It’s free Thursday here at Down on the Farm. Our work is 100% supported by our readers, so if you’re enjoying the newsletter please consider becoming a paid subscriber. Thanks for reading everyone!

Subscribe now

Share Down on the Farm

Hi, my name’s Matan. I’ve previously published baseball research on Medium. With Drew Haugen taking a well-deserved post with the Phillies, I’ve joined the team to help contribute to analysis in his stead. Given Down on the Farm’s focus on player development and the minor leagues, it seemed appropriate to begin by addressing a fundamental question: How important is contact ability for prospects? Without further ado, let’s begin with a simple plot…

Player Whiff% does not have a strong correlation with overall batting success at the major league level. This (lack of a) relationship exists between Whiff% and OPS as well. Given this information, it’s easy to construct a narrative which involves celebrating or lamenting (depending on one’s point of view) the centrality of the home run and corresponding devaluation of the “pure contact hitter” in modern baseball.

However, as many readers are doubtless aware, drawing this conclusion from the above plot is at best dangerous and at worst, flatly incorrect. MLB players are not a random sample of the baseball-playing world. They’re the cream of the crop, expected to perform at the highest level of their profession. Players with high whiff rates in major league baseball likely possess exceptional skill in other areas, which compensates for their contact-related deficiencies. In other words, there is ample potential for selection bias to cloud the importance of contact ability at the MLB level.

Selection bias can also have another, perhaps paradoxical, effect. If MLB players are chosen based on their ability in the most important facets of play, the resulting talent variation in those aspects may be smaller than the spread of less crucial skills at the major league level. For instance, consider the distribution of home runs and stolen bases in MLB during the 2023 season. A small number of speedsters, including Ronald Acuña Jr and Esteury Ruiz, accounted for the lion’s share of stolen bases. In contrast, the league-wide distribution of home runs was far more even. This difference can be seen in the Lorenz curves below, where a diagonal line would represent a perfectly equal distribution…

Many sluggers go entire seasons without stealing a base. Sticking in the majors with zero power is nearly impossible. Now imagine an alternate universe where a stolen base was worth a full run. The above curve would certainly look very different.

If selection bias may be clouding the importance of contact ability within MLB, it is imperative to understand how it impacts which players are in the majors. On the most basic level, there is a clear gap in the lower bounds of acceptable contact ability between the highest and lowest levels of the sport. For instance, from 2021 to 2023 there were 41 players with 100+ plate appearances and a whiff rate >= 40% in the 10 team Low A Florida State League, compared with just 26 in 30 team MLB (despite the large gulf in the quality of pitching opposition). For our purposes, a more apt comparison would be between MLB and AAA, the highest level of minor league baseball. As a primitive method of comparison, one can juxtapose the distribution of raw whiff rates at both levels…

Batter whiff rate distributions at AAA and MLB tend to be relatively similar, with a slight preponderance of contact bats in the majors. However, there is a clear issue with this comparison. MLB pitchers are significantly more talented than their minor league counterparts. Whiff rates may be an informative measure of contact ability within a given level of play, but they have less utility when comparing players between differing levels. Performance against various velocity and/or movement thresholds (such as 93+ mph fastballs) are often used to attempt to address this problem, but these thresholds can be arbitrary and may shrink the usable sample of pitches significantly.

In lieu of raw whiff rates or particular thresholds, it is possible to create a fairly objective standard for contact ability with a modeled approach of an “expected whiff rate.” This XGBoost model uses typical pitch quality inputs (such as pitch location, velocity, movement and release point) utilized in Pitching+, PitchingBot and other prominent pitch models to assess the likelihood of a whiff on a given swing. Instead of using the output to evaluate the quality of a pitch or pitcher, it can be employed to evaluate a batter’s ability to make contact on a given pitch compared with what the model would expect for a theoretical “league-average” batter. Importantly, the expected whiff rate model was trained exclusively on MLB data, so AAA pitches (and resultantly batters) should be evaluated on the same standard as their major league counterparts. Here’s how the distribution of whiff rates versus expected differed for full-time MLB batters, part-time major leaguers and AAA batters in 2023…

