The Kid Who Could Only Hit Doubles

March 9, 2018
Let’s imagine a baseball player.
Let’s imagine a player who has a unique and freakish skill: he can only hit doubles. No homers. No singles or triples. He can’t even walk…he just hits doubles. If a pitcher tries to walk him, he ends up on second base anyway. Our guy is Earl Webb, imagined through a Matt Christopher book.
Let’s go a little further in flushing his character out. Let’s say he has an on-base percentage of .300. Three out of every ten times he comes up to the plate, he gets a double. The other seven times he’s an easy out. That is an abysmal on-base percentage, of course.
But he would also have a .600 slugging percentage. If he collects six total bases per ten at-bats, that’s a .600 slugging percentage. You can probably live with a .300 on-base percentage if you’re getting a .600 slugging percentage, right? Let’s bat him fifth or sixth and call it a day.
Well hold on. Hold on for just a second. I have another twist to throw at you.
Because the best superheroes have layers of complexity, we’re going to throw in one moredetail about our guy’s oddly specific power: his ability to hit doubles only works if there is no one on base when he is batting. If someone’s on base, he’s a nearly useless hitter.
So maybe he’s not worth playing. I mean, it’ll be great to have a .600 slugging percentage in the batting order, but it’s not that great if he can’t use that slugging percentage to actually drive in runners. And he’s going to have a lot of at-bats where he’s hitting below replacement level. So he’s probably not worth it.
Except it turns out that he is a superlative defensive player at a key position. He saves a good number of runs on defense.
And he comes cheap. His annual salary is less than what Clayton Kershaw makes for one start.
What do you do with a guy like that? Would you want that kind of a headache? And how would you use him? What would be the best way to cash in on his particular skill set?
I think, if you thought it all the way through, you’d decide that he should play. Because you’d want to maximize the possibility of a double, you’d probably slot him in the leadoff spot. And because you’d want to maximize the chances of him hitting leadoff, you’d want him to play for an NL team.
*             *             *
Which is handy, because Billy Hamilton does play on an NL team.
I was thinking about Billy Hamilton last night, and I was thinking about the fact that his on-base percentage just does not do justice to who he is as a player. It is the one statistic that almost everyone who writes about Hamilton focuses on: his abysmal on-base percentage. Almost every article you read about Hamilton will be about:
a)       How he might improve his on-base percentage by trying one thing or another, or
b)      Why the Reds cannot bat him leadoff, because his on-base average is so terrible.
Billy Hamilton certainly does have a terrible on-base percentage: his career mark is .298. But he scores a lot when he gets on base. In an off year, Hamilton ranked 16th among qualified players in the percentage of times he scored after reaching base:
Aaron Judge
Jose Bautista
Cody Bellinger
Billy Hamilton
Paul Goldschmidt
Jose Ramirez
Curtis Granderson
Byron Buxton
The year before that, Hamilton ranked tied for fourth among qualified players:
Ian Kinsler
Adam Duvall
Rougned Odor
Charlie Blackmon
Billy Hamilton
Kris Bryant
Ian Desmond
The reason for this is very obvious: once he reaches base, it is extremely difficult to keep Hamilton from scoring. He is pretty bad at getting on base: he doesn’t have the power to knock one into the cheap seats, and he doesn’t have the contact skills to reach base regularly. But Billy Hamilton is excellent at moving around the bases.
Hamilton’s speed doesn’t show up in the traditional equations for on-base and slugging percentage, of course. But it’s possible to tweak those formulas to give him a leadoff slugging percentage.
How do we do that?
We convert his stolen bases into total bases, and we penalize him for the times he has been caught stealing. The result is a player who had a slightly worseon-base percentage, but a much better slugging percentage:
Actual OBP
Actual SLG
Leadoff OBP
Leadoff SLG
B. Hamilton
When you factor in the times he was caught stealing, Hamilton’s on-base percentage drops twenty points. But his slugging percentage, once you factor in his stolen bases, jumps nearly eighty points. He becomes last year’s version of Mark Trumbo:
B. Hamilton
Mark Trumbo
Let’s just stop a moment and appreciate the historical nature of this article: we’ve reached a point where we’re comparing Billy Hamilton to Mark Trumbo. Uncharted waters, I suppose.
But that measure actually undersells Hamilton, because it considers all of his at-bats. What about just looking at his at-bats as a leadoff hitter?
