clock menu more-arrow no yes mobile

Filed under:

You can’t blame analytics for the Yankees’ struggles with runners in scoring position

Analytics have nothing to do with the Yankees’ shortcomings on situational hitting.

If you buy something from an SB Nation link, Vox Media may earn a commission. See our ethics statement.

Divisional Round - Boston Red Sox v New York Yankees - Game Four Photo by Elsa/Getty Images

The Yankees are known as one of the most analytically-driven teams in MLB, as baseball insiders estimate that their research and development department is among the largest in the majors. Under the direction of Brian Cashman, that approach has guided the Yankees through their “rebuild” and current resurgence, leading them to a 100-win season in 2018.

An early exit in the ALDS at the hands of the Boston Red Sox, however, has many fans pointing to the Yankees’ reliance on analytics as a possible cause for their troubles. This is largely because many of the problems that plagued the Yankees throughout the season - inconsistent starting pitching, questionable bullpen usage, and an inability to hit with runners in scoring position - reared their ugly heads in the series.

Fans have every right to be frustrated with the Yankees’ RISP failures. However, the Yankees’ usage of analytics isn’t to blame here. The view that analytics somehow hinders a good approach to hitting in clutch situations in predicated upon a warped view of “analytics” and how it is used by MLB teams, as well as some pretty big assumptions on what constitutes a good approach in said situations.

As I see it, the argument that analytics is to blame for RISPfail goes along these lines:

A) The analytic approach emphasizes power over contact in terms of hitting, willing to look over an increase in strikeouts for more dingers and extra-base hits.

B) Clutch situations call for situational hitting - i.e.; sacrificing power for contact, aiming for singles and not home runs

C) Therefore, the analytic approach leads to bad performances in clutch situations.

This syllogism sounds clear-cut and reasonable, but actually there are many problems with it. Firstly, point A, though not entirely wrong, is a simplified and exaggerated view of what is generally considered to be a good approach to hitting in the analytics community.

At its most basic level, all the analytic approach dictates is that hitters should aim to hit balls in a way that is most conducive to positive outcomes - hits, that is. Statistical data has shown that avoiding ground balls and consistently generating a certain level of exit velocity is the way to go. The point, according to the analytic approach, is not to achieve the highest exit velocity or to always aim for the fences, although it might seem that way to the uninitiated. It is to hit line drives and fly balls reasonably hard, because, over the long run, those are the types of contact most likely to be hits.

It is true that the analytic approach might lead to increased strikeouts for some hitters. That does not mean that the analytic approach encourages low batting averages and high strikeout totals. That is a myth stemming from the early days of sabermetrics, when low-average, high-strikeout sluggers were undervalued because of those qualities despite having positive attributes like power and patience.

The analytic argument is not that strikeouts don’t matter; it’s just that they don’t matter as much as traditionalists would have you believe. Aaron Judge is proof enough that high strikeout totals and great batting lines can coexist. Of course, the best hitters are able to consistently generate quality contact while reining in their strikeouts. That truism is completely in line with what the advanced metrics show us, as the top five hitters by wRC+ in MLB in 2018 - Mike Trout, Mookie Betts, J.D. Martinez, Christian Yelich, Alex Bregman - all have strikeout rates around or below league average. The problem isn’t the analytic approach - if anything, it’s that the Yankees employ none of the above hitters.

As for point B, it’s far from clear that sacrificing power for contact is the best approach in clutch situations. Just making contact may be preferable to striking out in some situations, but the opposite can be said of routine grounders hit to second or short in a bases-loaded, one-out situation. If the objective is to get hits, why not just aim to hit hard line drives all the time, regardless of the situation?

Sure, there’s no guarantee that such an approach will always lead to timely hits. But there’s literally no way to always come through in the clutch. Regardless of approach, there will always be instances when the Yankees fail to deliver the big hit. Sometimes Nathan Eovaldi will have his best cutter to pair with his 100-mph fastball. Sometimes Craig Kimbrel will throw a perfect slider and the hitter will have no chance. Even when Kimbrel misses with his fastball, sometimes Gary Sanchez will get just under it. The best hitters in baseball hit .300. Failure, with or without runners in scoring position, is the norm, not the exception.

This, I think, is at the core of the anti-analytic argument. Those who are frustrated towards analytics place unreasonable expectations on them, and on other hitting philosophies as well. There is no way to guarantee consistent success in clutch situations. The best we can do is look at general trends, identify what is most conducive to success over the long run, and try to execute that game plan regardless of the situation. In other words, all we can do is play the percentages.

The rest is ungovernable by laws and statistics. It’s also why we play the damn games, rather than simulating them on a computer. Sometimes the Yankees will come through, and sometimes they won’t. There will never be any certainty around that. There had better not be, because why else would we watch baseball?