Ryan Howard Against LHPs
One interesting match-up in the upcoming World Series is all of the great left handed Phillies hitters against a Yankees’ rotation that features two solid left handed pitchers, CC Sabathia and Andy Pettitte. The Phillies will face a lefty starter in at least half of the World Series games no matter if the Yankees go with a three or four man rotation or how many games the series goes.
The biggest issue for the Phillies is Ryan Howard, who is a very bad hitter against LHPs.
On the splits page, you can see that the biggest difference is in his ISO and K rate. On the other hand, in recent years, the difference in his walk rate against LHP and RHP has decreased. Looking at a pitch-by-pitch basis he takes more called strikes against LHPs than RHPs (13% of pitches versus 11% of pitches). His whiff rate increases from 28% against RHPs to 38% (Mark Reynolds territory) against LHPs. Here is how his whiff rate varies by horizontal pitch location:
Interestingly, he whiffs less on inside pitches from LHPs than RHPs. Next, let’s check out his slugging on balls in play:
Here you see a big difference. Against RHPs he maintains high power across most of the plate (it peaks in the middle of the plate, not surprisingly). But against LHPs his power is relegated to the middle-away, and it falls off sharply inside.
Putting the graphs together explains his overall weakness against LHPs. On inside pitches, he doesn’t whiff that often, but has little power. On the outer portion of the plate, he can hit for some power, but he whiffs at a huge rate.
If you project Howard as roughly a .450 wOBA hitter versus RHBs and roughly a .300 wOBA hitter against LHPs that works out to a 0.13 run difference per at-bat. That is over half a run every four at-bats, an enormous difference and significant cause for concern for the Phillies in the games they face Sabathia and Pettitte.
Print This Post
Jayson Werth is the anti-Howard (almost). Giving enough bullpen arms, you can take a lot of the power out of the Phillies lineup, but there’s always someone who can hit a starter.
Good point. Werth looks like about .450 wOBA versus LHP and .350 versus RHP. Not as extreme as Howard, but helps to pick up the slack against a lefty starter.
Sounds like a job for a ROOGY.
Sounds like it would make sense to at least sit him against CC and use him as a pinch hitter against Joba or Hughes.
I just have no idea who should be used in his place.
couldn’t really happen regardless of how bad howard is. who would they replace him with? matt stairs? maybe i’m forgetting someone else…?
Stairs is a lefty, too, as is Dobbs. The best righty bench bat is Francisco, and he’ll probably start in left with Ibanez as the DH (or vice-versa). I don’t really think there’s a good replacement.
Really, regardless of what might make baseball sense by the numbers there’s absolutely zero chance of Manuel benching Howard in the world series.
There’s zero chance of this happening, and plus he’s had some big hits against lefties in this postseason.
I agree. Good grief. They are not going to sit Howard against Sabathia in a million years, nor should they.
Ryan is 1 for 3 vs. CC; 1 for 6 vs. Pettitte.
Didn’t the Phillies have a good record against lefties this season?
Great article. Just a note, Howard’s severe lefty difficulty seems to have become a bit less severe post all star game, as he has .351/.434 (OPS/slugging) post AS, and .265/.309 pre AS against lefties (still not great, but eh). Also, interesting that Howard has a “plus” fielding number. I would never have guessed…
I think the numbers bear out what I see when he struggles. He pulls off against LHP when he is struggling, commits too early to bad breaking pitches away and fouls off inside pitches in the 1B dugout or weakly into the shift
When he is red hot and using the whole field, he can handle fastballs away from lefties very well. Unfortunately the last few years, that hasn’t been very often
In the playoffs so far he has been doing a good job laying off borderline junk pitches (a problem a times) and taking his walks. This trend is a positive for his chances against LHPs, since it is the key to his success (or non failure if you will) against them
Finally CC is the guy who I am worried about him facing with the lower arm slot and the sweeper slider. Pettite has a more loopy breaking ball, not the profile of lefty that gives him the most trouble
wonder if those righties are pounding him up and in? It seems like they are aiming at a very small window when you look how Howard can kill them inside.
Dave,
I love your graphs. They are spectacular. Please keep it up. Do you have graphs like this for each player, or are you making them one-by-one?
Sal thanks. I make them one-by-one, but the process is, at this point, pretty automatable.
Here is a question:
Say we have two players. Both are left-handed. Player A has an actual wOBA versus RHP of .450. Against LHP it is .300. Again, those are actual numbers in say 3 full seasons.
Say player B is .350 versus RHP and also .300 versus LHP in the same number of PA (versus both RH and LH pitchers) in 3 seasons as well.
Who has the higher projection versus LHP and by around how much?
Do you know why I ask that question?
If I’m doing this right, Player B basically has a league-average split (actually .050 seems a bit high but I might be misremembering) and thus his projected wOBA against lefties is .300.
Player A has a demonstrated platoon split of .150 but that gets regressed most of the way to the mean (let’s say 3/4 with 3 seasons of hitting data). So his true estimated platoon split is .075. His average wOBA is .413; arranging the platoon splits around that puts his estimated RHP wOBA at .432 and his estimated LHP wOBA at .376.
Am I warm? Player A is way better against lefties.
I should add that I’m assuming that 25% of his PA are against lefthanders. Which might not be true if he’s been platooned in his first 3 seasons, but eh.
MGL: Because you like to fuck around with us, rather than simply stating your point?
;-)
Paul is 100% correct.
Nick, I’m not trying to f***k with you. I am trying to illustrate why the idea that Howard is terrible (or not much more than an average hitter) against lefty pitchers and should be platooned, as some analysis have even said, is (likely) nonsense.
As Paul has shown, how can two lefty hitters with the same playing time and the same exact numbers versus LHP have such different estimates of their true talent versus lefty pitchers? The both hit .300 (wOBA) against lefty pitchers, but one batter computes to a projected .376 versus lefty pitchers and the other computes to .300! How can that be?
