Clutching At Straws

Every time I watch an MLB game I feel like I hear commentators and fans say something like this, “This guy is the one you want on the plate in the pressure situation, he is a clutch hitter” or “Here comes Mr Clutch, right when we need him”. Somehow every team has one, or more, player(s) that are ‘Clutch’, the ones that will get on base or bring home the runs when you really need them to in these clutch scenarios.  Are the fans and commentators right? What even is a clutch scenario, let alone a clutch player?

To do that we have to decide how much impact each at bat has.  I will use the metric Leverage Index (LI) which is essentially a measure of how critical a particular situation is by measuring the swing of the possible change in win expectancy. If you’re interested on the calculation, you should read this explanation from Tom Tango, the creator of the version LI that I will be using.

Fangraphs breaks down at bats into three categories.  With an LI of 1 being average, anything below 0.85 is low leverage and anything above 2.0 is high leverage. So for my clutch analysis we are going to define a clutch scenario as a high leverage at bat. All of this data comes from the 2002 season onward as that is as far back as Fangraphs goes with its LI.

Since 2002 just over 10% of all plate appearances are deemed to be high leverage and the stats for these appearances show that despite a higher walk rate batters are performing worse to the tune of 8 pts for wOBA. I will be using wOBA (weighted On Base Average) for the rest of this article to determine batter performance as it is the best widely available metric for this. So who have had the best seasons in these high leverage scenarios (Batters with 50 or more high leverage plate appearances in the season).

You’ve got some top names there with two seasons from Ryan Howard and Barry Bonds, Robinson Cano and Joey Votto. But also some slight surprises Adam Dunn and Magglio Ordonez but they are from their stand out seasons in 2004 and 2007 respectively. Then two even more surprising ones from Raul Ibanez and Victor Martinez in seasons when they ended up with pretty average a WAR figure for the season 2.3 and 2.1 respectively.

That gives us some interesting names but the real value in a metric is how good it is at predicting the future. So I took the 2,536 player seasons where they had faced 50 or more high leverage plate appearances and compared their wOBA for that season to the next seasons wOBA for high leverage scenarios. This is what we got.

There is virtually no correlation between high leverage wOBA one season to the next, R2 of 0.02, which means that across the MLB we have no idea if a player who had a good in the ‘clutch’ last year will be good this year. This may be because we are only looking a lots of small sample sizes which means this analysis is impacted by randomness quite significantly but since we have a fairly large data set it is pretty conclusive.

So, we cannot assume someone will be good in the clutch just because they were last season. But there must be players that are good in the clutch consistently, so lets just look at players career numbers and see who is best. As before we only have the data post 2002, I took all the players who had 200+ high leverage plate appearances since then and ranked them by wOBA.

That is a pretty good top 10 for the last 15 years, these 10 have high leverage wOBAs above .400 which is excellent and even in non-high leverage scenarios the worst is hitting .345 which is above average. This shows what I was expecting that best ‘clutch’ hitters are actually just the best hitters in the game doing what they do. Interestingly Mike Trout is only player who has slight dip in performance but the player I want to highlight here is Joey Votto. He is second to Barry Bonds but Bonds is Bonds and being second to him in any baseball statistic makes you very good.

He is clear of third by a far margin, as of the time of writing he had 643 high leverage plate appearances which is basically a season so this isn’t skewed by being a small sample size. His slash line in these high leverage scenarios is a ridiculous .353/.495/.639, that on base percentage is helped by a walk rate of 21.5%.  There is a reason this site offers a t-shirt about Joey walking.

Now some people may define a clutch player as one who gets better in these high leverage moments, I don’t like that logic because it harder for good players to be great than it is for the bad hitters to become good. If we were to do this and order the batters by difference in high leverage wOBA to low/med leverage wOBA, the top player would be Daniel Descalso. He has a woeful .293 wOBA in low and medium leverage scenarios, which improves to very good .387 in the high leverage moments. Of the 20 players with the highest increases only Khris Davis, Joey Votto and Ryan Howard had a wOBA above .340 (the marker for above average) in low and medium scenarios. So all of these ‘clutch’ players are average or worse/

I don’t know about you but I am far more comfortable saying that Barry Bonds or Joey Votto is the best clutch player since 2002 than saying it is, 1.8 lifetime WAR, Daniel Descalso. So, let’s go with Bonds is the best (Votto if you want to exclude Bonds due to PEDs) and Descalso is the most improved. Since we have the data lets have look at who is worst, place your bets.

Bottom of the 563 players who have had 200+ high leverage plate appearances is a man whose only real use, on a baseball field, is on base or out in center field. He draws the ire of the great Bill James, which you will know if you follow him on twitter, and goes by the name Billy Hamilton. Did you guess right? I imagine a few people did. His high leverage wOBA is .195, yes that is right .195. Given that .290 is considered awful his performance can only be described as pathetic.

Billy also makes the top 10 for what I like to call the Choke List, players whose wOBA decreases most. Now all of this top 10 have seen somewhere between 200 and 300 high leverage plate appearances, so once again small sample size warning, but with exception of Billy all of them above average hitters and some of them are great hitters.

Seeing the likes of Jose Abreu, Kris Bryant, Jeff Bagwell and George Springer on this list really shocked me. The numbers might be a bit hard on Bagwell as we only have data on his last few years of his career but for the others three that is a slightly worrying trend. Is this something that White Sox, Cubs and Astros fans have noticed?

From looking at all of this I think the concept of a clutch player statistically doesn’t exist, invariably it is just the great players doing what the do all the time no matter what the score is or how many are on base. I know this will fall on deaf ears but please commentators & fans stop using these phrases.

If you want to search the player list I created you can do here, just remember if someone isn’t there then they haven’t had 200+ high leverage plate appearances since 2002.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.