Desert Island Hitter Metrics: Barrels, Walks, and Strikeouts

2 years ago by Jordan Rosenblum

Link copied to clipboard!

Barrels, walks, and strikeouts are three important indicators of hitting talent.

Previous research by the Baseball Data Machine Learning team has found that a simple metric that sums barrels and walks and subtracts strikeouts, and then divides by plate appearances, offers a good starting point for evaluating a hitter’s talent.

This article builds on that idea, looking at barrels, walks, and strikeouts to consider hitter outlooks for the rest of 2022.

Be sure to check all of our fantasy baseball lineup tools and weekly lineup resources:

Establishing an improved metric for barrels, strikeouts, walks

To start, I pulled data from Baseball Savant from 2015 to 2022, looking at season pairs where a hitter received a minimum of 90 plate appearances in adjacent seasons. This left me with a sample of 2,342 hitters. The correlation between barrels plus walks minus strikeouts per plate appearance in season one and wOBA in season two is .29 for this sample (note: wOBA is an all-encompassing metric of hitter talent, so new metrics of hitter talent might reasonably hope to have a strong correlation with wOBA).

For reference, the correlation between wOBA in season one and wOBA in season two is .38 for this sample (generally, metrics can be expected to have a strong correlation to themselves in the future relative to other metrics). Barrel plus walks minus strikeouts is a solid proxy of hitter talent then, but significantly lagging wOBA.

If, instead of simply summing walks and barrels and subtracting strikeouts, I assign weights to walks, barrels, and strikeouts generated from regressing season two wOBA on the three components, the correlation between the barrels-walks-strikeouts metric (henceforth BWS, because everyone loves a convoluted acronym, and I don't believe BWS has any pre-established negative connotations) and wOBA improves significantly. The optimal weights to maximize the predictive power of BWS in this sample are -1 for strikeouts, 1.3 for walks, and 4.3 for barrels. These weights improve the correlation between season one BWS and season two wOBA to .42 across the same 2,342 hitter sample, surpassing the strength between season one wOBA and season two wOBA.

Best Hitter Seasons by BWS Since 2015

The downside of adding these weights is that they are not easily computable in one’s head, and they are on an unfamiliar scale (league average is 11 in 2022, for instance). The table below shows the best hitters since 2015 by BWS (minimum 190 PA), but I convert them to a z-score (also known as standard score) for ease of interpretation (0 is league average, one is one standard deviation above league average, -1 is one standard deviation below league average, etc.).

Note the correlation between zBWS (z-score of BWS) and wOBA in the same season is .75. The two metrics of hitter talent are highly correlated, but each still offers unique insight.

Table note: all data for this article was collected on 8/15/2022.

Aaron Judge’s historic season rates second only to Juan Soto’s marvelous 2020. The rest of the list is comprised mostly of the game’s best hitters—what you want to see when designing a novel metric that purports to measure hitter talent—with each player also rated extremely well by wOBA – with the exception of Josh Donaldson in 2021, who was more good than great by wOBA, but excellent by zBWS. Donaldson also had an elite xwOBA in 2021, but he suffered from a low BABIP, and BWS does not account for close to everything. For instance, speed is not factored in, which likely played a role in the low BABIP for Donaldson (as opposed to solely bad luck).

Interesting BWS Cases in 2022

Next, the table below shows zBWS for interesting names in 2022. For inclusion, there needed to be: 1) a standard deviation difference of at least one in magnitude between zBWS and zwOBA (wOBA also converted to a z-score for ease of comparison); 2) meet my completely arbitrary and hastily determined definition of ‘interesting.’ Note that a one standard deviation shift in zBWS is equivalent to a 28-point shift in wOBA.

In 2022, league average barrel per pa is 5.2%, K% is 22.2%, and BB% is 8.2%. The table is sorted by magnitude of difference between zwOBA and zBWS.

Danny Jansen has cemented himself as a strong all-around catcher who’d see a boost in value if he left Toronto, where he’d get more playing time not having to compete with rising star Alejandro Kirk.

Abraham Toro, Austin Meadows, and Jesse Winker have each had down seasons in 2022, but Toro has maintained good BB% and K%, and Meadows and Winker have maintained elite BB% and K%, while all three have barrel per PA rates close to the MLB average. Winker and Meadows are two of my favorite buy low candidates for 2022 and beyond—particularly Meadows because of how quiet his 2022 has been on the surface (although Meadows still can’t manage to get healthy so buying low may need to wait until 2023).

Slowpoke Carlos Santana looks like a chronic zBWS underperformer and his playing time has become too inconsistent to roster in typical weekly formats, but I’m convinced he’s better than he’s been, and still an above-average hitter when he plays. Another old, slow bat, Nelson Cruz continues to play everyday and showcase solid BWS abilities. I am fine with continuing to run him out there as long as the playing time is there—he may yet have one last hot streak left in him.

Jose Miranda, Michael Harris II, Andres Gimenez, Andrew Vaughn, and Jake McCarthy comprise a group of young 2022 breakouts, to differing degrees. However, zBWS suggests a bit of caution is justified in evaluating each, as only Vaughn has a strong BB% minus K%, while only Harris has a barrel per PA rate significantly above average (7.7%). It is not surprising to see the weak BB% and K% for Harris in particular, as he went straight from Double-A to the majors. Aging growth should also be expected for this group in the longer term, however—they should look better by zBWS in 2023 and beyond. Further, Gimenez, McCarthy, and Harris have serious 2022 value even without super strong hitting skills because of the speed and playing time. I also think Vaughn has more barrel potential in the tank given his prospect pedigree.

Luis Arraez and Steven Kwan are two of baseball’s most extreme contact hitters. Arraez has grown into enough power to be appreciated by both zBWS and zwOBA, but Kwan needs to add a bit of oomph if he is to rate well by zBWS. The Big Three projection systems at FanGraphs appreciate Kwan and Arraez more than zBWS, tending to value both as comfortably above-average bats. zBWS may place a bit too much emphasis on barrels relative to strikeouts and walks and other indicators of contact quality—a future version might incorporate additional quality of contact measures beyond barrels to account for this potential bias.

Randy Arozarena also looks underrated by zBWS, as he possesses rare strong BABIP abilities, and the projections also tend to like him more than zBWS. Still, he’s a better fantasy bat than real life bat given the serious speed. The merely average-ish zwOBA and zBWS gives him some platoon risk, although he’s continued to prove that risk unfounded thus far. With elite K% and strong BB% and barrel per PA rates, Vinnie Pasquantino fully looks the part of the game's next slugging first basemen.

Conclusion

A focus on just three metrics, barrels, walks, and strikeouts, cuts through the noise to get at a hitter’s true talent level. Careful thinking and more in-depth research is required to account for the various aspects of hitting BWS does not account for.

Download Our Free News & Alerts Mobile App

Like what you see? Download our updated fantasy baseball app for iPhone and Android with 24x7 player news, injury alerts, sleepers, prospects & more. All free!