Strikeouts, Stabilization and Surprising Swings

You’ll hear us talk about statistical stabilization here, and link to pieces like Russell Carleton’s or Derek Carty’s. The basic idea is that there are thresholds at which a stat moves into a decent sample and becomes more meaningful.

Maybe you’ll come away thinking that we’ve said that ‘x stat is stable so that’s what you’ll get the rest of the way,’ and if so, that’s on us. That’s not what stabilization means in this context.

What it means is that the r-squared number that correlates a player’s past stats with his future stats in that category has passed .50. That’s a mouthful, here’s another try: if you were to try and predict future work in a category, you’d regress their current work against the league average. At the stabilization point, you can use half their own number plus half the league average in your calculations.

One more try, in the most colloquial language possible: Stabilization is the point at which a number in a category tells us more about their future work than the league average.

If that gets boiled down to more simple language, you can understand why. None of the above makes for hit monster posts.

But, there’s a little more to it: we play fantasy baseball. We don’t have the same stakes as a major league team, and if we’re in a redraft league, we have to move faster than the numbers, in effect. For example, power doesn’t stabilize until 550 plate appearances or so. There’s no way we can hold on to someone experiencing a power blackout for a full season in your standard 15-team redraft.

And so we simplify our language, at the cost of precision. “Time to worry about this player’s production because of this stat has stabilized” is something I’ll say. I understand it’s more complicated than that. I understand that shivers will go down spines. But I have moves to make on my fantasy teams, and I need to grasp at straws sometimes.

All of that is pre-amble. According to Derek Carty’s excellent piece linked above, strikeout rate for batters has stabilized for most regulars. Because we’re shaving the garlic here anyway, I’ve set the minimum at 90 plate appearances instead of the 100 he found. And I’ve set their current strikeout rate against last year’s, so that you can sort the table in different directions.

I’ll post the table at the bottom, after we point out some interesting movers and shakers. Strikeouts are negatively correlated to batting average, so that’s why you care about this. Even though most researchers and baseball people don’t care about batting average and might cringe at using stabilization this way, we have categories to protect. We must move quickly.

I’m picking up what Miguel Montero is laying down. Swing metrics are ahead of per-PA metrics, and he’s swinging less than he’s ever had, he’s reaching less than he’s ever had, and he’s making more contact than he’s ever made. At this strikeout rate, there’s plenty of room for regression while still showing batting average improvement. Monter’s hit .280+ in three of his five big-sample years, and I think he’ll do that again this year.

Brandon Moss isn’t seeing the same kind of radical changes in his swing profile, but the contact is up and the strikeout rate is down. This is interesting because he’s made more contact in the past, but seemed to go with the Big Swing for the Big Power. His owners own him for the power, so I’m a little more ambivalent about this big change.

Yoenis Cespedes is swinging less, reaching less, and making more per-pitch contact, so I’m buying this new contact rate. It’s not statistically important, but it’s nice to see it come with a better walk rate and comments from the player himself about shortening his swing. Since BABIP is hiding the fact that he’s improved, he makes for a decent buy-low, actually.

I love Jason Kipnis, still do.

Pedro Alvarez! God of Whiffs! What will we do if you change so fundamentally? We will go buy you. Because you have decided to swing less, reach less, and make more contact. And you haven’t cost your power. And you’re walking more. If only your batting average on balls in play (.161) loved you as much as I do right now.

On the other side of the spectrum, we’ve talked about Brad Miller enough, and it all looks bad. Every day makes him more droppable. Edwin Encarnacion, though, is getting a pass. But maybe he shouldn’t. At least, it doesn’t look like he’ll have his customary great strikeout rate this year. But it’s a testament to how good his strikeout rate used to be that he can fall this far and still have something close to an average strikeout rate. If the wrist is fine, I’m only slightly worried, because he can still provide a ton of value even with a worse batting average.