This gets trickier, because while we can get his hitting splits as a leadoff hitter, I can’t find anywhere that lists his stolen base numbers split on leading off an inning versus not leading off an inning. But we can at least make an approximate guess of things. 246 of Hamilton’s 633 plate appearances came when he was leading off an inning last year, or 39%. We can run his numbers leading off an inning, swapping in 39% of his stolen bases and caught stealing events, and figure out the kind of hitter Hamilton was at the start of innings.
We get:
Actual OBP
Actual SLG
Leadoff OBP
Leadoff SLG
B. Hamilton
Leading off an inning, Hamilton was…not bad.
B. Hamilton
Trey Mancini
He was opposite-world Trey Mancini…he was Trey Mancini if you took Mancini’s power hitting and converted it to speed on the bases.
In an ideal world, you wouldn’t want a hitter like Trey Mancini batting leadoff. In an ideal world, Trey Mancini would be hitting at the back of an offense, where that slugging percentage could convert into runners being driven in.
This is where the trick loses its magic. We can pretend all we want that Hamilton is a pull hitter all we want, but the reality is that the bases he advances don’t do the same thing as the bases that Mancini or Trumbo collect. Hamilton is the living incarnation of one-run strategies, and trying to understand him using multiple-run metrics like slugging percentage is a little like trying to understand a fish by looking at clouds.
*             *             *
There is nothing conclusive about the numbers I’ve shown you so far: I’m just trying to get us out of the mindset that a metric like on-base percentage doesn’t tell the whole story about Billy Hamilton.
But I am going to come around to a conclusion: I think the Cincinnati Reds should absolutely bat Billy Hamilton in the leadoff spot.
 Hamilton hit leadoff in 135 games last year, and the Reds went 61-74 in those games. That’s not good, of course…that translates to a .452 winning percentage.
But the Reds went an abysmal 7-20 in games where Hamilton didn’t lead-off. That’s a winning percentage of .259. They played like the 1962 Mets without Hamilton.
They scored 660 runs with Hamilton batting leadoff, an average of 4.9 runs per game. Without Hamilton, the Reds scored 92 runs in 27 games, or 3.4 runs per game.
And their pitchers did appreciably worse without Hamilton in centerfield. When Hamilton played, the staff allowed 5.2 runs per game. When he was out of the lineup, the Reds pitchers allowed 6.2 runs per game.
Let’s table that:
2017 Reds
With Hamilton Leading Off
Without Hamilton Leading Off
Most of Hamilton’s days off were rest days, but there was a stretch in September when Hamilton missed fourteen straight games. The Reds record without him was 5-9, a .357 winning percentage.
I don’t want to put too much stock in those run differentials: it is unlikely that Hamilton generated 1.5 runs in production more per game than whomever the Reds substituted in for him. And though I’m sure Hamilton’s stellar play in the outfield is a contributing factor to the difference in runs allowed, it is a big stretch to say that he is worth a run per game on defense.
But the Reds do seem to play betterwhen Billy Hamilton leads off. Maybe they shouldn’t do better - maybe the math suggests that Billy Hamilton is a liability - but it is difficult to argue with the results.
*             *             *
I’m not impartial on the subject of Billy Hamilton. He’s my favorite player to watch in baseball, and I root for him every year. I root for him because he is so different, and because sometimes it is difficult for teams to know what to do with players who are different. What do you do, really, with a kid who can only hit doubles?
The Reds efforts to win baseball games with Billy Hamilton on their roster has been one of the most interesting storylines I’ve followed over the last few years in baseball. Let us all remember, for a moment, that when Billy Hamilton first blipped on our radar as the guy charging at Vince Coleman’s minor league record for stolen bases, he was a right-hand-hitting shortstop prospect. He has transitioned to a centerfielder who now hits mostly left-handed. Let’s remember, too, that Hamilton developed those skills simultaneously, under a great deal of scrutiny and competing at the highest levels of play. At times, Hamilton’s story has been a frustrating story, and I’m sure that no one is more frustrated with the annual conversations we seem to have about Billy Hamilton than Hamilton himself.
Billy Hamilton is a good player. He isn’t going to win any MVP awards, but his presence on the Reds is a net positive on a team that is lacking in positives. Let him hit where he’s comfortable. Let him hit where he’s useful.  
Dave Fleming is a writer living in western Virginia. He can be reached on this site and at 

