fangraphs comment section… FIRST
“A little moron, the Phillies’ batting order.”
Ha ha. You said “moron”.
Sorry, I had to.
I wouldn’t put too much stock in the “optimal lineup”. First off, the Bill James’ projection for Brown is very optimistic. Usually, the fans projection (which tends to be high) is close to James’. In this case, James has Brown at .288/.346/.505; the fans have Brown at a much more realistic .271/.327/.451. Secondly, the Baseball Musing lineup optimizer does not take adequate account of baserunning ability. For this reason, Ryan Howard batting second is not in fact an optimal use of his skills compared with the other options that the Phillies have.
yeah, and the tool doesn’t consider platoon splits. I think the most useful thing the tool can do is show us how small the difference between Rollins at the top and Rollins in the middle would be. Negated mostly by his baserunning ability.
“James has Brown at . . .” I feeeeeeeeeeeeeeeel good.
The real problem with these things is that they assume a league average batter is in spots 1-9. This matters because it might want you to put a power hitter 3rd and then assume that a high OBP hitter is batting third when telling you who to pick for spot 4.
These things are just approximate, and the further you get from a manager’s typical batting order, the worse they are.
LaRussa’s a genius.
He does get hitting the pitcher 8th right and, generally, gets pitching changes correct as well. He may walk a marathon to and from the mound every night but he’s usually able to take advantage of platoon splits as well as anyone.
If they break camp as expected with Schumaker, Theriot and Punto in the starting lineup (David Freese not ready and Allen Craig primarily used in a platoon role with Lance Berkman), I think that he’d be justified batting Adam Wainwright 6th every fifth day.
Unfortunately the lineup optimizer doesn’t take into consideration the lineup construction that was put forth in The Book. That lineup wouldn’t have either Rollins or Utley batting first, as Rollins isn’t good enough and Utley has too much value in his ability to hit the ball out of the park.
The Book’s optimum lineup would probably have something like:
Victorino
Utley
Polanco
Howard
Ibanez
Rollins
Brown
Pitcher
Ruiz
something like that. Utley would definitely be 2nd and Howard 4th. Rollins, b/c of his speed, should hit 6th. You could switch Brown and Ruiz or even Ruiz and Polanco if you buy into the notion that Ruiz’s 2010 season was more the rule than the exception. It certainly wouldn’t have Utley 1st and Howard 2nd, however.
This is probably it. The lefty thing is a problem, but I guess you could bat Rollins fifth.
C’mon… pitcher batting 8th?
Everyone here knows that the Phils are NOT going to do anything that funky with the lineup card. There is no need.
(and really, who wants to explain to Chooch that its better that the pitcher is ahead of him on the lineup card, and that this makes sense???)
Using a “non-optimized” lineup would be a problem for the Phils, but NO other team in the league is using an optimized lineup! So there is no real penalty for the Phillies to use a more “traditional” lineup.
Within that traditional framework, I’m not overly concerned about who bats 1-2 in this lineup. Utley & Howard are gonna bat 3-4 (barring injury). I don’t have a problem with that either. Phils have been able to score lots of runs with them batting 3-4.
My concern remains that this lineup, no matter HOW you arrange it, is potentially somewhat vulnerable to LHPs. You can’t “construct” away 4 LHBs.
There is no RH power bat to keep the other team honest. And I’m frankly a little perplexed that the Phils would identify and sign a FOURTH pitching ace… but fail to locate a corner OF with a decent RHB.
It just seems odd to me.
And when Manny signed for a lousy 2 mill. ugh. If this team is going for it… would 2 mill be that much for a killer RH pinch hitter (and DH if they get to the series)?
Or maybe Ben Francisco is better than I think?
There is no cost to them. The cost is 1/4 of a run per game. If the changed, they would get 1/4 more runs per game.
Also, I think Manny wanted to sign for more than 2 million next year based on better performance than he delivered in Chicago at the end of last season ($5 million per RBI), so PHing for a NL team was not an option. Even waiting until the all-star break and then signing with an AL team looking for more power in the DH would have been preferable.
Not to get too nit-picky, but 1/4 run a game is 40 runs a year…
Think about it… if you’re a pitcher, would you ming if your ERA dropped from 4.00 to 3.75??? thats 1/4 run a game.
But there is no loss to the Phils anyway… because NO ONE ELSE is using an “optimized” lineup either!
You are spot on about Manny. I’m sure he doesn’t want to sign on for PH and (severely limited) DH duties. I just like to dream out loud.
@ Dave S — 40 runs a year is 4 wins! That could be the difference between winning the division or winning the wild card. It could be the difference between hosting playoff series or having to travel. It could (gasp!) even be the difference between making the playoffs and not.
And, aside from all the lefty hitters, you should also be concerned about the fact that the Phils are getting older and their hitters — particularly Rollins, Polanco, and Victorino — aren’t getting any better. It’s still a pretty good lineup but nowhere near the juggernaut that many people believe them to be.
