Pitching by the Numbers: Riding the hot hand

We’re in the home stretch and need to do something that seems on the surface very unsabermetric-like: look at the hot pitchers.

Big numbers beating small numbers is a general rule. And chopping up samples within a season tends to be ill-advised. However, I do believe in the hot pitcher given that there are so many little things that, when set right, can generate big results. So the guys who have found their stride since the All-Star break only need to keep that muscle memory going for us for five or six more weeks.

But I will focus more on the best of our foundational stats, (Strikeouts-Walks) divided by Innings Pitched. You will see in the list below that our category stats (ERA and WHIP) track very neatly with it and Ks are a part of it. Wins? I don’t bother thinking about them and you should not either. We want ERA and WHIP to track closely with our foundational stat because then we can more reasonable speculate cheaply on the pitcher who is good in the key stat but unlucky in run prevention. Alternatively, we can also find guys who may be freely available because their full-season stats are masking their second-half turnaround. The minimums here are 35 innings pitched since the All-Star break with five starts, giving us 84 qualifiers. Here are the top 40 in (K-BB)/IP:

The most gettable guys here are McCarthy (44% Yahoo! owned and pitching like an ace in both reality and fantasy), Colon (43%), Smyly (32%) and Norris (14%).

And that’s exactly the order they would come off the board right now for me.

Now I see why my Julio Teheran hate mail has died down. Ironically, since the break he’s actually underachieving his K-BB dominance in ERA. My point has always been that I do not see him as an upper-tier pitcher. That doesn’t mean he’s some bum or even that I dislike him. It just means to trade him if you can find someone who does think he’s all that.

Fister is a very strange pitcher. I’m riding him to Scoresheet glory, where his strikeouts don’t even matter. But his averages all year are just sick, making his low strikeout rate a non-issue in even shallow formats. But are they bettable going forward? Of course not, not at those levels. No one is bettable at them. Fister though just has something that defies all sabermetric reason. What can it be? Oh, his isolated-slugging is always great — .117 for his career and, many sabermetric guys will maintain, a totally random and unprojectable .113 this year. Shocker. Seriously, when are the database people going to just give us this stat instead of making us hunt for it in the deep weeds like it’s some dark secret?