“The scientist is not a person who gives the right answers, he is one who asks the right questions.”
—Claude Levi-Strauss
One of the aspects of writing a weekly column that I enjoy most is the reader feedback; sometimes it starts to arrive just minutes after the column is posted. Although I may not get to every email, rest assured that they’re all read and many of them populate my “future columns” queue. So in that vein–and as a thank you for asking the right questions–today we’ll answer two specific questions prompted by previous columns: one related to baserunning and, and another taking a look at plate discipline.
Taking Advantage Redux
One of the most frequently asked questions after last week’s column concerned the repeatability of all the baserunning metrics, including Equivalent Other Advancement Runs (EqOAR). Reader Aaron voiced it this way:
How much correlation is there year to year in the baserunning metrics? Is there significant year-to-year consistency in the individual metrics? Or in the combined sum of the various baserunning metrics?
I ask because given the relatively modest range in values on a year-to-year basis, I’m wondering how much predictive value these metrics have in estimating a player’s future worth.
That’s a question that can be answered definitively, so the following table records the correlation coefficient for all consecutive seasons by any runner who meets or exceeds the opportunity thresholds noted for the given metric:
Metric Opps r EqHAR 30 0.36 EqSBR 10 0.19 EqGAR 10 0.18 EqOAR 100 0.03 EqAAR 10 0.02 --------------------- Total 25 0.26
Advancing on hits is the clear leader, followed by stolen base runs and advancing on ground outs, with other advancement and advancing on fly outs bringing up the rear. The reason the latter two lag behind is the combination of their scarcity and the disparate impact that individual events can have on seasonal totals. That is, runners simply don’t get very many opportunities to advance on fly outs, wild pitches, and passed balls; when they get thrown out (which is rare) the impact can significantly drag down their total for an entire season.
To answer Aaron’s question, we can say that none of the metrics is that strongly correlated from season to season, and so outside of the top runners like Chone Figgins (who you could reasonably assume will contribute between five and 10 runs) and those at the bottom like Paul Konerko or Jorge Posada (who cost their teams around 4.5 runs), I wouldn’t bet a great deal on repeatability in runs contributed through baserunning.
With that said, when grouping seasons into even and odd years and then performing correlations on the totals, the numbers–as you would expect–go up:
Metric r EqHAR 0.60 EqSBR 0.55 EqGAR 0.30 EqOAR 0.07 EqAAR 0.07 ---------------- Total 0.37
Finally, this discussion also provides the opportunity to update a graph that I created last summer that puts each metric into perspective in terms of its overall contribution to the running game, as measured in runs:
So, not only is EqHAR the most repeatable–because it allows good baserunners to differentiate themselves from their peers and is therefore less susceptible to randomness–but it also has the highest aggregate impact because of the greater number of opportunities. On the flip side, EqAAR has a relatively high impact since it primarily deals with scoring on sacrifice flies but encompasses relatively few opportunities and has a low skill component.
But Fish Aren’t Square?
After our discussion two weeks ago on plate discipline for hitters, several readers wondered if a similar version could be created for pitchers. Typical was this question from Bryce:
Your Fish/Eye charts for hitters were very interesting and makes me wonder how the pitchers would stack up? Which pitchers encourage chasing more than others?
Using the same metrics as with hitters, we can get a pretty good sense of a pitcher’s stuff. But before taking a look at the pitching leaders and trailers in the four metrics, it should be noted that I’ve tweaked the calculations a bit this time around. Previously, the area of the strike zone was defined by the regulation width (17″) and the height as recorded in the PITCHf/x data. Now I’m accounting for any part of the baseball touching the zone, and so have expanded the area an inch and a half (the radius of a baseball) in all directions.
