It's gotta be nice to have a day like Tom Boswell had yesterday. Looking for something positive to write about, he keyed in on Danny Espinosa Danny then went out last night and went 3-4 with 2 homers. Sure it wasn't exactly related to what he wrote about (he was talking about an improvement in BABIP and home runs aren't in play) but the gist was that Espinosa was going to be good and hell if he wasn't last night.
Of course I can't let a Boswell column pass without scrutinizing it for being overly optimistic and dammit it Boz hasn't done it again, with two big errors made, one mathematical and one logical.
Let's start off by agreeing that the general idea behind Boswell's column is sound. Espinosa's BABIP is crazy low. It will get better. I guarantee. But Boswell believes that "get better" means get back to at least league average (.297) if not Espinosa's minor league number (.323). I have to take issue with that sunshine and ho-hos view.
You see Danny has a terribly low BABIP but he's earning it. The easiest way for a player to get a hit on a ball in play is to hit a line drive. As long as it's not right at someone they should be ok. The average player has about 19% of his hits classified as line drives. Danny's LD% is at 10.7%, the lowest in the majors. In a broad sense his BABIP is right in line with what you'd expect based on the LD%.
But we can go even farther than that. If you don't get line drives than you have to rely on either ground balls to get through, which isn't nearly as likely, or fly balls to drop in, which there's an even smaller chance of happening. By looking at the BABIP for each of these specific types of hit we can see how exactly Danny is getting unlucky (or if he is at all)
The BABIPs by hit type are (roughly using this year's NL numbers - but it's pretty stable)
LD: .716 GB: .235 FB: .135
And Danny's numbers are
LD: .714 GB: .226 FB: .093
You can see Danny isn't just hitting LD right at people. He's in-line with the NL. His issues come with GBs and FBs. OK so you say that Danny should see more bloops fall in and seeing eye hits go through. Well... maybe. One of the way you can get a hit on a GB is the infield single. Danny is fast enough to get several of these a year. However his IFH number is only one higher than the super slow Adam LaRoche. If the number of infield hits he gets is more in line with what you'd expect, his GB% would end up being normal as well. So expect him to beat out a couple more GBs like last night. As for his FB number, well Danny's issue may not be bloops that aren't falling. His Pop-up percentage is among the highest in the majors (8th). It may be that he had a couple bloops grabbed by an infielder but I think it's more likely that he's prone to hitting pop-ups which are very easily caught flyballs.
The way Danny is hitting now his BABIP won't be getting back to league average anytime soon. Doesn't matter how many at bats he has. He gets no line drives and hits a ton of pop-ups. That's a recipe for a low BABIP. It will go up, sure, but it doesn't necessarily have to jump up 70 points to the league average like Boz suggests initially. It may only go up to .240 or .250 as long as he hits like this.
So here comes in the math error. Boswell notes that if he hit the league average BABIP he would have 14 more hits (5 doubles) and a line of .258 / .341 / .502 or closing in on star-level. That's not right. Or at least not right as I figure it. (Someone correct me if I'm wrong, please) If he had an BABIP of .297 from seasons start that means you take the number of balls in play he's had from season's start and multiply by .297 right? He's had 127 balls in play (180 ABs -46 Ks - 10 HRs + 3 SFs) * .297 = 38 hits on balls in play. That's the number he should have using the league average number. He has only 29 hits on balls in play so that's only an extra 9 hits. Figuring in Boz's slugging guess... you're looking at 6 singles, 3 doubles or a line of .266 / .361 /.527 (Jesus - what did Boz exactly do here? Am I wrong? This makes sense to me. Like why would his isoOBP drop in Boz's projection from what it is now? It shouldn't because the number of walks are stable. I see how he got 14 hits (he used the ABs rather than balls in play), but that doesn't explain why my 9 hits are making Danny better than his 14. I'm guessing when he added those 14 hits he back-added 14 ABs too, which isn't how it works on a couple levels. I'd have to check that though and I'm too lazy. I'm going out on a limb and saying I'm right. Someone knock me down please)
My estimate would be, well let's be generous and say .270 BABIP, and most of those NOT being XBH (because it's not LDs or deep fly balls that he's getting unlucky on). That's about 5 extra hits, and let's say 1 double. That's a line of .244 / .339 / .494. Less than the superstar turn Boz's BABIp idea would suggest, or the mere occasional all-star Boz's bad math would have you believe. It's good, but only bordering on something more.