Because how a batter does overall (versus lefties and righties) tells us more than just how he does against LH or RH pitchers alone, even if we want to estimate how they will do in the future against either lefties or righties alone.
IOW, how a batter does against RH pitchers informs us on how he will likely do against LH pitchers and vice versa. Why? Because there is not much a spread in true platoon splits among ML baseball players yet there is a large spread in overall true hitting talent among ML baseball players. So if we see a large platoon split, like for a player like Howard, it is likely a fluke. So if a player does really well versus RH pitchers but terrible against LH pitchers, both the “really well” and the “terrible” numbers are likely fluky and the “truth” is somewhere in between.
Howard has a .741 OPS in the last 3 years versus LHP. How would be estimate his “true” OPS versus LHP? You might be tempted to just use the .741, which is not too good (not terrible of course) or you might be tempted to use the .741 and then regress that toward the league average for a LH batter of Howard’s physical characteristics, which might be around the same – I don’t know. Both of these methods would be wrong. You cannot ignore the fact that he also hit 1.042 in OPS versus RHP over the same time period (last 3 years) and in many more PA. This suggests that he is a very good hitter overall (which he is) and that the .741 is somewhat of a fluke. The fact that it is only in around 750 PA, the equivalent of a little over a year of full time play against all pitchers, makes it very reasonable that it could be flukey. We see flukey results in one season of play all the time. Heck, one standard deviation due to chance in 750 PA is around 30 points of OPS.
Anyway, do what would be the correct method for estimating Howard’s true OPS versus LH pitchers using his last 3 years of stats? Basically the same as Paul did above.
First we’ll estimate his overall true OPS. In his career, it is .966. We’ll regress that and call it .930, which is a typical projection for him. Now we have to take his observed platoon ratio (I like to use ratio – some people use a differential) and regress that. His observed platoon ratio for those 3 years is 1.042/.714, or 1.406. For the average lefty, it is 1.20 and we just don’t see that much variation among players in their true platoon splits. IOW, that 1.406 is likely very (but not completely) flukey. We might regress that 1.406 80% toward the league average of 1.20, to get 1.24. That is Howard’s “true” estimated platoon split.
Now we simply apply that to his overall estimated true OPS of .930 and the fact that he faced 62% RHP. That gives us an OPS of .810 versus LHP and 1.004 versus RHP.
.810 is a far cry from .741
So if we are going to talk about how Howard is likely to do versus LHP, let’s at least get our numbers right. .810 is higher than Werth and Jeter versus RHP. We don’t talk about how terrible Jeter and Werth are versus RHP, which they face most of the time, do we?
Just a quick comment – .810 (or .791 or .741) is worse than Jeter vs. RHP both this year and over his career (.817 this year, and .826 career). And for a first baseman playing in Philadelphia – it is bad. Not replacement level bad, and not really a candidate for a platoon given the alternatives on the Phillie bench – but possibly a reason to drop him a couple of places in the lineup against Sabathia. Not that any manager would actually do that and take the chance of looking stupid.
Edit: I was using the ESPN “3 year stats” as “last 3 years” but of course they are 06-08. But my point is still the same.
I made some mistakes with the numbers. Here are the real numbers and this time I used Howard’s last 4 years of data versus LH and RH pitchers.
First we’ll estimate his overall true OPS. In his career, it is .966. We’ll regress that and call it .930, which is a typical projection for him. Now we have to take his observed platoon ratio (I like to use ratio – some people use a differential) and regress that. His observed platoon ratio for those 4 years is 1.052/.719, or 1.46. For the average lefty, it is 1.20 and we just don’t see that much variation among players in their true platoon splits. IOW, that 1.46 is likely very (but not completely) flukey. We might regress that 1.46 80% toward the league average of 1.20, to get 1.25. That is Howard’s “true” estimated platoon split.
Now we simply apply that to his overall estimated true OPS of .930 and the fact that he faced 62% RHP. That gives us an OPS of .805 versus LHP and 1.006 versus RHP.
So .805 is an estimate of his true OPS versus LHP.
MGL,
thanks for stopping by and clearing this up. I have always been in the platoon Howard camp, but this makes a lot of sense.
How much data is enough to stop regressing his platoon split so far back to the mean? I’m not saying that none of his platoon split is luck – but 80% seems fairly arbitrary for a player with over 1,000 career PA against LHP.
I think that he will improve over time against lefties as he becomes a better hitter – but that doesn’t mean that his past performance has been based on luck or that we should expect him to substantially improve on his performance this year.
Curious – this would imply that the logic of using a platoon (such as the Gabes in Tampa) to cheaply approximate a much better player doesn’t gain as much of an advantage as one would think. Do players such as Xavier Nady not deserve their “designated lefty-masher” tags?
“Do players such as Xavier Nady not deserve their “designated lefty-masher” tags?”
That is exactly correct. I wrote about that over 5 years ago. There is almost no spread in true talent platoon splits among RHB. Therefore any large platoon split you see for a RHB is almost certainly a fluke, even over a large number of PA.
Ken, I regressed his platoon ratio too much. According to The Book, for 750 PA, you regress about 57%, not 80%. That would put his true OPS versus LHB at .791 and not .805, not a gigantic difference.
You also have the issue of how much to weight each year, which I don’t know. I typically don’t do any weighting at all for platoon splits, but that is probably incorrect, as you want to do some weighting if you think that the underlying talent can change over time. In Howard’s case, of course, his platoon ratio has been getting larger. So if do any weighting, his estimated true OPS versus LHP would get worse, maybe in the .775 range. Of course if you do weighting, you have to reduce your effective sample size as well, and then regress more.