Hitter Strikeout Rate Regressers
Yesterday, I shared a list of the 11 hitters who have enjoyed the most sizable improvement in strikeout rate versus last season. If sustained, there’s a good chance that each of the hitters produce a sweet profit for their owners. On the other side of the coin are the hitters whose strikeout rates have increased the most. This is the danger zone. Let’s dive in.
| Name | 2016 K% | 2015 K% | Diff |
| Justin Upton | 41.1% | 25.6% | 15.5% |
| Leonys Martin | 36.4% | 22.3% | 14.1% |
| Brad Miller | 32.1% | 20.3% | 11.8% |
| Mitch Moreland | 33.3% | 21.7% | 11.6% |
| Yoenis Cespedes | 31.8% | 20.9% | 10.9% |
| Brian McCann | 28.8% | 18.1% | 10.7% |
| Alex Gordon | 32.4% | 21.8% | 10.6% |
| Edwin Encarnacion | 26.2% | 15.7% | 10.5% |
| Lorenzo Cain | 25.3% | 16.2% | 9.1% |
| Jason Kipnis | 25.7% | 16.7% | 9.0% |
| Carlos Beltran | 25.0% | 16.0% | 9.0% |
It’s going to be hard for Justin Upton to lead the league in runs scored, as I boldly predicted, while striking out 40% of the time. And to add to the misery, his walk rate has tumbled to a pathetic 4.1%, less than half his previous career low. Hey, at least he’s hit line drives at a 30% clip and has yet to pop up! His SwStk% is up, but it’s not absurd and certainly doesn’t match with the strikeout rate. Obviously, he’ll get better, but this is worrisome, to say the least. And that he hasn’t even attempted a steal yet, which fueled a chunk of his value, is a concern.
Leonys Martin is already 60% of the way to his 2015 home run total and nearly half his 2014 mark. Is he selling out for power? The stats certainly suggest so — his Hard%, FB%, SwStk%, and K% have all surged. I have no idea if Martin 2.0 is here to stay or if this is just a one month blip. It will be interesting to see how he continues from here though.
Brad Miller is doing his best to once again fail to break out, ensuring he makes next year’s sleeper lists…again. We knew the move to Tampa Bay wouldn’t exactly be a boon for his fantasy value, but he owns some decent skills that with just marginal improvement could have resulted in a nice fantasy season. Somehow, his strikeout rate is through the roof, despite a SwStk% that’s actually down from last year and a Swing% that’s up. Plus, his Contact% is actually above his career mark. This slow start in the strikeout department looks to be a fluke. I would suggest him as a decent trade target in AL-Only leagues, but I feel like he doesn’t possess a whole lot of job security. So if his struggles continue, he may be out of a starting job.
Most of Mitch Moreland’s underlying skills are normal, aside from the fact that he has suddenly refuses to go the opposite way, instead opting to pull everything. But he’s swinging at pitches outside the zone far more, and making contact less. He’s even whiffing inside the zone more frequently, as his overall SwStk% is up significantly. For the sake of my sad Tout Wars team, my fingers are crossed that he’ll snap out of this swing and miss funk soon.
Yoenis Cespedes is hitting the ball harder than ever before (career best Hard%), while also hitting a ton of fly balls. Paired with the high strikeout rate, it seems he’s doing his darnedest to ensure he exceeds 30 homers again. Since his walk rate sits at a career high, this version of himself isn’t so terrible. But that batted profile in no way supports a .364 BABIP.
What happens when a slow-footed catcher with a career .279 BABIP suddenly sees his strikeout rate shoot beyond the 20% mark? Bad things. You wouldn’t know it thanks to an inflated .367 BABIP at the moment, but if Brian McCann doesn’t get that strikeout rate back down, there’s a real risk he delivers a .200 batting average. Luckily, his SwStk% has remained excellent and he has stopped swinging at pitches outside the zone. His strikeout rate should rebound substantially over the rest of the season.
With just mediocre power and speed that has dried up, Alex Gordon has prided himself on being neutral on your batting average (yes, I asked him, and he’s very proud to have neither helped, nor hurt his fantasy owners in the category over the years). Unfortunately, he’s swinging and missing more than ever before, which puts him at risk of being worthless in shallow mixed leagues if he continues these shenanigans.
What had made Edwin Encarnacion such a fantastic hitter is that he has paired his great power with an excellent strikeout rate. But that hasn’t been the case so far this season. Besides seeing his SwStk% jump above 10% for the first time since his 2005 debut, he has also stopped taking the free pass, thanks to a sharp increase in Swing%. It’s a bit of a concern for a 33 year old, as you never know when old age is going to suddenly take its toll.
Lorenzo Cain enjoyed a breakout last year and there wasn’t anything overly fluky about his performance. This year, his power has come crashing back down and he has flashed a new set of plate discipline skills — his walk rate has doubled, while his strikeout rate has spiked. This would be interesting if it came along with another surge in power, but it hasn’t. So I’m not sure what’s going on here with his approach.