COMMENTS (24 Comments, most recent shown first)

Brock pointed out a flaw in Dave's analysis, and here's another:

In the comments, Dave mentioned that the Cubs leadoff hitters had a better OBP and SLG than the Reds' leadoff hitters, but had a similar total of runs scored.

This ignores that the Reds actually had better slugging percentages than the Cubs at the 3-4-5 spots.

3rd spot: Reds .581, Cubs .462
4th spot: Reds .495, Cubs .480
5th spot: Reds .475, Cubs .429

Those numbers have nothing to do with Hamilton. To ignore them is like ignoring his outs on the bases -- it makes him look more valuable than he really is.

So yeah, the Braves had 100 runs from their leadoff spot -- which is less than the Reds -- but they also had Matt Kemp batting cleanup. The Giants scored 102 runs from their leadoff spot -- with their No. 3 hitters having a .709 OPS. The Phillies were at 104 -- with Maikel Franco and Tommy Joseph batting cleanup.

Hamilton's also not exactly playing in the Astrodome. At home the Reds had an OBP 12 points higher and a SLG 23 points higher than on the road.
1:48 PM Mar 16th
Brock Hanke
I actually did an analysis like this, back in the 1980s or 90s. I was interested in seeing what Vince Coleman's stats would look like if you added SB to Total Bases when computing Slugging Percentage, and removed CS from Times on Base when computing OBP. I then looked for the closest player that year to Coleman's adjusted stats - the closest OBP and closest SLG. It turned out to be Ron Kittle, not in a great year, but a hilarious comparison to Coleman. The one real complaint I have about your article is that, when you have a leadoff man who steals a lot of baes, he will score a lot of runs himself, but he is STEALING some of those runs from later hitters who did not get a change to hit because your speedster had been thrown out trying to steal. Evaluating a speedster by your method overrates him, just as it did for me when I tried it.
2:36 AM Mar 14th
Thanks for the interesting article, Dave. I'm anticipating your annual "here's the team to watch this season" piece. Hope it's in the pipeline. It's one I look forward to each spring.
8:27 PM Mar 11th
Plausible :-) but the one I meant was, If Henry Cotto is a baseball player, I'm a jet airplane.
6:31 AM Mar 11th
Was it..."you might as well run up a flag saying, 'we don't know what the hell we're doing!'"
12:36 AM Mar 11th
.....remember Bill's line about Cotto? :-)

Cruel, and not sure it was fair, but funny (if you're not Henry Cotto) and memorable.
8:59 PM Mar 10th
Don't forget old Henry Cotto.
6:51 PM Mar 10th
Isn't Billy Hamilton the second coming of Omar Moreno?

How many more years does Hamilton have before he is taken by the fate of the low-OBP speedsters, drifting from team to team, demonstrating that speed on the basepaths alone is a weak weapon? I'm guess two. Two more seasons with the Reds, and then he sets off on that odyssey that took Juan Pierre to five more teams, Vince Coleman to 5, and Otis Nixon to 8...
4:07 PM Mar 10th
That's a lot of outs to start a game with.

let's bat him 9th, go for the double leadoff hitter thing.
3:07 PM Mar 10th
Steve: You're kind of right. :-)
I do see them as subtly different but very close.

That was part of my point: "Mediocre" was only slightly unfair, at most.
10:15 AM Mar 10th
Another way to phrase CharlesSeager's point about the Reds leadoff runs scored....

The Reds leadoff hitters finished dead last in the NL in on-base percentage. They were dead-last in slugging percentage. They were dead-last in wRC or wOBA, or whatever other metric you'd like to look at. They weren't just last...they were comfortably in last...the Pirates were 14th in leadoff on-base percentage at .309...the Reds were at .295. That's a big distance. In every hitting metric, the Reds leadoff batters, mostly Hamilton, were hot garbage at leading off.

But they Reds scored 105 runs from the leadoff slot, which ranked 9th in the league. Their leadoff men scored one fewer run than the Cubs (106). The Cubs had a 30-point advantage in on-base percentage and an 80-point edge in slugging percentage, and their leadoff hitters scored one more run than the Reds.

It is possible to look at the Reds leadoff numbers and come away thinking it's a problem...but it's also possible to come away from those numbers thinking that there's SOMETHING there that we're not adequately crediting. Every hitting metric says that Reds leadoff hitters should have finished dead-last in the NL in runs scored. They didn't....they finished in the middle of the pack. Middle of the pack ain't great, but it's better than they had any right to be. If mediocre means 'average' I agree that the Reds leadoff hitters were mediocre. But if we're using the term 'mediocre' to mean 'the Reds absolutely have to change that', well, I disagree.
8:09 AM Mar 10th
What strikes me about that first chart is the placing just a few slots below of Byron Buxton. He's at least Hamilton's equal as a centerfielder and probably has more offensive upside. So if you're looking for a really exciting player...