Norichika Aoki and Garrett Jones are both older gentlemen with flawed games. At their current rates, they’re playing themselves out of mixed league relevance quickly.

The next two are more important to more of yous. Mike Trout is probably unassailable, and it’s worth noticing that most of his swing rates are fine, it’s just a little drop in contact rate. This might need more attention. But there is reason to worry about Martin Prado. His swing rates are unchanged, but his ability to make contact outside the zone — not a skill that ages well — dropped seven points. That’s how you turn a small change in contact rate into a huge change in strikeout rate. Given his career strikeout rate is 10.6%, and he’s at 17% now, even with regression, we’re likely looking at Prado’s worst year for strikeouts. And his game is built on contact.

We’ll try to look into the rest of these guys, but here’s a sortable table that you can use for yourself:

Name PA K% 2013K% 14K-13K% BABIP AVG
Miguel Montero 105 8.6% 23.2% -14.6% 0.273 0.271
Brandon Moss 96 18.8% 27.7% -8.9% 0.295 0.268
Yoenis Cespedes 95 15.8% 23.9% -8.1% 0.242 0.238
Jason Kipnis 107 14.0% 21.7% -7.7% 0.264 0.247
Pedro Alvarez 107 23.4% 30.3% -6.9% 0.161 0.172
James Loney 97 6.2% 12.9% -6.7% 0.316 0.306
Matt Adams 98 18.4% 25.1% -6.7% 0.408 0.337
Alex Gordon 101 13.9% 20.1% -6.2% 0.313 0.277
Emilio Bonifacio 99 16.2% 22.3% -6.1% 0.405 0.333
Mike Napoli 106 26.4% 32.4% -6.0% 0.390 0.304
B.J. Upton 100 28.0% 33.9% -5.9% 0.295 0.211
Carlos Gonzalez 101 21.8% 27.1% -5.3% 0.265 0.234
Starlin Castro 100 13.0% 18.3% -5.3% 0.304 0.292
Justin Morneau 97 12.4% 17.3% -4.9% 0.360 0.356
Jose Altuve 116 7.8% 12.6% -4.8% 0.309 0.292
Chase Utley 99 10.1% 14.9% -4.8% 0.377 0.360
Jhonny Peralta 93 17.2% 21.9% -4.7% 0.167 0.195
Ben Zobrist 114 8.8% 13.0% -4.2% 0.321 0.313
Albert Pujols 110 8.2% 12.4% -4.2% 0.244 0.290
Alfonso Soriano 101 20.8% 24.9% -4.1% 0.286 0.258
Aramis Ramirez 102 11.8% 15.7% -3.9% 0.288 0.280
Neil Walker 111 11.7% 15.4% -3.7% 0.217 0.235
Prince Fielder 109 12.8% 16.4% -3.6% 0.227 0.209
Troy Tulowitzki 98 13.3% 16.6% -3.3% 0.349 0.342
Asdrubal Cabrera 100 17.0% 20.3% -3.3% 0.250 0.211
Jed Lowrie 114 10.5% 13.7% -3.2% 0.316 0.292
Evan Longoria 109 20.2% 23.4% -3.2% 0.351 0.296
Michael Brantley 102 7.8% 11.0% -3.2% 0.253 0.264