@chuckb… YES. I know 40 runs is 4 wins. I GET IT!!! That was what I was saying!!!
But the Phillies will NOT use an unconventional “optimized” lineup.
AND… it will NOT hurt them.
and it wouldn’t matter if it was 400 runs…
Because NO ONE ELSE USES ONE EITHER!!!!
There is NO advantage or disadvantage to a strategy that NO ONE USES.
And yes… I completely agree with your thoughts about the Phils lineup. Its getting old. They had injuries last year. I assume they’ll have some this year. I don’t like 4 LHBs in the lineup everyday. Werth will be missed against LHP. But overall, they’re still good.
And 4 aces covers a lot.
No matter what, the Phils are going to have an “odd” sort of batting order (at least I think so). The 4 LHBs. In the end, I don;t think it will matter much. They hit well enough. Other teams have won World Series with 4 LHBs in the everyday lineup. So… we’ll see how it goes.
Even thouogh it doesn;t appear to matter a whole lot… I remain acutely interested in the lineups Charlie Manuel will be putting out there.
@ Dave S — thanks for all the all-caps words and exclamation points. I wouldn’t have understood your argument without them.
Your argument is that a loss of 40 runs, or 4 wins, is irrelevant b/c every other team will forgo the optimal lineup as well. How about this: the Phillies should just decide not to play Carlos Ruiz at all this year. He was a 4 win player last year and can be replaced by a replacement level player. It’ll cost the Phillies 4 wins but that shouldn’t matter b/c no other team is using their optimum lineup. Make sense?
40 runs is 40 runs and 4 wins is 4 wins. The Phils are costing themselves the opportunity to win 4 more games independent of what every other team in baseball is doing. The point is that it does matter. They have no control over what any other team in baseball is doing. Perhaps in a game theory scenario where teams could conspire together it would make sense to forgo those 4 wins but, in this scenario, every team’s lineup construction is independent of every other team’s lineup construction.
The Phils are losing 4 wins with this lineup. It matters. That the Braves and every other team in baseball is also losing some wins through poor lineup construction doesn’t give those 4 wins back. If it truly doesn’t matter, they should just sit Carlos Ruiz this year.
chuckb… I was not arguing pro or con on the “optimized” lineup.
I was simply saying that the Phils incur no penalty for failing to use one. Because no other team in the NL is using one either.
So, in that respect, if every team in the league had a 4 win catcher, and left him off the team… it would not hurt the Phils to leave their 4 win catcher off the team either.
Is that clear enough… or do I need to go ALLCAPS and exclamation points again? LOL
Dave S: your argument makes no sense!
they’re costing themselves 4 wins in the standings by not optimizing their lineup. how is that not hurting them?
now, if every team then decided to follow the phillies lead and optimize their lineup, maybe those extra 4 wins disappear. but that doesn’t change the fact that right now, they could choose to increase their win expectation this season by 4 wins and they’re probably going to choose not to do it. that’s the definition of hurting themselves.
also, i think that the phillies use one of the most poorly optimized lineups in mlb. the difference for most other teams between their projected lineup and most optimized lineup is probably on the order of 1-2 wins.
@Jono411: what argument?
please re-read, and I quote:
“… I was not arguing pro or con on the “optimized” lineup.
I was simply saying that the Phils incur no penalty for failing to use one. Because no other team in the NL is using one either.”
Is that unclear?
Dave S: very clear, and very wrong. their penalty for not using an optimize lineup is 4 expected wins in the standings. this is true regardless of what other teams do because what the phillies do with their lineup is independent of what other teams choose to do with their lineups.
There is no “actual” 4 win advantage to an optimized lineup.
There is a potential 4 win advantage.
There is also a very real potential that your crazy lineup with the pitcher batting 8th, backfires, causing you to lose a pennant, and 3 million Phillie fans come screaming for your head on a platter…
This fact is exacerbated by the knowledge that the Phils are favored already. They have no reason to assume the risk inherent in using your flaky lineup.
Law of competitive balance would tell you that a team that is chasing the Phils should be introducing this “optimization”. Not the Phillies.
Has any team won a World Series with 4 LHBs in their everyday lineup???
I don’t know… I’m honestly asking.
The 2002 Angels primarily started LHB’s Adam Kennedy, Darin Erstad, Garrett Anderson and Brad Fullmer.
I’d guess a lot have, but the most recent was the 2009 Yankees: Cano; Damon; Cabrera; Matsui.
Cabrera is a switch hitter
I guess I should qualify that…
any NL team, or pre-DH AL team won a World series with 4 LHBs in the everyday lineup?
2001 Diamondbacks: Mark Grace; Tony Womack; Luis Gonzalez; Steve Finley. Four lefties.
And, two top bench players (in terms of PA’s) were Counsell (lefty) and Delucci (lefty).
Yes… I answered that below a few hours before you posted. I should have replied directly to my own question.
