Defensive metrics can be extremely confusing and sometimes unintuitive. You don’t necessarily need to proceed with caution when using them in analysis, rather it’s a good idea to also watch video on fielders to get a grasp of a player’s routes, reads, footwork, and more. Luckily, Statcast’s breakdown of their defensive metrics, Outs Above Average (OAA), is helpful in doing this, making it my personal go-to. It usually better aligns with the eye test than say, UZR/DEF.
With that in mind, my go to WAR methodology for position players is indeed fWAR. Because OAA and UZR/Def can vary from time to time depending on the player, it’s important to have an appropriate level of skepticism on a player’s fWAR if a chunk of their value is skewed by defensive metrics. All that said, I thought it was a good time to compare how and if those discrepancies were apparent for some Yankee players this season and if their total value may have been higher or lower.
Joey Gallo is the first player that comes to mind when thinking about these defensive tidbits. He’s hurt by the positional adjustment for left fielders when it comes to Def and ends up with -6.9 defensive runs. It brings his fWAR down to 3.5. Alternatively, Statcast pegs him in the 92nd percentile for OAA. In other words, he is actually an elite defender and not a below average one.
Something like that could significantly skew how we perceive his season. Don’t get me wrong, ~1 WAR is within the margin of error, but if I told you Gallo was a 4.5-win player and still has so much left to tap into with the bat, what would you say? It’s an interesting thought exercise.
Aaron Judge’s case is slightly different to Gallo’s, and probably validates our argument that Gallo was more valuable than just 3.5 WAR. Judge graded at -4.2 runs below average, just slightly better than Gallo. Def hasn’t ever really loved Judge like OAA once did in his first few seasons. With Judge slowing down in terms of sprint speed and admitting he is now playing a more passive outfield, it makes sense to see his defensive metrics getting worse, as Def and OAA (40th percentile) both indicate.
However, Def says Judge was a more valuable defender than Gallo, while Statcast has Gallo grading out loads better. The logic definitely tracks for Judge, but for Gallo, I’m still sold on him being an elite defender as OAA says. In reality, Gallo is most definitely the better defender, but still trails behind Judge in total value because of their offensive gap.
The last player I want to look at is Gio Urshela. Yankees fans have this understanding that Urshela is a great defender. We’ve seen him flash the leather countless times, making Machado-esque plays. Yet, these defensive metrics are not at all a fan of him, and they never have been. He obviously has great hands. They’re smooth like butter and quick as one would want. However, his feet aren’t very quick despite the very sound fundamentals.
This season, with time spent at third base and shortstop, he accumulated -4 OAA but had 3.0 Def. The positional adjustment certainly helps him. It’s very tough to question that he has superb true defensive talent based on watching him make difficult play after difficult play. Despite that, it’s also tough to ignore his woeful OAA year in, year out. This is why defensive metrics are so confusing no matter which one you prefer. This type of disagreement is significant enough to make Urshela a replacement player, or not if he is going to be an average hitter.
Our public understanding of defensive talent has improved over time, but there is probably still a long way to go when it comes to reaching a consensus methodology that is close to flawless. Let me put it like this — if most of a player’s value comes from offensive output, it’s hard to make a case against them as an elite player. If a player with a similar WAR gets most of their output from defensive value, then I’d be much more inclined to trust that the player with the better offense is truly the more valuable player. I know that has holes of its own, but as long as these metrics are still separated as much as they can be, I’ll continue to rely more on offensive output than defensive when I’m evaluating a player’s entire package.