Introducing a Simple New Plate-Discipline Metric
Not all walks are created equal. This is codified most directly in the concept of the intentional walk, which became such standard and tedious practice that it can now be accomplished with a simple signal. While some intentional walks are strictly based on setting up a better opportunity (say, it allows for a double play, more convenient force outs, or the pitcher is batting next), they often indicate that simply allowing the hitter to take first is a safer alternative than giving them the chance of doing anything else.
More generally, while walks are good for the batter, they're worse than getting a hit. On the flip side, while walks are bad for the pitcher, they're better than giving up a hit. To put that another way, getting a walk instead of getting a hit is a missed opportunity for the batter and a minor victory for the pitcher, but getting a walk instead of getting an out is a minor victory for the batter and a missed opportunity for the pitcher. This is pretty straightforward, but the distinction isn't really captured in most plate-discipline statistics. On one hand, something simple like walk percentage very intentionally sets aside the question of what else might have happened in the plate appearance, or if those walks are coming at the expense of hits, given the uncertainty of a batted ball becoming a hit. On the other hand, swing decision data combined with Statcast batted-ball data could provide a pretty good idea of what the alternatives could have been for a given walk - potentially even to the level of whether there was a hittable pitch earlier in the plate appearance and what the batter could have been expected to do with it - but would require some fairly detailed analysis to determine how beneficial the walk was compared to what could have been. Is there a middle ground there?
One approach is to reframe the issue slightly. If a batter gets a hit instead of walking, that shouldn't reflect negatively on their plate discipline. If a batter gets a hit instead of getting out, that doesn't directly relate to plate discipline either. With that in mind, you can focus directly on the ratio between how often a batter walks and how often they get out. While there are various ways to represent the resulting statistic, the simplest is to just divide walks by outs for a direct Walks Over Outs value. Since most data sources don't have 'outs' as a direct batting statistic, the formula would be as follows:
WOO = BB/(1-BA)*AB
If that feels too results-oriented, you could substitute xBA for BA (this would probably be the more useful version, but the end results are generally fairly similar, so I'll stick with the simpler version here). Either way, leaders in WOO generally line up pretty closely with leaders in BB%, but good hitters aren't 'punished' for getting hits instead of walks, and those anomalies who walk a lot but do relatively little else (sorry, Vogelbach) are held in less high regard. For reference, here are the top 30 seasons by WOO over the past five full seasons, along with the BB% for those players.
I also thought it would be interesting, given the large number of high-leverage walks in yesterday's game, to look at a quick snapshot of this metric on a team level for the entertainingly small sample size of one game.
Which I think confirms what we were all suspecting: the Mariners are currently among the league leaders in WOO.