BTW Maris: 'mediocre' means 'middling', or at least it used to.
8:02 AM Mar 10th
Hi Dave,

As a Reds fan, and I suspect like many Reds fans, it's hard for us to look at Billy Hamilton and see anything but disappointment. Perhaps it's not fair, and I don't want to give up on a 27-year old, but we all watched with great anticipation when Billy was coming up through the farm system. He hinted at in 2011 when he played for my local team (Dayton Dragons) and stole over 100 bases, and then really blossomed the next year (mostly at Pensacola and Bakersfield). It wasn't merely the 155 steals in was that he hit over .300 AND he drew 86 walks, resulting in a .410 OBP. It was the combination of amazing speed AND seemingly the ability to get on base that got everyone excited. When he came up and it turned out that one of those skills was seemingly a mirage, it was a big let down.

I will say that, watching him up close, he's a spectacular defensive player to watch, especially when he gets a chance to leverage his speed to track down a fly ball.

I hear he's continuing to work on his bunting, this time with Rod Carew. Maybe that will help some.

He also appears to not have benefited much from switch-hitting. His BB rate per PA is twice as high as a left handed hitter vs. as a right handed hitter, and his OBP is 40 points higher as a left handed hitter. I don't know if it's too late for him to change to lefty-only at this point after being a switch hitter for so long, and I'm sure it's easier said than done, but it makes one wonder.....

7:39 AM Mar 10th
The number of runs Hamilton scored was mediocre.

Why am I saying this? Last year, an average team’s leadoff man scored 109 runs. Hamilton’s team, the Reds, got 105 runs from its leadoff men.

Obviously, being four runs below average isn’t awful, and it’s largely owing to Reds leadoff men hitting only 9 home runs—average was over 20 (!)—but as a sheer run scorer, Hamilton is only so-so, and if he loses any foot speed, he’s hosed.
6:58 PM Mar 9th
Very interesting article, Dave. I've always wondered about the balance between speed and on-base acumen for leadoff hitters; specifically, could one quantify the break-even point for comparing the effectiveness of speedy, low-OBP leadoff men versus slow, high-OBP players? Would you rather have a .330 OBP leadoff man who ran the bases like Juan Pierre, or a .370 OBP leadoff man who ran the bases like Bruce Bochte? I think since everyone on BJOL is a student of Bill's work, many of us might lean towards OBP as the end-all be-all quality of leadoff effectiveness; but, there has to be a balancing point at which the faster player becomes more effective even if on base less frequently.

Thinking back to Bill's wonderful essay "Rain Delay", from the 1988 Abstract, where he eventually compared Tim Raines and Wade Boggs, both at the peak of their careers, to determine the best player in baseball ...of course there were many things to consider, but perhaps the most glaring contrast was their baserunning ability. Raines was one of the greatest baserunners of all time, while Boggs was decidedly slower than average (though not Ernie Lombardian). But would that distinction lead you to prefer Raines over Boggs at leadoff, considering Boggs' OBP may have been 30-40 points higher? I think in that case I would prefer Raines, since he certainly had a very good OBP -- but there has to be a point at which slow and steady wins the race, no?

Billy Hamilton is one of the most extreme cases; his OBP is SO bad that it does make you question his fitness for the leadoff role -- and yet he is SO fast that, as you've demonstrated, it might still be worth it to lead him off, as he does seem to score so frequently when he does get on base. I'd love to see a study about this question (tl;dr - what's better, a good OBP slow leadoff man or a low OBP speedy leadoff man; and what's the statistical point at which the superiority of one or the other becomes clear?) but I wouldn't know how to go about it -- there are so many other variables (power, quality of hitters behind the leadoff man, ballpark effects, etc). If I could phrase it right I'd Hey Bill it, but it's kind of a jumble in my mind, as you can probably tell.

Anyway, thanks for another interesting read!
5:03 PM Mar 9th
BTW: Is he that great a fielder?
I have no impression of my own -- haven't seen him enough.
For what it's worth, Win Shares shows him as good-but-not great, except in his first full season, where it does show as great.
3:06 PM Mar 9th
Let me add to that: I don't think they're necessarily unanswerable.
In fact we could probably get some hints (although not easy answers) by looking at the stats of the guys hitting behind him 'with' Hamilton on base and 'without.' One reason it wouldn't be easy (or straightforward) is the confounding factor that whether Hamilton gets on base or not might be correlated with how well the pitcher is going.