Jay Bruce 103 23.3% 26.5% -3.2% 0.271 0.224
Anthony Rendon 112 14.3% 17.5% -3.2% 0.326 0.298
Adam LaRoche 104 19.2% 22.2% -3.0% 0.359 0.307
Chris Davis 94 26.6% 29.6% -3.0% 0.333 0.250
Hunter Pence 109 13.8% 16.7% -2.9% 0.282 0.253
Alejandro De Aza 90 18.9% 21.8% -2.9% 0.190 0.183
Ryan Howard 107 27.1% 30.0% -2.9% 0.295 0.245
Yasiel Puig 96 19.8% 22.5% -2.7% 0.311 0.265
Trevor Plouffe 105 19.0% 21.5% -2.5% 0.388 0.310
Freddie Freeman 107 16.8% 19.2% -2.4% 0.371 0.344
Alex Rios 107 14.0% 16.3% -2.3% 0.376 0.330
Todd Frazier 97 18.6% 20.8% -2.2% 0.266 0.244
Eric Hosmer 104 12.5% 14.7% -2.2% 0.341 0.295
Nolan Arenado 109 11.9% 14.0% -2.1% 0.303 0.298
Anthony Rizzo 104 16.3% 18.4% -2.1% 0.319 0.284
Nelson Cruz 101 21.8% 23.9% -2.1% 0.305 0.284
Nick Markakis 111 9.0% 10.9% -1.9% 0.330 0.300
Dan Uggla 93 30.1% 31.8% -1.7% 0.281 0.209
Ian Kinsler 96 8.3% 9.6% -1.3% 0.284 0.278
Allen Craig 103 16.5% 17.8% -1.3% 0.203 0.177
Elvis Andrus 110 12.7% 13.9% -1.2% 0.284 0.253
Domonic Brown 101 16.8% 18.0% -1.2% 0.311 0.264
Joey Votto 112 17.9% 19.0% -1.1% 0.328 0.287
Dustin Pedroia 115 9.6% 10.4% -0.8% 0.305 0.274
Jean Segura 94 12.8% 13.5% -0.7% 0.267 0.239
Jimmy Rollins 102 13.7% 14.0% -0.3% 0.280 0.261
Chris Carter 92 35.9% 36.2% -0.3% 0.238 0.169
Matt Holliday 107 14.0% 14.3% -0.3% 0.312 0.272
Howie Kendrick 111 17.1% 17.3% -0.2% 0.354 0.300
David Ortiz 103 14.6% 14.7% -0.1% 0.250 0.253
Yonder Alonso 97 12.4% 12.5% -0.1% 0.198 0.174
Alexei Ramirez 109 10.1% 10.1% 0.0% 0.360 0.350
Omar Infante 97 9.3% 9.2% 0.1% 0.289 0.279
Paul Goldschmidt 124 21.0% 20.4% 0.6% 0.369 0.306
Jonathan Lucroy 95 12.6% 11.9% 0.7% 0.347 0.306
Giancarlo Stanton 112 28.6% 27.8% 0.8% 0.328 0.270
Daniel Murphy 104 14.4% 13.6% 0.8% 0.337 0.289
Jayson Werth 111 19.8% 19.0% 0.8% 0.338 0.281
Yunel Escobar 97 13.4% 12.6% 0.8% 0.253 0.225
Leonys Martin 93 21.5% 20.5% 1.0% 0.400 0.309
Jacoby Ellsbury 103 15.5% 14.5% 1.0% 0.372 0.312
Gerardo Parra 118 16.1% 15.1% 1.0% 0.292 0.250
Dexter Fowler 94 22.3% 21.3% 1.0% 0.295 0.238
Jose Bautista 112 17.0% 15.9% 1.1% 0.305 0.294
Brett Gardner 91 22.0% 20.9% 1.1% 0.350 0.272
Robinson Cano 102 13.7% 12.5% 1.2% 0.342 0.301
Jason Castro 90 27.8% 26.5% 1.3% 0.265 0.221
Salvador Perez 97 13.4% 12.0% 1.4% 0.270 0.239
Erick Aybar 94 11.7% 10.0% 1.7% 0.286 0.261
Andrew McCutchen 120 16.7% 15.0% 1.7% 0.320 0.286
