Remember me

Who Will Throw the Next No-Hitter?

June 6, 2017

Over the weekend, Edinson Volquez threw his first career no-hitter. It was a special evening as the current Marlins and former Royals starter dedicated his performance to Jose Fernandez and Yordano Ventura. It was also a fascinating game from purely a baseball perspective. Volquez has had a bounce-back season following a rough 2016 campaign that featured a 5.37 ERA, but neither his current strikeout rate (8.3 per nine) nor his current walk rate (4.8 per nine) suggested this no-hitter was coming. In fact, based on the calculation we use in The Bill James Handbook, Volquez was just the 53rd most likely starter to throw a no-hitter.

The 10 most likely pitchers to throw a no-hitter include the usual suspects like Clayton Kershaw and Max Scherzer, but they also include a few more surprising names.

Most Likely No-Hitter
Player Chance of No-Hitter
Max Scherzer 36%
Clayton Kershaw 34%
Chris Sale 31%
Lance McCullers 26%
Stephen Strasburg 21%
Yu Darvish 19%
Carlos Martinez 16%
Dan Straily 16%
Dallas Keuchel 15%
Ervin Santana 15%


Lance McCullers may not be on the tip of baseball fans’ tongues when talking about elite starters, but based on his performance so far in 2017, he should be. He is one of only 13 qualified starters who is striking out at least 10 batters per 9 innings. He and teammate Dallas Keuchel are both in the top five in AL ERA and are a big reason why the Astros have the best record in baseball and a commanding 13.5-game lead in the AL West.

Volquez was not even the most likely Marlins starter to throw a no-no. That distinction belongs to Dan Straily, who is enjoying a breakout season. His strikeout rate is up nearly two strikeouts per nine compared to last season, and he has a 3.56 ERA.

Ervin Santana is the final surprising name. He does not possess the elite strikeout rate one would expect from a leading no-hitter candidate, but because of a combination of the Twins' elite defense, a repertoire of pitches that has traditionally induced weaker contact, and likely a fair bit of luck, Santana has limited batters to a .153 batting average on balls in play (BABIP) this season, easily the lowest in baseball.


The Handbook probabilities are based on, among other things, the batting average allowed projections from the Bill James pitcher projections. Because pitchers have logged quite a few starts since the offseason, we updated those batting average allowed projections with our in-season daily fantasy projection system that also incorporates performance to date.


COMMENTS (32 Comments, most recent shown first)

Looking at an old column from John, I think the percentage is for a no hitter at some point in a career. He says it's in the "Career Targets" section of the Bill James Handbook in the paragraph right before the table here:
3:13 AM Jun 11th
.....but probably a few of them will go up to John and whisper to him about it. I can't imagine they won't.
Call it a prediction. :-)

He might then take the article down and re-do it.​
11:39 AM Jun 10th
>(John, I hope you might take a look here -- for the sake of >anywhere else that you might be putting this, or any similar work >you might do later.)

An excellent suggestion, Maris, but unfortunately too late. There is a link to the article in the This Week in SABR email newsletter, so presumably the entire membership can read it and wonder.
12:38 AM Jun 10th
.....Well, I did something else to sort of check out the "walks" thing -- and now I'm even more at a loss to fathom where the article is coming from on that.

I looked at the actual 'Walks' records of the pitchers who have had no-hitters since 2010.
All of them.

Well there's an asterisk on that.
I looked at all of them, but I didn't include all of them in what I considered.

I excluded the pitchers that are known as 'great' or 'near great.'
Because, they're obviously among the pitchers who are most likely to have no-hitters, and almost without exception of course they have low walk-rates -- but using that data as in indicator of likelihood to throw a no-hitter is a logical red herring, because by far the dominant factors are that they have low hit-rates and that they're likely to be in their teams' regular rotations and far more likely than other pitchers to pitch deep into games.

So I ignored them -- guys like Kershaw, Arrieta, Scherzer, King Felix.

I'll digress for a moment.....

I'd guess that that's something very few serious sabmetricians would do -- to their detriment, I'm saying. I can imagine that someone like Bill might, but I doubt most would. They'd be inclined to do it more systematically, without such subjective judgment; they'd look at the walk rates of all of the pitchers who had no-hitters.

What I try to do in such things is to focus on what's really relevant, and eliminate things that only dilute what we're looking for, even if technically they're part of what we're looking at. BTW, that seems to me to be one of the hallmarks of the construction of the Win Shares system, far more than what I've seen of other systems, and a big reason that I find it so appealing.