A .378 BABIP has masked Jason Kipnis’s sudden issues with the strikeout, but his SwStk% remains well below the league average, a good sign. However, he is swinging less than ever before, which has certainly been one of the drivers of the increased strikeout rate. Strangely, the walk rate hasn’t followed. I’m not concerned much here.
Speaking of old age, I figured it would be worth ending at 39 year old Carlos Beltran. While his power has been there so far, his plate discipline has not. His SwStk% sits at a career high and his walk rate is a puny 2.9%. Since his Swing% is consistent with past years, I’m not sure what’s fueling to disappearance of his walks. But at this age, any signs that he’s done have to be taken seriously.
Mike Podhorzer is the founder of ProjectingX IQ, an advanced fantasy baseball analytics platform that transforms projection data and in-season performance signals into actionable intelligence. He is the 2015 Fantasy Sports Writers Association Baseball Writer of the Year and three-time Tout Wars champion. He is the author of the eBook Projecting X 2.0: How to Forecast Baseball Player Performance, which teaches you how to project players yourself. Follow Mike on X@MikePodhorzer and contact him via email.
Not quite as dramatic but Mookie Betts at 20.9 vs 12.5 has been very surprising so far. His K/BB is Adam Jonesian at the moment. Hopefully just an adjustment phase to the league working him away as much as possible.
I’ve been tracking Mookie closely. After two weeks he was sitting at a 27.5% K rate, a shocking number for a guy who’s normally in the low teens. Since then he’s steadily declined and is now about to go under 20% again. I would expect he finishes in the mid- to low-teens after all is said and done, even though his swing profile supports a higher K rate so far this year.
Probably worth noting that EE had a 20%+ K% in April and May last year as well, before it came back down
A dissatisfied Russell Martin owner says:
Russell Martin: 20.9% in 2015 to 42.4% in 2016 = +21.5%
Woah, he must have missed the qualified cutoff. Had no idea, that’s crazy!
Crazy indeed. And to think, I traded Logan Forsythe (AL-only, hybrid categories) to get him 🙁
Always good when you have Martin, J. Upton, Cespedes and Kipnis on your team! Plus Sano. I feel like my whole team is a strikeout machine.
Watching Kipnis this year, I’ve noticed that he’s had trouble laying off the high fastball out of the zone when he gets down two strikes in the count. Also, a lot of the Indians have been watching pitches go right down the pipe for strikes rather than swinging at them.
I hate to be “that guy”, but I’m afraid your samples are just too darn small.
They are small, for sure, but strikeout rate “becomes reliable” after 60 PAs and shouldn’t be ignored at this point.
Encarnacion is pretty interesting. He had a strikeout rate of 21% last year in April. The 25% has me a tad uneasy. I think he figures it out but I could see him finish with his highest strikeout rate as a member of the Blue Jays
Agree I think it’s rust, he had a late start to the season due to his spring injury and it always feels like he takes some time to warm up.
Did DWright not make this cut because he’s no longer fantasy relevant (sadly)? 35.6 k% almost double his career avg.
I’m confused, in my doubles-only league Justin Upton has been very good! Say it ain’t so!
Miller’s hard% is up and his FB% is up at least over last year.
I thought I would explore if the K% change was predictive for a different sample. One, because if it was, it would be very useful. Two, because the sample seems too small.
April 2015 Strikeout Rate Regressors
2014: 18.2%
April 2015: 27.0%
Rest of 2015: 16.7%
2014: 127
April 2015: 102
Rest of 2015: 128
So for my particular sample, overall it did not yield any significant results. Besides Adam LaRoche (-7%) and Brian McCann (-4%), the other 8 players came within -1% of their 2014 K rates (or improved).
April 2015 K Rate Improvers
K%
2014: 22.3%
April 2015: 14.2%
Rest of 2015: 18.2%
wRC+
2014: 121
April 2015: 133
Rest of 2015: 128
It seems for this particular sample, overall there is a nontrivial improvement in K% and wRC+.
This is just one sample, it may be that April K% changes are more predictive than shown here. But I think that because these players are the extremes, the K% is inherently going to be less reliable and thus require more data.
Cool, thanks for doing this. I went back myself and just eye balled some players. It would be time consuming to go back multiple years and do the work, so I haven’t done that. But it would certainly be interesting.
Stabilization concept in general needs to be reviewed. I am inclined to think it is being misused. I looked at a list from May 2013 (when the claim was that this data stabilized at only 100PA rather than todays 65) and the rest of year results did not correlate well with the early results.
It may be a problem of expectations. It is one thing to say that after a sample of 65 for a highly correlated variable we have moderate level of confidence that the sample reflects a range around the true level. However, since most of the changes fall within or close to the expected range it is difficult to assign much significance. To expect variations of only 10% to be meaningful at such small sample sizes may not be reasonalbe and this practice should certainly be reviewed for past reliability. This link contains another link to a PDF which has a table showing sample sizes needed for different r levels, confidence levels, and stabilization levels. http://www.r-bloggers.com/at-what-sample-size-do-correlations-stabilize/