I’m not sure there is anything meaningful here, but I thought it would be neat to take a look at how the Rays fared over the course of a game by each pitch number. After reworking the data a bit I was able to line up all the pitchers in each game and extend out the number of pitches so that it would be easy to analyze the data. Note: This does not contain our pitchers which may have thrown off a couple of games, but I didn’t think it would be a huge deal with this sample of over 25K pitches. Again, I chose to use SLGCON here (recall that SLGCON is Bases/Balls in play. For example, the Rays swung at 7 first pitches of the game with a single and a HR and 5 outs to show for their efforts. 5 bases/7 balls in play is approximately .714 (league average is around .500)), mostly because it’s pretty easy to calculate within the pitch f/x database. After finding the SLGCON for each pitch (number 1, number 2, number 3, etc…) I plotted the SLGCONs into the below chart. I removed the data points which were a random mess all over the place and installed these trendlines to give an idea of how the game moved along.
You’ll notice the blue line is a 10-pitch moving average, while the black line is a linear trendline. The latter shows us about what you would expect, over the course of the game teams tend to make better contact the more pitches that have been thrown. The blue line lets us look at how we tend to move around the linear trendline over the course of the game. You shouldn’t be surprised to see a nice peak through the first 20 pitches as that will typically be our best hitters and studies have shown that most pitchers need a few to get their fastball up to speed and get a feel for their secondary stuff. We can then see that as we move through the cycle of the line-up that around the 60th pitch we start to make real strong contact, on average, over the course of 2010. This continues with a bit of fluctuation up to around 100 pitches when we will typically see the game handed off to the bullpen. The Rays had a nice uptick from pitches 120 to about 135 then had a serious downturn before seeing better results over pitches 140-150. I stopped there because the Rays averaged 151 pitches per game and the data becomes a bit more sparse after this mean.
I think you see about what you would expect from a team over the first 150 pitches of the game and I’d like to do this for another team to get an idea of how we compare. Does anyone have any suggestions about a team or two that they would like looked at? Bear in mind that this is a pretty long process to get the refined results below, but I still think it would be cool to see if this is the norm or if that’s how bad our offense performed through the first couple of times through the order. Don’t forget that these don’t include walks, which were a pretty important part of our offense. I would love to try calculating wOBA or BA/OBP/SLG on these, but it’s a bit of a pain in the neck using this approach. I also tried using linear weights on the pitch results to get an even better idea of what’s happening (this approach would factor in balls and strikes as well as balls in play), but, sadly, I have not been taught a reliable way to do this, yet. This is just a first look, and as I mentioned, I’m not sure how reliable the data is below since there are some small sample issues for some of the pitches, but I think it’s pretty cool and seems to jive with the flow of the game that I’m already starting to miss. As always, click the image to enlarge in a new tab/window and let me know which teams you would like to see us compared to and, in the meantime, I will find better ways to present this since I find it pretty cool.