Without further ado, here are the leaders and trailers for the 204 hurlers who’ve recorded 350 or more pitches:
Pitchers Sorted by Fish Name Pitches Fish Square BadBall Eye ------------------------------------------------------------ Cole Hamels 445 0.423 0.788 0.738 0.252 Scott Baker 1140 0.384 0.878 0.699 0.189 Cla Meredith 757 0.383 0.857 0.567 0.202 John Smoltz 1893 0.376 0.829 0.576 0.261 Takashi Saito 682 0.357 0.757 0.594 0.216 Johan Santana 729 0.354 0.713 0.558 0.204 Jonathan Broxton 926 0.350 0.799 0.562 0.220 Bobby Howry 640 0.348 0.826 0.791 0.185 Jeremy Bonderman 980 0.348 0.899 0.649 0.240 Aaron Harang 1236 0.347 0.857 0.623 0.234 ------------------------------------------------------------ Matt Chico 375 0.227 0.846 0.841 0.267 Jorge de la Rosa 492 0.222 0.903 0.653 0.243 Brian Bannister 1258 0.221 0.933 0.762 0.267 Ryan Feierabend 529 0.220 0.894 0.784 0.220 Jo-Jo Reyes 619 0.216 0.907 0.848 0.248 Jamey Wright 848 0.214 0.929 0.707 0.246 Steve Trachsel 598 0.210 0.920 0.754 0.275 J.D. Durbin 539 0.206 0.920 0.704 0.283 Cliff Lee 371 0.201 0.906 0.759 0.249 Mike MacDougal 359 0.178 0.815 0.639 0.271
When used to evaluate pitchers, Fish can be seen as a measure of how often the pitcher enticed the batter to swing at a pitch out of the strike zone and so is a direct answer to Bryce’s question. Cole Hamels is separated a bit from the pack at over 42 percent, although the sample size is a bit smaller than it is for the other pitchers. Unlike with hitters, although I didn’t actually run the numbers, it’s clearly the case that the performance of the pitchers at the top of the list far exceeds those at the bottom, indicating that their stuff is more deceptive. The average value for this metric is 29 percent; attentive readers will note that this average will be lower since the strike zone is expanded.
Pitchers Sorted by Square Name Pitches Fish Square BadBall Eye ------------------------------------------------------------ Franklin Morales 378 0.239 0.955 0.681 0.282 Jon Lester 428 0.320 0.941 0.648 0.264 Aaron Laffey 354 0.280 0.938 0.815 0.284 Kenny Rogers 461 0.315 0.934 0.725 0.238 Odalis Perez 388 0.287 0.934 0.650 0.333 Livan Hernandez 1060 0.228 0.934 0.788 0.260 Brian Bannister 1258 0.221 0.933 0.762 0.267 Mike Mussina 526 0.322 0.932 0.886 0.319 Jamey Wright 848 0.214 0.929 0.707 0.246 Tom Glavine 626 0.265 0.924 0.825 0.245 ------------------------------------------------------------ Cole Hamels 445 0.423 0.788 0.738 0.252 Ron Mahay 403 0.236 0.785 0.509 0.248 Brandon Morrow 820 0.269 0.782 0.641 0.252 Ryan Rowland-Smith 376 0.257 0.774 0.773 0.279 Scott Kazmir 704 0.272 0.774 0.582 0.232 Takashi Saito 682 0.357 0.757 0.594 0.216 Joaquin Benoit 862 0.330 0.751 0.591 0.187 Michael Wuertz 365 0.310 0.750 0.397 0.256 Johan Santana 729 0.354 0.713 0.558 0.204 Santiago Casilla 389 0.320 0.707 0.452 0.268
Square records how frequently hitters make contact with pitches that they swing at that are in the strike zone. An average value is 86.5 percent, and Rockies rookie Franklin Morales–who has had mixed success thus far–leads at over 95 percent. He’s somewhat atypical of the group, since many of the rest are soft-tossers who don’t induce many swinging strikes. On the other end of the spectrum, Santiago Casilla (the suddenly aged A’s reliever formerly known as Jairo Garcia) is down around 71 percent. Although the performance difference between those at the top and at the bottom is not as obvious as in the above table, we could reasonably interpret low Square values for pitchers as an indication of pitchers with exceptionally good stuff, since hitters are having difficulty in getting a bat on the ball, even when the pitches are in the strike zone.
