Pitcher Expectations and Trends

It took some doing, but I’ve finally set up some workbooks to look at pitchers in the same way that I’ve been looking at hitters. Before stepping over to that I want to provide the most recently updated charts for hitters to give a refresher on this stuff and to keep you, the beautiful public, apprised of what has happened since the last running of the numbers. Here’s a look at the summary:

Readily evident is how the offense has regressed to their mean since the last update. Their wOBA is basically in line with expectations with the team outperforming their xwRAA by just over a run. If black holes were the genesis of the universe then consider Jose Molina the Great Progenitor. Logan Forsythe has hit bad even for his meager expectations, but joining these two are Zobrist and Longoria, hitting similarly, and vastly under performing their expectations.

The other side of the coin shows Loney still doin his thang thang and the suddenly hot Yunel Escobar not far off. In fact, here’s a look at Yunel on a 20 PA rolling basis:

After a slow start he played around what you should expect and then has been scorching over the last 75 or so plate appearances. Let’s move on to the daily run observations:

I know the concept is a bit confusing so to simplify think of xRuns as the expected runs based on the matchup, aRuns as the expected run total based on the actual wOBA hit for, and Runs as the actual number of runs scored. The difference between xRuns and aRuns can be thought of as the offense under or over performing based on matchups and the difference between aRuns and Runs can be chalked up heavily to the timing of events and luck. If the team has a .218 wOBA, as they did last night in the opener against Oakland, you should expect them to score 1.4 runs. The fact that they actually scored zero runs shows poor luck and/or timing of their offensive production as measure by wOBA.

Taking an overview, we can see that after the initial stretch of poor performance and poor luck we went through a period where offense greatly exceeded expectations and actual run scoring was right in line with what you would expect. As the offense stayed above the norm you can start to see some poor luck coming in as Runs trails aRuns and then this most recent tailspin shows offense well below the expected level and Runs basically in line with aRuns. Not good. Let’s get to the original point here and focus on the pitchers:

Starting at the bottom we see just how much David Price has under-performed. He has pitched worse than team average when he should have one of the lowest allowed production levels. Based on matchups and regressed platoon splits he should have allowed five less runs than the average, but he has allowed over three runs MORE than the average. Sum that and you can see that he has been around 8 runs worse than you would expect. Here’s a look at how that has evolved:

The recently deceased Bell is in that same boat. U-G-L-Y. Cesar Ramos was a good man to take on more than he could chew, but thankfully Alex Cobb is back so Ramos can slot back in where he belongs. Speaking of Cobb, he was an absolute monster through his first two starts, and it’s not hard to see how much he has been missed.

Erik Bedard has really righted the ship as he has booked a couple nice outings in a row. Here’s what that looks like on a 20 PA rolling average:

I didn’t think we would get anything meaningful from the guy, but boy was I wrong. He could get hurt or turn into a pumpkin at any moment, but he has been a nice port in the storm that has been our starting rotation.

McGee, Beliveau (very limited), and Oviedo have turned out to be very reliable bullpen options. While I’m apologizing for my Bedard bastardry now would be a good time to point out that I wanted to see Oviedo do it before anointing him a reliable option. I was wrong on that, he has been good, and I hope it continues. There’s a lot here that I’m not going to touch on, but these images work great as a reference sheet so have at it. I now want to turn over to the runs framework established previously, but this time from the perspective of the pitcher:

The season started off nice enough with strong over performance and the pitching wasn’t all that far off expectation from there, though timing/luck of production certainly outpaced how many runs a team should have scored against us. Since the peak we’ve continued to see teams hit our pitchers better than you would expect, but the opposition has not scored runs quite at the rate you would expect so there has been a little bit of luck there. It could have been a lot worse, but we’ve seen some gritty clutch luck or whatever else you want to chalk it up to when a team doesn’t score as many runs as they should based on offensive output.

The crazy thing is when we look at the totals on the pitching side. The opponent has put up a .323 wOBA compared to a .313 expectations. These convert to 205 aRuns, 190 xRuns, and 204 Runs. So while our pitchers have under performed expectations by around 15 runs, the actual amount of runs scored is basically right in line with what you would expect based on the opponent’s offensive output. So while the batters have scored around 15 runs less than they should have the pitching is basically right in line. Luck should turn around on the offensive side leading to run production commensurate with wOBA, but on the pitching side there has been little luck involved and with Alex Cobb coming back and Ramos going to the pen we should be lowering the expected runs allowed a bit further.

We’re not staring into the sun here, but there may be a few rays of light creeping into the dark tunnel that this season has been thus far. Keep hoping that things play out to expectations and good things should start happening soon.


About Jason Hanselman

Rays fan.
This entry was posted in Batter Analysis. Bookmark the permalink.

2 Responses to Pitcher Expectations and Trends

  1. This is great stuff–excellent work. One quick question based upon my unfamiliarity with the data – Is Cesar Ramos in any way penalized for being a starter vs. his expectations I assume are based upon projections as him as an RP? That is a round-about way of asking how the expected statistics are calculated.

    • Jason Hanselman says:

      Great question. Because projection systems do not give a wOBA projection I use my own very rudimentary system for projecting their performance. I take overall career wOBA and regress with 500 PA of league average wOBA. A positive externality of this is that it does not adjust for relief or starting since it’s using the pitcher’s career numbers regardless of role. The downside is that it’s a less sophisticated projection, but perfectly adequate for our needs here.

      I’m coding in whether the pitcher was used in relief or as a starter and hope this will lead to interesting research down the road. Thanks for the comment and being a reader. My hope is that others find this as fascinating as I do.

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s