Now that’s a significant difference! Full-time (500+ PA) MLB batters have an average whiff rate over expected ~6% lower than the AAA group, while the MLB part-timers enjoy a typical ~2.5% advantage. Unsurprisingly, there seems to be a large gap in contact ability between AAA and MLB players. Perhaps of more shock value is the disparity within the ranks of MLB as well. Contact ability appears to be a highly valued trait, maybe even a prerequisite, for everyday MLB hitters. In fact, the (only) 16 full-time players with whiff rates over expected greater than 5% also posted a ridiculous average barrel rate of 13% (versus 9% for all 500+ PA batters). While it’s difficult to tease out the effects of selection versus a power over contact hitting approach, it certainly appears that hitters with below average contact skills must compensate with elite raw power (or other skills) to succeed as impact players.

There are no hard-and-fast whiff rate over expected rules, but here are some useful heuristics:

  • In 2023, the league average expected whiff rate in MLB was ~26% versus ~22% in AAA. This should be reflective of a difference in pitch quality on swings.

  • While MLB batters actually whiffed ~26% of the time, AAA batters swung and missed at a 27 percent rate. As such, this seems to be reflective of a ~5% gap in pure (independent of swing decisions) contact ability between the mean of both levels. This is approximately a full standard deviation of MLB contact performance.

  • There were just 4 (of 136) MLB batters with 500+ PAs and a whiff rate over expected of 10% or more in 2023. They were James Outman, Brent Rooker, J.D. Davis and MJ Melendez. Aaron Judge (who was injured for much of 2023) is the singular star above the 10% mark. This seems to be a reasonable, though crude, lower bound for contact ability for consistent MLB performers.

Using whiff rate over expected to compare the contact ability of AAA and MLB players is useful in theory but requires empirical validation. Just under 100 players took 100+ plate appearances at both the AAA and MLB levels in 2023. In AAA, those players had a mean whiff rate over expected of +3.4%. That cohort’s actual whiff rate rose 2.7% in MLB (though their whiff rate over expected dropped a bit more, to +2.2%). This slightly lower figure is largely unsurprising. MLB teams are more likely to call up (“select”) batters that they believe will adjust well to major league stuff. Overall, the results of these league-switchers suggest that the whiff rate over expected model does a reasonably good job of standardizing contact ability for level of competition.

An important factor to keep in mind is that a given batter’s expected whiff rate is largely dependent on their swing decisions, while whiff rate over expected adjusts for this factor. A glaring example of this is Javier Báez’s 2023 season. Báez had a shockingly high expected whiff rate of 34%, due to his expansive approach. However, his actual whiff rate was 1.5 percent lower. A contrasting example is Brett Phillips’ 2023 AAA campaign. His whiff rate was under 30 percent, but his +11% whiff rate over expected was suggestive of subpar contact ability.

Given that contact ability is a key ingredient for MLB success, whiff rate over expected may be a useful performance indicator for AAA players. Here are the expected whiff rate numbers for a few notable prospects…

Everson Pereira, Brennen Davis and Marco Luciano stand out in a negative sense, as their high whiff rates over expected are ruby red flags for their viability as everyday players. Jackson Holliday’s +9% figure is also alarming, though it is perhaps not as concerning given his youth. One more notable point is Pete Crow-Armstrong’s 28% expected whiff rate, shockingly high for a AAA batter. This is due to his high chase rate, which stands at 43%.

Complete 2024 AAA whiff rate over expected data can be found here.

On the surface, contact ability does not appear to be a key driver of success at the MLB level. In truth, batters with subpar contact skills rarely have lasting careers in the major leagues and those that do tend to possess other elite talents. With this in mind, whiff rate over expected serves as a useful benchmark of translatable contact ability for minor league prospects.