The Starting Pitcher Strikeout Rate Surge Candidates

As we near the end of the month, our samples remain far too small to perform any serious analysis on. But, that’s really only true when we’re using plate appearances or innings pitched as our denominator. My xK% equation is based on per-pitch metrics, which stabilize much more quickly than anything based on innings. No, I don’t know the actual stabilization point, but since a pitcher has thrown more pitches than he has innings pitched, that’s what’s going to happen.

So let’s take a look at those starting pitchers whose xK% marks are most above their actual strikeout rates. These are the guys with significant upside who should enjoy a strikeout rate surge in the near future.

Name xK% K% Diff
Chris Sale 27.1% 21.6% -5.5%
Julio Teheran 24.4% 19.4% -5.0%
Archie Bradley 21.3% 16.9% -4.4%
Nick Martinez 15.8% 11.4% -4.4%
Mike Pelfrey 15.5% 11.1% -4.4%
Tim Hudson 16.3% 12.2% -4.1%
David Buchanan 15.5% 11.5% -4.0%
Gio Gonzalez 20.9% 17.3% -3.6%
Daniel Norris 22.7% 19.3% -3.4%
Shane Greene 14.9% 11.8% -3.1%
Jeff Samardzija 17.8% 14.8% -3.0%

Chris Sale got a late start as he was recovering from a foot injury, but he has still thrown 19.0 innings so far. His current strikeout rate would easily be a career low, so it might be tempting to believe that maybe his foot is affecting him. But according to xK%, he has been the unluckiest pitcher in baseball, as his xK% mark sits at exactly his career average. Nothing to see here, everything looks normal and Sale’s strikeouts will return sooner rather than later.

Given his two seasons of SIERA outperformance, I have felt that Julio Teheran has been massively overvalued. But the velocity decline he suffered last year just raised another red flag. His xK% suggests that his strikeout rate shouldn’t continue tumbling, but there’s still other things to worry about, namely his control. This is not the post to dive into all his issues, but his total strike percentage is down significantly from last year and he has thrown first pitch strikes less than 50% of the time, after posting marks above 60% in his first two full seasons. Since his value was already inflated in my eyes, I don’t consider him a buy low candidate right now.

Archie Bradley received tons of hype despite terrible control and solid, albeit unspectacular strikeout rates in the minors. In his first three Major League starts, his results look shiny, but he has actually pitched rather poorly, striking out just one more batter than he has walked. His minor league track record tells us that of course his strikeout rate should improve and his xK% agrees, but also believes he has been a bit unlucky in that department. However, there’s a caveat here — his xK% is driven largely by a huge looking strike rate, a mark that ranks second in all of baseball. It’s simply not a sustainable mark. I’m quite bearish on Bradley, but given his hype and sub-2.00 ERA to begin his Major League career, he looks like a prime sell high guy.

So maybe Nick Martinez shouldn’t be absolutely terrible, just bad?

Same goes for Mike Pelfrey!

Tim Hudson is still pitching? His current strikeout rate sits at what would be a career low. At age 39, you never know when the gas tank is fully emptied, but xK% tells us the strikeout rate should sit right in line with his career mark.

If David Buchanan’s strikeout rate actually did approach his xK%, he would have struck out more than he has walked, which is not something he is currently doing! He has a weak track record of strikeouts in the minors, so there’s little upside here. I’d rather own any middle reliever than Buchanan, even in an NL-Only league.

Gio Gonzalez pitched yesterday and struck out 32% of opposing batters, so his xK%-K% differential has likely narrowed. With a 3.44 xFIP now, there’s nothing to worry about. That .394 BABIP is going to come crashing down soon enough.

I think Daniel Norris is my new favorite player after hearing his story, seemingly told every time he pitches, about him living in a van. He seems like a cool, down to earth guy. I like that. Anyhow, his season has started off just as expected, with major problems throwing strikes, but the ability to strike hitters out. That latter ability hasn’t been as strong as his minor league record has suggested though, which has led to an uninspiring performance so far. Most concerning is his 53.4% fly ball rate. The AL East and a home run friendly home park aren’t the places to allow such a high fly ball rate.

Sometimes regression to the mean can happen in an instant. Shane Greene found that out, as he got bombed for eight runs in just four innings literally the same night I suggested selling him high that morning. In that post, I said that his strikeout rate should definitely rise and xK% agrees. But even a 14.9% strikeout rate is a far cry that we all expected. Even Steamer, which was much more pessimistic than I, and likely many fantasy owners, forecasted a mid-high teens strikeout rate. He’s obviously not a buy low for the strikeout rate surge and I’m guessing the window to sell high just got smaller. Maybe he has another good start to erase the memory of the previous one, which opens that window back up.

We all figured that Jeff Samardzija would be hurt by his move to the South side of Chicago. After four starts, his ERA sits at 3.33, which was likely better than most expected. So all’s good, right? Nope. His strikeouts have vanished. His 17.8% xK% suggests better ahead, but even that mark is significantly below what we have been accustomed to from him. All of the components of his xK% are in line with his history, except for his swinging strikes, the most important one. The problem hasn’t come on just one pitch. In fact, every one of his non-fastballs have experienced a drop in SwStk%. His fastball velocity is down just a tick, but it’s normal this early in the season and certainly nothing to worry about. I’m not sure why all his pitches have lost their swing and miss effectiveness, but I would be concerned if I were a Samardzija owner.





Mike Podhorzer is the 2015 Fantasy Sports Writers Association Baseball Writer of the Year. He produces player projections using his own forecasting system and is the author of the eBook Projecting X 2.0: How to Forecast Baseball Player Performance, which teaches you how to project players yourself. His projections helped him win the inaugural 2013 Tout Wars mixed draft league. Follow Mike on Twitter @MikePodhorzer and contact him via email.

21 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
CDMcLean
8 years ago

FYI: The Nick Martinez link is a Johan Santa.

Good article, though. Is the difference between xK% and K% usually due to sequencing or do some pitchers have problems in 2-strike counts that prevent them from performing near their xK% (or vice versa for overperformers)?

novaether
8 years ago
Reply to  Mike Podhorzer

Ideally, you would probably to split out the swinging strike and looking strike rates in 2-strike counts. It’d probably be easier to have an indicator in the regression for a “put-away” pitch (or lack thereof). It would at least explain why a guy like Pelfrey shows up on this list.

Tom
8 years ago
Reply to  Mike Podhorzer

Not on your list, but with regard to lack of putaway pitch, Schilling went on for a while in last nights game about Eovaldi’s lack of a putaway pitch. Where does Eovaldi rank?

novaether
8 years ago
Reply to  Mike Podhorzer

No Mike, that’s precisely why it would work. Backing into an equation that calculates strikeouts in a round-about way is going to have a higher R^2 with strikeouts than the current xK% equation.

All that really does is illuminate the fact that a high R^2 in a regression doesn’t mean that the relationship is causal or repeatable. Leaving a pitch sequencing variable off the regression leads to omitted variable bias, which isn’t the end of the world, but it means that the goodness of fit overstates the predictive power of the regression.

Even more, though, the entire predictive power of the regression is predicated on the fact that the components are supposed to stabilize quickly. However, strikeout rate itself stabilized quickly. I did a quick regression of 2013 xK% on 2014 actual K% (50 IP min both years) and got an R^2 = 0.50. I regressed 2013 actual K% against 2014 actual K% and got an R^2 = 0.52.

Based on that result, it seems clear that we shouldn’t discount pitch sequencing or other effects when discussing xK% in the predictive sense. It might be overkill when discussing if a pitcher “earned” his past results or not, but it would technically improve the discussion.