Back to the subject...
Looking at those other pitchers who've had no-hitters since 2010:

There isn't any overall tendency toward low walk-rates.
If I were one of those serious sabermetricians I'd present the data. I'm not, because I didn't write it down, and I know that this may keep you from believing it, and I don't blame you.
I wasn't doing it to convince anybody, just to see for myself, and I'm sharing it with you. If some of you are interested enough to ask for the data, I'll get it together.

Those pitchers as a group probably did have a better-than-average walk rate -- but very little better than average. A couple of them had very good walk rates, but most were about average, and two or three had poor walk rates. Several had worse walk rates than Volquez has had for the last few years.
(BTW, John was talking about Volquez's walk rate just for this year, which arguably wasn't the best kind of sample to look at anyway. What I looked at was for the couple of prior years for each pitcher as well as the no-hitter year, unless the guy didn't have prior years or something like that.)

The walks thing in the article remains a mystery.
11:32 PM Jun 8th
(John, I hope you might take a look here -- for the sake of anywhere else that you might be putting this, or any similar work you might do later.)

Presentation counts. Clarity counts. Knowing what your stuff looks like counts.
10:24 PM Jun 8th
Just curious if I am understanding this correctly. This is saying that Scherzer, Kershaw, and Sale have a 36%, 34%, and 31% chance to throw a no hitter this year. That would mean there is a 71% chance that one of the three throws a no hitter this year. That seems very high, doesn't it?
10:17 PM Jun 8th
So......what's the verdict on the "walks" factor?
Or at least consensus?
Or best guess? :-)

Is John right that it is a factor in likelihood for a pitcher having a no-hitter?

As I said, maybe he's not going particularly on any "reasons," but just on actual data -- but if so, it would be important to know whether the data really looked at whether 'walks' is an independent factor, as opposed to just being a coincidental datum (good word) :-) because of being correlated to how good the pitcher is.
11:05 AM Jun 8th
In some circumstances, more walks may signal fewer chances for a hit, if you have a power pitcher who would rather walk people than compromise. I recall Bill talking about Nolan Ryan one time, and how Ryan--for better or worse--adamantly refused to throw an easy strike even with three balls, with the result that Ryan had lots of walks but also lots of strikeouts and no-hitters.
8:10 AM Jun 8th
Well, that's a thing where walks ARE pretty relevant!:-)

It's a big part of why a guy like Mantle had only 1 game with even 3 HR's,
and why a guy like Ted Williams never had a long-long hitting streak.
9:34 PM Jun 7th
While we're at it, who was most likely to have the next four-homer game?
9:22 PM Jun 7th
I love John's work in general, but it's not a good job when a chart is labeled so unclearly and when a supposed factor that is used is neither explained nor self-evident. We here are "experts" on this stuff (in varying degrees) :-) and we're left scrambling to figure out what he meant -- in vain.
8:17 PM Jun 7th
MichaelPat wrote: "The method attempts to estimate the likelihood of predicting a no hitter in the rest of a pitcher's career."

So the table has very little to do with who is most likely to throw the next no hitter. It looks like the method is saying that, of the people on the list, Scherzer is by far the most likely to get a no hitter this year, since he is older than most of the rest. Followed perhaps by Santana. But of course, the next no hitter will almost certainly be by someone not on the list.
3:45 PM Jun 7th
(join the club) :-)
3:08 PM Jun 7th
Is this probability table implying that there's roughly a 100% chance that one of Scherzer, Kershaw, and Sale will throw a no-hitter this year? And another ~100% probability that one of Strasburg, Darvish, Martinez, Straily, Keuchel, and Santana will also throw a no-hitter? That seems kind of extreme... how many expected no-hitters does this add up to, for 2017? Or am I misunderstanding what these probabilities represent?

- Matthew Namee
11:53 AM Jun 7th
Michael: My posts answer what you're asking. I can't do any better on it.

Mahd: Good point!!
You guys are coming up with some things that do weigh in favor of the idea that walks decrease the chance of a no-hitter. None of them strike me as strong factors, but all of them taken together maybe do add up to something significant.

As to whether John was thinking of any of those things: Really too bad he's not replying.
His answer might just be, "I wasn't thinking of anything, it's just that pitchers' likelihood of getting no-hitters is correlated with lesser tendency to give up walks."

To which I'd say, "Yeah, but have you (or anybody) shown it really to be an independent factor? Isn't it just that better pitchers tend to give up fewer walks, and better pitchers are the ones who are most likely to throw no-hitters?"