BTW, regarding my having noted that the field of sabermetrics thinks (I think) that this question has been largely answered and that there's no significant benefit, maybe none at all, I would suggest that even if this is so in general, that wouldn't necessarily apply to an extreme case like Hamilton.
2:51 PM Mar 9th
I think that's entirely right Maris: a lot of Hamilton's value probably lies in the unanswered (and unanswerable) questions. How much does his presence in the batting order/on the bases influence the opposing pitcher? What benefit do hitters behind Hamilton get when they're getting fastballs? Does it help the pitchers to know that he's out there in center field, ready to gobble up any fly ball that stays in the yard? What is he like as a he a net positive in the clubhouse or a net negative?

The Raines comp wasn't supposed to be taken seriously, of course. It was a weird year for Raines...he had one of the best on-base percentage of his career, but he just didn't score that much. Just a fluke event...the next year Raines was on base about seven or eight fewer times, but he scored 123 runs. Runs scored rates fluctuates wildly among players, often for reasons that are hard to deduce.
2:41 PM Mar 9th
......but I have to say, you're right that it was unfair to say "mediocre."
Middling would be more accurate; I don't think your figures say anything beyond that.

I would also offer the thought that for a player like Hamilton, his actual offensive value depends a lot on the unanswered question (I consider it unanswered) of how such a player affects the production of the hitters behind him. I know that sabermetrics largely considers this an answered question: There's no significant benefit, and if mishandled, it can be a detriment. I'm not convinced, and I think it's possible he's a significant positive for other hitters when he reaches base.
10:31 AM Mar 9th
Dave, we can say this: You really love the guy. :-)
I think you're trying a little too hard.
Like, the comparison to Tim Raines in 1986:
Hamilton's run context was about 11-12% higher (taking account of both league contexts, which is 10% higher, and park factor, which is 1-2% higher).​
10:24 AM Mar 9th
What I'm looking for in a baseball article is "here's a new/fun way of looking at something."

This does that perfectly. Thanks Dave!
10:09 AM Mar 9th
You state that 'how many runs (Hamilton) scores' is 'mediocre.'

Let's unpack that. Is that actually true, or does it just seem true?

Hamilton scored 85 runs in 633 plate appearances, which calculates to 13.4 runs per 100 PA's. How does that compare to the other lead-off hitters in his division?

Hamilton (135 games) - 13.4

Jon Jay (51 games) - 15.0
Zobrist ( 40 games) - 11.7

Villar (70 games) - 11.7
Sogard (50 games) - 12.4

Carpenter (89 games) - 14.6
Fowler (50 games) - 13.8

Frazier (63 games) - 12.1
Marte (49 games) - 11.4

Hamilton's rate of runs scored per plate appearance is well behind Jon Jay and Matt Carpenter, and about even with Dexter Fowler. He's ahead of everyone else.

Of course, Jon Jay benefited from a platoon split, as did MOST of these guys. Hamilton was the only steady leadoff guy in the division...the only guy who didn't get a platoon break...and he holds his own against the competition. I don't think the number of runs he scores is actually just seems mediocre because a) he spent a lot of his earlier years hitting in the 8th slot, and b) NL leadoff hitters get dinged compared to their AL counterparts b/c they're often hitting after a pitcher makes an out.

Here's a fun one, just to throw out. In 1986 Tim Raines led the NL in batting and on-base percentage, the only time he led his league in those catagories. How'd he do in runs scored per 100 at-bats?

He was basically Hamilton: 91 runs in 664 PA = 13.7 R/100.
9:54 AM Mar 9th
P.S. Since last year was his best run-scoring year and while his number of runs was mediocre, but, as you said, he missed a lot of playing time, I looked at how he would have ranked in the league if we fill in all the missing time (and also fill in the missing time for other guys in the league who were up there).
He'd be about 20th, which is similar to his 'unadjusted' rank (tied for 23rd-26th).
I'm not sure we needed to give him this benefit of filling the missed time, but to be fair, I figured I'd look at it.
9:48 AM Mar 9th
This is an odd article. The premise is interesting but bottom-line-wise, I don't think it tells anything.

You talk about Hamilton's runs-scored per times on base, and you trumpet it as a highly significant thing.
It's interesting, but it isn't -- because the basic thing is just how many runs he scores, and that figure is mediocre.
9:23 AM Mar 9th
©2018 Be Jolly, Inc. All Rights Reserved.|Web site design and development by|Terms & Conditions|Privacy Policy