Pitchers Sorted by BadBall Name Pitches Fish Square BadBall Eye ------------------------------------------------------------ Mike Mussina 526 0.322 0.932 0.886 0.319 Paul Byrd 887 0.309 0.918 0.870 0.291 Kyle Kendrick 654 0.347 0.906 0.851 0.276 Jo-Jo Reyes 619 0.216 0.907 0.848 0.248 Carlos Silva 1394 0.326 0.902 0.846 0.243 Woody Williams 1088 0.304 0.914 0.844 0.263 Curt Schilling 699 0.305 0.869 0.843 0.277 Matt Chico 375 0.227 0.846 0.841 0.267 Horacio Ramirez 1270 0.276 0.912 0.836 0.282 Vicente Padilla 1570 0.283 0.913 0.836 0.250 ------------------------------------------------------------ A.J. Burnett 1495 0.299 0.866 0.556 0.268 Tyler Yates 640 0.290 0.800 0.556 0.203 Sean Green 512 0.301 0.842 0.545 0.266 Erik Bedard 564 0.293 0.873 0.525 0.281 Ian Snell 775 0.274 0.862 0.514 0.243 Ron Mahay 403 0.236 0.785 0.509 0.248 Carlos Marmol 748 0.296 0.801 0.496 0.368 Francisco Rodriguez 689 0.282 0.802 0.462 0.332 Santiago Casilla 389 0.320 0.707 0.452 0.268 Michael Wuertz 365 0.310 0.750 0.397 0.256
As you might expect, the top of the list is populated primarily by pitchers who don’t throw all that hard, and it should come as no surprise that Square and BadBall are pretty strongly correlated (r=.52). The bottom of the list contains pitchers like Michael Wuertz, Casilla, K-Rod, and Carlos Marmol, all of whom sport nasty sliders that make it difficult for hitters to make contact on balls out of the strike zone. The average for everybody is 70 percent.
Pitchers Sorted by Eye Name Pitches Fish Square BadBall Eye ------------------------------------------------------------ Scot Shields 839 0.263 0.895 0.663 0.379 Carlos Marmol 748 0.296 0.801 0.496 0.368 Dustin Moseley 890 0.321 0.923 0.789 0.347 Odalis Perez 388 0.287 0.934 0.650 0.333 Francisco Rodriguez 689 0.282 0.802 0.462 0.332 Justin Germano 1596 0.253 0.906 0.745 0.328 Heath Bell 892 0.326 0.855 0.592 0.327 Mike Mussina 526 0.322 0.932 0.886 0.319 Darren Oliver 432 0.303 0.819 0.775 0.318 Byung-Hyun Kim 687 0.245 0.835 0.679 0.317 ------------------------------------------------------------ Tyler Yates 640 0.290 0.800 0.556 0.203 Cla Meredith 757 0.383 0.857 0.567 0.202 Chien-Ming Wang 608 0.275 0.913 0.667 0.199 Chris Young 2072 0.336 0.816 0.660 0.196 Eric Stults 371 0.274 0.822 0.722 0.195 Rafael Soriano 610 0.298 0.805 0.631 0.190 Scott Baker 1140 0.384 0.878 0.699 0.189 Joaquin Benoit 862 0.330 0.751 0.591 0.187 Bobby Howry 640 0.348 0.826 0.791 0.185 Alan Embree 535 0.310 0.819 0.610 0.183
In this metric, the link to performance as a clear indicator of stuff is probably even a little more mixed than with the others. For example, although Marmol and Rodriguez once again make appearances at the top, so do pitchers with lesser stuff, guys like Darren Oliver and Odalis Perez. The average here is 26 percent.
Using this information, we can now create a similar chart to the one we created for hitters. However, instead of plotting Fish versus Eye, in our desire to get a feel for which pitchers have the best stuff or are the nastiest to stand in against, we’ll instead plot Fish versus Square. Using Fish, we see how frequently the pitcher enticed the hitter to chase, while with Square we can capture how difficult it is for hitters to make contact, even with pitches in the strike zone. Since Fish-Square has no catchy ring to it, we’ll call our resulting plot simply “Stuff”.
One of the things you’re likely to notice here is that the pitchers are bunched up much more than the hitters were in the previous article. Indeed, there is less variability among pitchers, a result I chalk up to both the larger samples involved with pitchers, and the fact that hitters’ actions are what is being recorded, so you might expect more differentiation in approaches at the plate to show themselves.
Even so, from the plot we can see that Cole Hamels, Johan Santana, Takashi Saito, Jonathan Broxton, and John Smoltz appear in the upper right-hand quadrant, and are among those pitchers who we can crown with the title of “Nastiest Stuff”.
You’ve Got Mail
As this is the final Schrodinger’s Bat during the 2007 regular season, I’d like to also take this opportunity to thank all those who have written me to offer their opinions and ideas. Please keep them coming throughout the offseason, as they are greatly appreciated.
Thank you for reading
This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.
Subscribe now