AH! You're dismissing that .323 minor league BABIP, You say! Boswell notes in his chat that a players BABIP is pretty stable from the minors to the majors. He even shows examples! Not quite. (let's forget about the PA for Epsinosa are on the low end in comparison - and what's up with shooting ABs in there - that's not PAs. Stop cheating!) This is where the logic error comes in. Boz suffers from selection bias when comparing players. All the players he looked at were successful major leaguers. You can't do that. By only looking at successes, you are implicitly saying Danny will be successful too. Of course they had high BABIPs in the minors - that's what got them to the majors. And of course they had high BABIPs in the majors - that's why they are starters now. For a different comparison I took a look at Dan Uggla (who to me seems offensively very Espinosa like and Boz even mentioned in the column). His BABIP numbers went from .372 to .294. Of course Uggla is pretty successful too and that minor league number is crazy high. So I looked at old Nat Anderson Hernandez, a failed 2B man. His number went from .314 to .283. You can lose BABIp, it's just that if you lose it, most MI guys don't have the peripheral skills to keep a full-time major league job then. Danny is different because of his home run power.
In the end this seems a long winded way of saying that Danny isn't that good. That's the funny part. Danny is good enough right now (even before last night). It's all relative. Look at NL second basemen. He currently stands at second best in OPS and even before last night's outburst he was smack in the middle. That's with the little bit of bad luck and being a rookie. Chances are even with the low average, Danny would be a Top 5 2nd baseman in the NL for years to come. So after all this work to show that he's probably not going ot be as good as Boz thinks the point is - it doesn't matter. Even at where he is now, batting around .200, he's already good enough to stick around and he'll only get better.
Think we're looking at a legitimate ROTY candidate?
ReplyDeleteIn a normal year he'd be borderline, not because he isn't worth a vote but because it's hard for voters to look past a .220 average. But it's not a great year for NL rookies right now so yes, he's going to be right in the middle of the discussion.
ReplyDeleteI think he needs to hit 240 to really have a good shot unless he's 35 hr or something silly. He''s putting up a case for gold glove honors. Espi's play and Zimmermann's boucneback year make this season a positive.
ReplyDeleteBTW, Most interesting thing from Boz's chat yesterday was his numbers on Riggs being a huge neg P-thag guy. His numbers big outlier in record being much worse than ytou expect. Boz takes some heat but pretty cool giving the eam prez a brakout of manager's record vs. predicted record and showing how off riggs is.
What would his AVG be if he has a .270 or so BABIP the rest of the year? I think that's the more interesting question, rather than where would he be if he had a decent BABIP to date. The past is the past, is his AVG going to get up to .250?
ReplyDeleteHis average is lacking, but it isn't like the Nats have tons of options to replace him. To the degree that new talent comes in either via trade or FA, I think that 2nd base is down the list of areas that need improvement. He is a rookie and one would expect his numbers to improve over the next year or two. Will he be a superstar, probably not, but he probably won't be an embarrassment either.
ReplyDeleteHoo - to have a strong shot you are probably right but as long as no one pulls ahead from the pack people are going to be looking at him and evaluating.
ReplyDeleteI'm not sure Boz is the first to mention it but he's been on this since the end of last year.
(http://www.washingtonpost.com/wp-dyn/content/discussion/2010/09/10/DI2010091005803.html) Statistically it makes more sense to say "Riggleman's just an outlier with super bad luck" but when you are looking for reasons to get rid of the guy that does stand out.
cks - not with a .270 BABIP no. A .270 BABIp so far would have gotten him a .244 average - all things being equal (and that means homer and K rates remain the same) then you can take that .244 and spread it out for the year. You're just basically assuming the same performance in each block of time until the season ends. Something like .280 would make it close.