But thanks for checking.
The Arizona team is similar to the Phils. Old. Two stud aces at the top of the rotation (Big Unit and Schilling).
I dare say their best bench hitter was Erubiel Durazo, also a lefty.
They did have Matt Williams and Reggie Sanders in the lineup. The Phils have no RH bat like those in their lineup.
The 1979 Pirates also won a series with 4 LHBs in lineup. They got very good years from their top 4 stating pitchers. Their most dangerous bench hitter, Milner, was a LHB also. Pirates had Bill Robinson in their lineup as a RH power threat… and Madlock of course was a very good RHB (without the power).
For all the optimism around the Phils, I’m not excited about the RHBs the Phils will run out this season: Rollins, Ruiz, Victorino, Polanco.
Almost all the power is in the LHBs: Howard, Utley, Ibanez, Brown.
I’m surprised the Phils don’t go after someone like Glaus?
The big limitation of Baseball Musings’ lineup optimizer, unless it’s changed a lot since I last saw it, isn’t that it doesn’t take into account baserunning, or platoon splits, or something odd in The Book, or anything like that. It’s that it calculates the contributions of each hitter independently of the others, based on how hitters in each spot have contributed throughout history. It suggests unorthodox things, like hitting pitchers 8th (sometimes 7th or 6th!) and sluggers 2nd, but the contributions of the other hitters are calculated as if those spots had been filled more traditionally (because that’s how they have been filled throughout baseball history).
It’s an interesting little tool and it gets people thinking, but I don’t think the process is close enough to correct to put much stock in the results… you’d really want to run the lineups through a simulator a bunch of times. That costs time and CPU (surely someone has written a program to do this, it isn’t that hard), but you could easily incorporate more detailed stats than just OBP and SLG and maybe even account for baserunning skill (at least in a blunt and subjective way).
The lineup tool makes the classic mistake of extrapolating outside of available data. The classic lineup has a very good hitter batting third and a poor one batting seventh; it is not hard to believe upgrading the #7 hitter from poor to fair might be worth more runs than upgrading a good #3 hitter to very good. But the lineup tool, using simple linear regression, projects that out to the absurdity of swapping #3 and #7. The tool is worthless for its intended purpose of selecting optimum batting orders.
answering own question: 2001 Diamondbacks won with 4 LHBs in the lineup.
1979 Pirates too.
And they had LHBs Parker-Stargell batting 3-4. But they’d split them with Bill Robinson against LH starting pitchers though.
I’m guessing that there’s a difference between LOOGY’s and left handed starting pitching. A lot of starting pitchers don’t have extreme lefty-righty splits, otherwise they probably wouldn’t be starting pitchers.
So Utley might not have big splits, but that’s because he’s mostly facing lefty starting pitchers. I bet if you look at his splits vs. relief lefties, there’d be a bigger one.
I don’t think that LOOGYs are better against LHB than LH starters are against LHB. Rather, I think LOOGYs are worse against RHB than LH starters are against RHB.
A teams optimal lineup would be quite different against righties versus lefties, when taking most players plsits into consideration. If a team used its optimal lineup against the appropriate handed pitcher what would the increase in their total runs be? (i.e If the phillies didnt consistently hit howard, poor hitting vs L, third in the order how much would it increase runs?)
assuming you mean optimizing it by hand as opposed to using a typical traditional lineup, my guess would be that its on the same order of magnitude as what we’re seeing here for going to a “neutral” pitcher, but probably a bit more. so maybe 50 runs per season for the most extreme case (ie the phillies), and probably more like 10-20 runs for the majority of teams. optimizing lineup construction really has a much bigger impact than most people assume.
also, about your howard in the 3 spot vs. lefties, that’s probably actually the best place to put him vs. lefties since he’s a high power, low obp guy (vs. lefties – he’s great in both vs. righties), which is ideal for the 3 spot (you want your best hitters in the 1, 2, 4 spots, then splitting your next two best hitters between the 3 and 5 spot with the higher power guy in the 3 spot, better on base guy in the 5 spot). related to this, i think the reason the phillies lineup is so far from optimal is because they put their best hitter – utley – in the 3 spot, which is just criminally bad.
Someone earlier said that “no one is using optimized lineups”. If LaRussa batting a pitcher 8th in the NL is using one, I’d say what Maddon does down in Tampa counts too.
larussa is definitely NOT using an optimized lineup. he has pujols hitting 3rd when he almost certainly should be hitting 2nd or 4th. that probably more than offsets whatever marginal gain there is for hitting the pitcher 8th instead of 9th.
as for maddon, i don’t really follow the rays so i have no idea.
LaRussa gave Skip Schumaker over 500 ABs last year. That’s de facto “non-optimazation”! LOL
“non-optimization”. duh
Maddon is in a DH league. Doesn’t count. Getting rid of the pitcher as a batter surely helps optimize any lineup.