Alberto Callaspo 91 11.0% 9.1% 1.9% 0.279 0.272
Josh Donaldson 119 18.5% 16.5% 2.0% 0.291 0.278
Carlos Gomez 112 26.8% 24.7% 2.1% 0.373 0.297
Adeiny Hechavarria 101 18.8% 16.6% 2.2% 0.329 0.269
Wil Myers 101 26.7% 24.4% 2.3% 0.306 0.231
Juan Uribe 103 21.4% 19.0% 2.4% 0.360 0.310
Yadier Molina 95 12.6% 10.2% 2.4% 0.368 0.344
Buster Posey 91 14.3% 11.8% 2.5% 0.226 0.238
Hanley Ramirez 105 18.1% 15.5% 2.6% 0.324 0.280
Ben Revere 96 13.5% 10.7% 2.8% 0.363 0.312
Nick Swisher 113 24.8% 21.8% 3.0% 0.278 0.218
Miguel Cabrera 91 17.6% 14.4% 3.2% 0.299 0.259
Angel Pagan 100 15.0% 11.8% 3.2% 0.385 0.333
Carlos Santana 103 20.4% 17.1% 3.3% 0.150 0.122
Justin Smoak 92 26.1% 22.8% 3.3% 0.298 0.241
Billy Butler 102 18.6% 15.3% 3.3% 0.257 0.209
Carlos Beltran 98 18.4% 15.0% 3.4% 0.290 0.275
Colby Rasmus 91 33.0% 29.5% 3.5% 0.235 0.188
Adam Jones 103 23.3% 19.7% 3.6% 0.338 0.265
Melky Cabrera 117 16.2% 12.6% 3.6% 0.382 0.345
Brian Dozier 113 23.0% 19.3% 3.7% 0.217 0.217
Matt Dominguez 94 20.2% 16.3% 3.9% 0.234 0.221
Brett Lawrie 103 19.4% 15.4% 4.0% 0.157 0.179
Desmond Jennings 95 23.2% 19.1% 4.1% 0.357 0.269
Bryce Harper 91 23.1% 18.9% 4.2% 0.377 0.289
Kyle Seager 91 22.0% 17.6% 4.4% 0.241 0.228
Marlon Byrd 102 29.4% 24.9% 4.5% 0.385 0.278
Chris Johnson 92 26.1% 21.2% 4.9% 0.317 0.241
Joe Mauer 111 23.4% 17.5% 5.9% 0.358 0.266
Jedd Gyorko 101 29.7% 23.4% 6.3% 0.197 0.144
Will Venable 92 29.3% 22.9% 6.4% 0.295 0.205
Justin Upton 101 31.7% 25.0% 6.7% 0.440 0.330
Adrian Gonzalez 109 22.0% 15.3% 6.7% 0.333 0.313
Jason Heyward 107 23.4% 16.6% 6.8% 0.239 0.191
Aaron Hill 118 20.3% 13.3% 7.0% 0.310 0.255
Starling Marte 117 31.6% 24.4% 7.2% 0.343 0.229
Matt Carpenter 114 21.1% 13.7% 7.4% 0.356 0.281
Pablo Sandoval 99 21.2% 13.5% 7.7% 0.212 0.180
Eric Young 106 24.5% 16.7% 7.8% 0.302 0.216
David Wright 113 23.9% 16.1% 7.8% 0.355 0.275
Brandon Belt 104 29.8% 21.9% 7.9% 0.311 0.263
Everth Cabrera 109 23.9% 15.9% 8.0% 0.403 0.301
Ian Desmond 109 30.3% 22.1% 8.2% 0.313 0.243
Brandon Phillips 107 23.4% 14.7% 8.7% 0.338 0.262
Mike Trout 111 27.9% 19.0% 8.9% 0.413 0.320
Martin Prado 112 17.0% 8.0% 9.0% 0.284 0.234
Garrett Jones 102 32.4% 23.0% 9.4% 0.316 0.237
Norichika Aoki 102 15.7% 5.9% 9.8% 0.333 0.277
Brad Miller 90 28.9% 15.5% 13.4% 0.211 0.174
Edwin Encarnacion 107 23.4% 10.0% 13.4% 0.319 0.242



