SOBB – My favorite metric! Why? Maybe because I came up with it. In all seriousness, even FanGraphs has seen its value and added it to their charts in its simple, no-fun, non-catchy nickname form, K-BB%. At least they saved us the trouble of manually calculating it. Instead of K/9 and BB/9, which are valuable tools in themselves, K% and BB% give us a slightly better picture because it accounts for number of batters seen (it uses plate appearances, not at-bats). If a pitcher sees more batters in an inning as a result of errors or his own lack of control (walks), etc., he has more chances to make his K/9 look better.
It rarely happens, but just think about the four-strikeout inning. The K/9 for that one inning pitched is 36.0, which is obviously a ridiculously high number. It doesn’t matter if the pitcher allowed four hits in the inning and walked two others, the K/9 would still be 36.0. However, K% adjusts for the number of batters faced, which is much more telling and significant. Let’s go back to the four-strikeout inning example, as rare as it is. If a pitcher struck out all four batters without allowing anyone else on base, his strikeout percentage would be 100. However, using the same numbers above with four hits and two walks, we now have 10 hitters with a plate appearance, changing the calculation from 4/4 = 100% to 4/10 = 40%.
Here’s a full-season example. A pitcher with 200 innings pitched and 200 strikeouts would have a 9.0 K/9. That’s a solid number. That same pitcher could have also allowed 250 hits and walked 100 batters, which is extremely poor, and still have that respectable 9.0 K/9. However, using K% would give us 950 batters faced.
((200 innings*3) +250 + 100) = 950
The K% would be 21.1 (200 divided by 950). If that pitcher had better control with half as many walks, and he allowed fewer hits (we’ll say 75 less), we’d have a much better K% of 24.2.
(200/(200*3) + 175 + 50) = 24.2
Enough math for you? Basically, K% is significantly more telling of a pitcher’s success and ability to strike out batters, and the same goes for BB% and avoiding walks.
The next step is taking K% and subtracting BB%, which gives us SOBB, or StrikeOut percentage minus Base on Balls percentage. SOBB gives a nice all-around look at a pitcher’s ability to get batters out simply based on his ability. An average K% is around 18-19 percent and BB% around 7-8. Therefore, an average SOBB will sit right around 11-12 percent.
As mentioned, SOBB gives us tremendous insight to a pitcher’s true ability to get batters out using only his skill, and that’s why it’s also one of the best predictive stats available. BABIP, Hard%, etc. are all great statistics to use, but none is as valuable as SOBB (or K-BB% if you insist). Using SOBB isn’t foolproof, but it is the most accurate stat you can use to predict future performance for pitchers and/or evaluate if pitchers are “for real.” Don’t head into your next Fantasy Baseball season without this tool in your bag.
Below is a table to show you the range of starting pitcher SOBBs (K-BB%) and the values they hold.
There is a second table for SOBB ranges. On average, relievers have higher SOBBs than starters do. It makes sense when you think about it, as relievers can go “all out” due to their brief outings. If we ranked all pitchers each year in SOBB, we would find one, maybe two starters inside the Top 10. As a result, we have to value relievers differently with higher thresholds for each tier.
Below is a table to show you the range of relief pitcher SOBBs (K-BB%) and the values they hold.