Valuing 2015 Pitcher Performances | The Process Report

Valuing 2015 Pitcher Performances

Recently, my good friend Dave turned me on to this outrageously excellent post from a user named   over at Minor League Ball. Go read it. I’ll still be here. Being the colossal nerd that I am I figured I would take a stab at replicating something like this at the Major League level for pitchers and batters. Today we’ll look at the findings from the actor portion of the plate appearance.

One of the truly great things that exists in this universe is Daren Willman’s site Baseball Savant. While the walk, hit by pitch and strikeout numbers are easy to find at numerous sites, Daren makes it possible for any noob to pull specific data regarding virtually every pitch thrown since 2007. I would not have been able to pull the ball in play (BIP) data without his excellent site so go give him money or your time or whatever else he politely demands.

Mimicking the CTL (BB+HBP/PA) and K (K/PA) numbers was very easy. Once I had that stuff in my workbook it was even easier to rank each pitcher to get an idea of who excels at each of these two very important parts of getting batters out. I was able to pull data for infield fly balls, line drives, groundballs, and outfield fly balls with a bit of a grind by isolating the handedness of the batter and whether the ball was hit to the opposite field, up the middle, or pull side based on the fielder that made the play. From there I used Joe Sheehan’s run values, which adjust each result for the count in which it was hit to create an idea of the value gleaned from each pitcher based on the trajectory and direction of the ball hit. To adjust for the fact that not all pulled line drives, for instance, end up with the same result I have regressed these figures to the average run value for that trajectory and direction using a weight of the average number of balls in play plus one standard deviation.

As an example let’s look at Chris Archer. By my count he yielded 34 pulled line drives in 2016 and had an average run value of 0.468 runs per pulled line drive. The weighted league average for that result was 49.9 pulled liners at a run value of 0.439. Instantly we can see that he averaged fewer occasions of this odorous result, but did yield slightly more runs than the average pitcher. The regression formula would look like this:

((34*0.468)+((49.9+14.4)*0.439)/(34+49.9+14.4) = 0.449

We can do this for all pitchers for all trajectories and directions to get an idea of their regressed run values per BIP. Furthermore, when we multiply this regressed figure per BIP we can then get an idea of the total run values yielded for each type. This allows us to compare every pitcher (with at least 300 batters faced) across each type to give us those nice percentages. For those that are interested here’s the weighted average run values and number of balls in play plus one standard deviation:

averages

Sum the regressed RV multiplied by the number of occurrences for each type and you get a total run value allowed figure, which easily allows us to compare all pitchers. Now that we have the BIP component to add into the CTL and K parts we can get a good idea of how these pitchers rank. The inspiration for this used weights of .23, .43, and .33 for the CTL, K, and BIP figures, respectively, but I want to tweak this just a nudge. Let’s drop those slightly to .2, .4. and .3, but also introduce a weight of .1 for the number of plate appearances so that we can also lend some favor to those that not only provide quantity, but do it with volume, as well. You can find this figure in the “Overall” column. This yields something like this:

Top40

In all instances you want to be closer to 100% so a guy like Max Scherzer being in the 98th percentile for walks and 99th percentile for strikeouts is such a beast because he’s not trading off walks to get more strikeouts like, say, Tyson Ross or Francisco Liriano. This certainly seems to pass the smell test as all of these guys are quality pitchers. Let’s move on to the next set:

2nd 40

As you work your way down the first image you start to see guys having to trade off one thing for another and that is the case nearly across the board for this group. The newest Marlin Wei-Yin Chen has very good control, but that comes with average strikeouts and results on balls in play. Mark Buehrle featured even better control and very good results on his balls in play, but that comes at the price of not striking anybody out. Ian Kennedy is a free agent that is still available and you can see that he’s near average in walks, but very highly rated at getting punchouts. His balls in play are brutal and that stems from all those line drives and fly balls. I’ve long thought he would be a perfect fit on a team like the Royals that turns those hard hit balls into outs more often than other teams. Let’s move on to the next forty guys:

3rd 40

Now we’re getting into the guys that are likely to either be really good at one thing and really bad at the others or just ok across the board. Karns is a good example as a guy that can give you some strikeouts, but that comes at the cost of lots of walks and getting hit fairly hard on balls in play. Former Ray Jeremy Hellickson is around the average for the stuff that he solely controls, but well below that when batters hit it between the chalk. Let’s move on to the back-40:

Last 40

You may find a few guys that manage their free bases, but none of these guys got strikeouts and the vast majority suffered when batters made contact. There you have all of the pitchers that faced at least 300 batters last year, but his being a Rays blog and all let’s dig in further focusing solely on the four Rays pitchers that met our qualifications last year.

Rays

Ranking by Overall we saw that Archer finished 17th, Odorizzi 39th, Ramirez 62nd and Karns 86th. They got there in a variety of ways. The Youth row is the 2015 Overall score multiplied by the Age percentile at .8 and .2 weights, respectively. Archer was every bit the ace last year combining league average control with elite strikeout stuff and fairly impressive scores on balls in play. He did get hurt on line drives back up the middle and ground balls to the opposite field, lesser so, but was average or above on every other type of ball in play. Grounders were his biggest weakness just slightly outpacing liners, but he really excelled when batters hit the ball up in the air to the outfield.

Odorizzi was also very impressive as he showed the best control of this group, while still getting oodles of strikeouts. His balls in play were his weakest component due nearly entirely to the line drives that he gave up, especially those that came back through the middle. His grounders were merely ok, but his fly balls went are where he really shines.

Erasmo wasn’t quite as good as his peers at getting strikeouts, but that says more about this staff as he was still pretty close to league average. He is, however, very good at limiting damage on balls in play. Unlike these first two he does get hurt quite a bit by the fly ball, particularly those that are pulled or hit to the opposite field. He is pretty adept at limiting the damage on grounders and liners, however.

The departed Karns was very good at getting the strikeout, but his control and balls in play left a lot to be desired. Where he really got killed was on grounders, especially those to the opposite field and up the middle. He wasn’t hurt as much on his liners and his fly balls were around league average, overall, with those to the opposite field going to die a ton.

In the next installment I will take a look at the hitters that saw at least 300 plate appearances in 2015 viewed through this same exact lens. I hope you enjoyed this and if you would like to receive this workbook please shoot me an e-mail at sandykazmir@dhazebay.com

 



4 Comments

  1. […] CLICK HERE TO READ MORE […]

  2. OTownRaysFan wrote:

    nice work – keep ’em coming!

  3. […] Valuing 2015 Pitcher Performances […]

  4. […] recently I published the findings of my deep dive into rating pitchers using walks, strikeouts, and ball in play data. The first two parts are something you’re […]

Leave a Reply

#layout { padding-left:20px; }