Print This Post

Graphs: Baseball, Roto, Beer, brats (OK, no graphs for that...yet), repeat. Follow him on Twitter @enosarris.

36 Responses to “Strikeouts, Stabilization and Surprising Swings”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Close to the Edge says:

    Someone offered me Cespedes for Heyward in a 14 team 5×5 OBP redraft league. Normally would dismiss it out of hand but do you see this as getting to be a closer call?

    Vote -1 Vote +1

  2. Been thinking about targeting Alvarez in an 8-team keeper, but couldn’t decide on who to offer. Seager seemed like too much. Currently have Rendon playing in the 3B slot every day. Have several OFs, should I try to sell high on Blackmon and get something else in addition to Alvarez back?

    Vote -1 Vote +1

    • I should note this is an AVG+OPS league.

      Vote -1 Vote +1

    • Travis L says:

      My opinion is worth what it costs you, but I probably wouldn’t trade Alvarez for Seager straight up (in an OBP/SLG/R/RBI/SB league).

      I’m currently trying to acquire Alvarez myself, and he was a keeper for our league (keep 8). Probably going to offer Cashner + Headley + Bonifacio.

      Vote -1 Vote +1

  3. Burton Cummings says:

    In a league that counts AVG and SLG, I should be less concerned about Adrian Gonzalez, who, I’ve seen noted, might be putting swing-for-power ahead of swing-for-contact these days, yes? I won’t pretend to know my way around these numbers, but this metric wouldn’t indicate changes in ISO, would it?

    Vote -1 Vote +1

  4. stonepie says:

    what’s the correlation between contact rate and K%? looking at a guy like de aza, his K% is down while his contact % is also down a whole 5%. shouldn’t we expect his strikeouts to shoot (way) up?

    Vote -1 Vote +1

  5. Sgt. Hulka says:

    Would anyone consider a Bryce Harper for Cespedes trade?

    Vote -1 Vote +1

    • Jeff says:

      I don’t think the guy getting Harper would take it at this point, but it’s worth a shot. In keeper leagues, definitely the Harper side (probably too obvious to say). Two months is a long time.

      Vote -1 Vote +1

      • Sgt. Hulka says:

        Yeah, I was wondering if the Cespedes owner was frustrated by the lack of steals so far and starstruck by the name. Probably a closer deal in H2H than in Roto.

        Vote -1 Vote +1

  6. JKB says:

    Moustakas seems to be improving YOY in most of his plate discipline metrics as well, and his BABIP is really low. Breakout candidate or SSDD?

    Vote -1 Vote +1

    • Emcee Peepants says:

      Does SSDD mean “Surely Sucks Donkey Dick”? If so, that one.

      Vote -1 Vote +1

    • Sgt. Hulka says:

      His batted ball profile definitely looks different this year, but it would have to be an AL only, or super deep mixed, league for me to think about buying low on him. I just don’t trust him.

      Vote -1 Vote +1

  7. Jeff says:

    Eno, does Gordon’s K% mean he went back to the line-drive swing he gave up near the end of last year? Haven’t seen any Royals games.

    Vote -1 Vote +1

  8. Jack says:

    Right rank: Beltran, Butler, Springer, Soriano, Joyce?

    Vote -1 Vote +1

  9. novaether says:

    “What it means is that the r-squared number that correlates a player’s past stats with his future stats in that category has passed .50.”

    Eno, I believe this is wrong. I’m nitpicking here, but stabilization means that more than half of a player’s performance in the given sample was due to his skill level, while the rest is noise. This past skill level is not guaranteed to be the player’s future skill level, and Russel Carleton warns against that in one of this stabilization articles. He goes on to verify that one of the metrics (I think it was K%) is only 25% predictive at the stabilization point.

    Side note: “skill level” is probably the wrong choice of words since this is also dependent on the pitchers and the situation and a hundred other things. The point is that this underlying metric in the first, say, 60 PA, is not guaranteed to be the same value in the second 60 PA.

