In this week’s GPS Location Report we are going to take a slightly different approach. Recently, Matt Williams approached me asking if it would be possible to track some of the Data Monster metrics on a game-by-game basis. As we all know pitchers can evolve mightily over the course of a season and he was wondering if we would be able to build a tool to help track that. So I decided this week to build a proof of concept of sorts to see if there are some interesting trends that we can parse out of the data.
In order to do this, I decided to take a look at one of the hottest pitchers in baseball over the last month, Tarik Skubal. If you go by season-long numbers, you’ll be mightily disappointed by Skubal. However, the young left-hander has been absolutely dominant in his last four starts. He has posted a K% over 30% in his last four starts and has also hit that number in five of his last six. He has allowed five earned runs in 21 innings in that same span. Quite frankly he has taken a massive step forward. So the question is, have any of his underlying Data Monster numbers changed as well?
When you look at Skubal’s numbers on a game-by-game basis, one big trend jumps out, his SwStr% is up.
As you can see outside of a random spike early in the season, his last four starts have easily been his best with an SwStr% well above his season average. This has been a huge reason for his success and has led to a massive jump in his strikeouts. So what exactly is causing this jump? Pitch mix change, stuff improvement, or pitch locations?
Looking at his 2021 pitch mix chart, we don’t see any major changes in his profile recently. His fastball percentage is dropping but instead of offsetting that with more breaking/offspeed pitches, he seems to replace some four-seamers with sinkers. So nows let’s see if there’s a change in “Stuff” mainly by looking at his vertical movement charts.
Again, looking at this we don’t really see anything beyond the normal variation for a pitcher showing that this has held fairly steady. So this leaves us with one final explanation for the improvement, location.
If you look at the chart above which plots his location-based xWhiff by game, you’ll see that it generally follows the trends of his actual SwStr%. Early in the season, Skubal saw a big spike, and then over the last four games, he has been above his overall average. This leads me to believe that the recent success that Skubal has had, has been due to improvements in his ability to locate for whiffs. While the overall shape of the pitches has not changed, the execution has been much better helping to lead him to a big jump in K% and overall success. In terms of the underlying results, the xWhiff difference is not much, however, locating pitches better can also cause a particular pitch to “play up” and can allow it to be even more successful. This will be the key with Skubal, can he continue to locate better than previous starts in order to generate more and more whiffs.
Based on this simple proof of concept I think there’s a ton of value in a game log type of catalog and as a result, the ability to generate graphs like these will be added to the Data Monster shortly.