My guess is no - he doesn't reach .250, not unless he starts "straightening out" those homers into LDs.
Great analysis. A few comments.
ReplyDelete(a) I think there are two Danny Espinosas - RH Danny and LH Danny. You'll see wildly different results on BABIP depending on handedness. Not sure if I have a conclusion from that, just sayin'.
(b) totally agree with your concluding paragraph - the dude is good right now. 1.8 WAR for the season already (1/3 of the way there). But not sure if I understand your guess of .244 / .339 / .494. being just 'good, but only bordering on something more'? Wouldn't an .830 OPS hitter with above average baserunning and defense at 2B be more than just good?
(c) saw this the other day: Ryan Zimmerman has more WAR than all but 4 position players for the Nats. (didn't check it). Dude has played 8 games. How sad/hilarious is that?
Anon - I think he's more than just "not an embarrasment". With his slugging, fielding and cost I think he's something other teams would love to have, low average be damned.
ReplyDeleteWally
a) I'd love to see the BABIP by type of hit split by side of the plate. Then again how many at bats are we talking about at that point. We can see he hits even fewer LDs as a lefty. I think he has the same approach from both sides though but not the same power. Since the pop-ups aren't worse from the left side it probably means lots of lazy fly balls.
b) I wanted to go back and put in a "very" in there but decided why bother. Thanks for calling me out. But I'll stand by it in the sense that I tend to break players down to star, good, acceptable and bad and I don't think that pushes Danny into star territory. It might SHOULD do it, but it won't.
c) They were a great 8 games (they really were) but part of that is a defensive component that was projecting out to be his once in a lifetime defensive year. But even-ing that out you could still say hsi 8 games were worth as much as nearly ever other player on the Nats.
Harper - I found BABIP quickly on B-R, but not LD%, GB% and FB%. I'll keep looking, but here is overall BABIP.
ReplyDeleteLH BABIP - .191 (154 PAs)
RH BABIP - .333 (57 PAs)
Pretty stark, but you are right about the relative number of PAs. My guess, without more data, is that he has swing issues LH, and that a disproportionate amount of those popups come LH. But I do not think that he would necessarily maintain the same rates hitting RH against righties, which is why I said earlier that I do not necessarily have a conclusion from this. The better question is can he improve on his LH issues (I would be willing to bet on him, at least for another year or two), or failing that, would he do better RH v RH, or these numbers?
OK, a little more data.
ReplyDeleteLD% - 9.5% LH, 13.9% RH. Still low, but worse LH. Supports my thesis.
However, this doesn't:
IFFB% - 17% LH, 22% RH.
Also:
GB% - 42% LH, 36% RH
FB% - 48% LH, 50% RH
Not sure what to make out of this.
The flaw in Boz's logic is while most pitchers have a BABIP hovering around .300, BABIP varies from hitter to hitter. Boswell assumes that Espinosa is an average hitter. That still remains to be seen. Thanks for furthering quantifying those prospects, at least based on his current performance, admittedly a small sample size.
ReplyDeleteStill it is nice to see more of the new statistical models creeping into the mainstream press, even if used imperfectly.
We should be keeping better statistics on Boswell!
ReplyDeleteHis MO is to start with a premise that makes for great reading (and that's difficult with the home team he covers), and then pad it with all of the statistics he can, damn the truth. A Mark Twain writer.
But we love to read his columns.
For what its worth, I think Espinosa should be in the conversation for ROY.
Wally - I was looking at those same numbers when I made my last comment. I think it means a lack of power from the left side. Not a terrible amount less but when you hit 50% FBs even a little bit matters.
ReplyDeleteDSK - You're sorta right. Pitcher should float back to the average, hitters not so much. I think his real flaw is not understaning two things (1) sure most people are between .280 and .330 but like 7% aren't on the high side and 10% on the low side. 17% isn't trivial. (2) and actually it would be a lot more on the low side it's just that most of those guys never make it long enough to stick in an analysis. Danny could be one of those guys. I don't think so but you can't rule it out.
ND - I don't think he's purposely trying to mislead though. Rather he's only looking for the good news. It's not telling half-truths as much as reporting half-stories.