3for3: Yes, that's a factor that blunts any negative effect of walks, not just because walks create DP possibilities but also because they permit forces at 2nd on plays where there might be no chance to get the batter.
9:00 PM Jun 6th
As against that, the more walks, the more DPs.
7:59 PM Jun 6th
The more pitches you throw, the greater the chance someone will hit one for a base hit.
In what circumstances would that be untrue?
7:16 PM Jun 6th
So you are saying that the chance of giving up a hit is the same whether a pitcher faces 30 batters or he faces 27.
The only way I can see that being true is if the three extra batters are waved to first base.
If you pitch to more batters, the chance that you will give up a hit increases.
I don't see how this cannot be true....
7:13 PM Jun 6th
Maris, How about this? More walks equals more opportunities for top of the order hitters to get a hit.​
5:46 PM Jun 6th
OK, I'll try. :-)

Walks do increase the number of PA's, but that's largely irrelevant to likelihood of a hit, because:
-- There are always exactly 27 outs, and
-- Likelihood of a hit within any number of outs depends simply on batting average.

You pointed out the rate of hits per PA. That's completely misleading, because if the number of PA's is elevated by there being more walks, it won't result in more hits -- because of that second reason up there: Likelihood of a hit within any number of outs depends simply on batting average.

The operative denominator isn't "plate appearances"; it's outs.


That said, I'd add that there are reasons, relatively subtle ones, that could decrease the chance of a no-hitter if a pitcher walks more batters rather than fewer. I mentioned one of them before; bhal mentioned another. And also, when a pitcher gives more walks, it adds to his pitch count, which perhaps makes it less likely for him to continue being effective for 9 innings.

But I'm doubting that John meant those things in bringing up the walk rate. Too bad he doesn't usually reply to comments......
5:29 PM Jun 6th
No, I can't state it more clearly, nor do I think anyone else could. :-)

It's right there, in clear detail.
5:14 PM Jun 6th
Delete 'of predicting' in the second sentence, please. It should read...

The method attempts to estimate the likelihood a no hitter in the rest of a pitcher's career. It's a little like the Favourite Toy.

4:44 PM Jun 6th
This has been around for at least a decade... can't find the original piece, but as Dewan notes it's in the BJ Handbook every year. (Ahhh... so that's why James hates abbreviations!)

The method attempts to estimate the likelihood of predicting a no hitter in the rest of a pitcher's career. it's a little like the Favourite Toy.

Maris: I did not understand your point at all (although I do remember that bit in Ball Four). Can you state it more clearly?

Every pitch thrown has some small chance of resulting in a base hit. The more pitches thrown, the greater the chance one of them will be safely hit. You're arguing against that?

4:27 PM Jun 6th
BTW, anyone get those "percentages"?
3:21 PM Jun 6th
Michael: Your post seems to ignore my point, so I can't tell why you think it wouldn't apply.

Bhal: Yes indeed, that's a factor, but it's so rare that a pitcher gets pulled at all when he has a no-hitter that this would be only a tiny factor in the overall likelihood of getting one.
3:21 PM Jun 6th
MarisFan, well given pitcher usage nowadays, the more walks the more pitches thrown, thus increasing the likelihood you will be pulled before getting 27 outs, no-hitter or not
1:29 PM Jun 6th
Batters average hits in 23% to 25% of their plate appearances, so the more plate appearances you have, the greater the chance of a hit...
12:57 PM Jun 6th
......What you said is pretty much exactly like a thing that Bouton talked about in "Ball Four," about whether a pitcher tends to get more strikeouts in a high-scoring game.

Some of his bullpen mates were arguing that you can get more strikeouts in a high-scoring game because more guys come to bat, and you can get more strikeouts it you face more batters. Bouton argued back that there were just 27 outs per game regardless, and there was no reason to think that the proportion of K's vs. other kinds of outs would be greater in a high-scoring game. I agree with Bouton. They didn't. :-)​
12:25 PM Jun 6th
Not so!

Sure, it's more batters. But it's not more likelihood of getting a hit. Walks are negligible with regard to that; they're basically throwaways.
(Except to the extent that perhaps it's more likely for the next batters to get a hit with a man on 1st base, which I think is possible, but I doubt that's what John meant, and anyway it's far from a definite thing.)
12:21 PM Jun 6th
More walks = more batters = more opportunities for a hit
11:59 AM Jun 6th
.....and I quickly see another issue (there may be more -- stay tuned): :-)

What the heck do you mean by those PERCENTAGES for throwing a no-hitter?

Surely you don't mean that's the percentage chance of each of those guys throwing a no-hitter in a given game.
And I don't imagine you could possibly mean either that it's their chances of throwing a no-hitter during this season. The percentages seem high for that too.

Do you mean ever?
11:37 AM Jun 6th
What does WALK RATE have to do with likelihood of throwing a no-hitter?
I would have thought, almost nothing.
11:32 AM Jun 6th
©2024 Be Jolly, Inc. All Rights Reserved.|Powered by Sports Info Solutions|Terms & Conditions|Privacy Policy