    Vote -1 Vote +1

      • Eno Sarris says:

        From that piece “By doing this I could answer the question “If you took Player X and gave him 60 PA, and then gave him another 60 PA in roughly the same circumstances, how well would his performance in each match up?” The rest was a matter of number-crunching to find out where the correlation crossed .70, which is the point where the R-squared crosses 50%. At that point, the signal outweighs the noise. There will always be random variation driving results in baseball, but at r = .70, the sample size was big enough that we could assume that the ups and downs had evened out. I had my defensible number with which I could set my minimum inclusion criteria!”

        As for the warning about the circumstances being the same in the second 60 PA, that’s where fantasy has to take a leap of faith that would make him cringe. As a fantasy player, I can’t wait all season for proof. So this moment becomes a meaningful moment for prediction for me.

        Vote -1 Vote +1

      • novaether says:

        The quote you gave isn’t saying that the 60 PA in “roughly the same circumstances” come directly after the first 60 PA. If I’m understanding the methodology correctly, he takes 120 PA and splits them into every possible 60/60 split via a mathematical trick (KR-21). For example, one set could contain all of the odd numbered PA while the other set of 60 contains all of the even numbered PA. While one of the splits incorporated in the analysis is first60/last60, it’s not the only set. That’s why he uses the phrase “in roughly the same circumstances”. When he examines first60/last60, he gets an r = .5 (about), which is of course and r^2 = .25.

        That being said, I agree 100% that you can’t wait for the season to be over before making a move. The 2013 K% aren’t 100% predictive in their own rights, so there’s a give and take between history and new results. I suppose I’d just be a little slower to condemn Aoki and pat Montero on the back.

        Vote -1 Vote +1

      • Eno Sarris says:

        Ah I see how I might have misread that part. Still, if a current stat explains more than 50% of the variance, it can be useful for us to look forward even if a more strict statistician wouldn’t use it.

        Vote -1 Vote +1

  10. MLB Rainmaker says:

    Any thoughts on what is going on with Matt Adams?

    The AVG is there, K% is down, but so is BB%, and ISO has vanished for a guy that was supposed to have super power.

    Vote -1 Vote +1

  11. Poinmonster says:

    What are we thinking about this Ian Desmond fella? I kept thinking he was going to rebound but I’m starting to wonder. His BABIP is disturbingly normal and the Ks are up and walks are down.

    Vote -1 Vote +1

  12. Satoshi Nakamoto says:

    What on Earth is wrong with Fielder?
    Just in relaxation mode because his contract is guaranteed for the next 6 seasons?
    He keeps dribbling weak grounders in the infield.
    And he doesn’t look too upset about it.
    Just doesn’t care?

    Vote -1 Vote +1

    • Satoshi Nakamoto says:

      Continuing the Rant against my Rangers, when is Washington going to realize that Andrus is Not a #2 hitter?

      It’s so annoying watching Andrus look up at the video replay of his weak groundouts. Every single time he makes an out, he looks up at the big screen.

      Vote -1 Vote +1

    • Ruki Motomiya says:

      Fielder has been declining in most stats for the past 3 years, especially in power, and changes to Rangers park may have also made it less hitter friendly. Basically, he was a terrible choice at the start of the year.

      Vote -1 Vote +1

  13. zer0grav says:

    Looking at K% differential definitely gives a very interesting perspective. The guy who wrote an article on Upton the other day made some very interesting points.

    I’d like to point out Kyle Seager on this list as one of the more interesting guys.

    It feels like Fangraphs writers are collectively pretty down on him, but I see tremendous upside. Yeah, his strikeout rate is at a career high. However, both his swing, and O-swing, percentages are at career lows. As a result, his walk rate is now at a career high. Seager’s BABIP is .042 points below his career average, and with his newly retooled swing, I see good things in his future despite his career high strikeout rate.

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current day month ye@r *