QU Dashboard - A Progressive Understanding in Player Development Part 1: Pitchers
Long before I took this current role I have with the Quincy University Baseball Program, I stood at a table with a current scouting analyst with the New York Yankees. We were at the MLB Winter Meetings in Las Vegas. We discussed a lot of things, particularly was building a Baseball Savant-esque web app for college baseball. It was hard to do that for me at that time, since I did not have access to data. Now, in my role, I have access to copious amounts of data, which I then began my project. A web application that provides visualizations and data tables to help players better understand their strengths and improve upon their repertoire/hitting style.
Without further ado, I present to you the QU Dashboard, which will be broken down page by page.
As you can see, our dashboard has a four tabs, color coded by our team colors.
First, we will start with pitchers. Right under the pitchers’ name, there is a date range that filters data based on the input. I will show you an example shortly. When you select on a pitcher, velo and spin data with appear, that contains average velocity, max velocity, average spin rate, max spin rate, and (not pictured) bauer units, or spin rate divided by velocity. This is helpful in understanding what kind of velo and spin a pitcher has and how often he is throwing each pitch. We also use Motus sleeves which tell us the stress of each pitch, among other things we use for analysis, though not included in the app.
Next, in that dropdown, we can select another option, like Game Box Score.
As you can see, I changed the Date Range as well to October 16th in the left range, that lets the web app know, I would like to see this pitcher’s box score on October 16th, as well as any other corresponding data. In our box score, we have innings, hits, runs, walks, strikeouts, first pitch strike percentage, strike percentage, and offspeed strike percentage. Useful in understanding what pitches worked and what didn’t.
The last dropdown are season stats, in which I zoomed out so we can see all of them all once. The stats we have listed are innings pitched, Earned Run Average (ERA), Strikeout Percentage (K%), Walk Percentage (BB%), Strikeout-to-Walk Ratio (K/BB), Batting Average Against (BAA), Fielding Independent Pitching (FIP), Batting Average on Balls in Play (BABIP), Walks and Hits Per Innings Pitched (WHIP), Pitches per plate appearance, First Pitch Strike %, Strike %, Offspeed Strike %, Groundball Percentage (GB%), Line Drive Percentage (LD%), Flyball Percentage (FB%), Pop Up Percentage (PU%), and Hard Hit % (Balls hit 90+ MPH at the Division II level). This allows us coaches and each player to track their progress as the fall season progresses.
While we move down the list, I’m going to skip ahead to the line charts and density plots of velocity and spin rate, then come back to the Video and Illustrator tab.
With these charts, we can get a clear picture of how hard Pitcher X is throwing each pitch, as well as how much it is spinning. Also, we have an idea of how frequently this pitcher throws around a velo and spin rate. We can see that this pitcher has a fairly consistent spin rate across his pitches and throws his fastball around 88 mph most of the time. The cool part (pictured below) is again this data will subset based on the date range we select, so we can track progress over time.
From this pitcher’s last start we can see that his Fastball velocity was down a bit and that he threw mostly fastballs with about 4 or 5 sliders and changeups each mixed in. Want to isolate a certain pitch to see only that on the graph? Well we can do that too.
By double clicking a pitch, we can see only the data of that pitch. Now, we move on to the Illustrator part of the app.
The illustrator has five dropdown options, the first being a heatmap. This pitch heatmap shows where this particular pitcher throws for the most part. We can see here that this pitcher lives middle-to-up in the zone, away to left-handed hitters. This allows us to quickly understand why this pitcher is struggling and is among the highest in hard-hit-rate.
Secondly, we have total pitches, displayed in a hexbin chart, much like this example from Baseball Savant. This lets us know in another format how frequently a pitcher throws in a particular spot. Seven pitches down the middle? Not great, given this pitcher’s spin characteristics on his fastball (see first image). Pitcher X’s strengths are working down and in to RHH, and down and away to LHH. Again, color coded by our team colors.
Next, we have release point charts across all pitches. Again, the data subsets by date and we can isolate a certain pitch. This is the only plot that has a pitcher’s view. We see here that a lot is going on. Early on, we saw that this pitcher was releasing close to zero on the side and varying heights. Pitches were getting left up and over the plate, leading to a 33.3 percent hard-hit rate. We decided that it would be best for this pitcher to move off the right side of the rubber (pitcher’s view) and set up on the middle of the rubber. Here’s how he fared his last start with his release point chart, pitch heatmap, and pitch hexbin chart.
Now, the game plan of going down and away is coming more to fruition from moving the release point, but again, we need more data to support that cause.
The fourth feature is pitch movement, with an ellipsis that shows us the potential movement of each pitch. Movements are fairly similar across all three pitches.
Our last feature is your typical pitch charts, where you can select four subsets, all pitches, swings, whiffs, and balls in play. Again, the data will subset by date and subset by what selection is chose from the options in the sentence above. This data allows us to understand what pitches generate swings and misses in what locations, as well as what pitches generate contact in what locations.
Lastly, we have video for every pitch and every plate appearance. In the dropdown we have these options:
Our views so far are Home Plate, Side View, CF View, Quad View, and Second and Third Innings are also Quad View.
In our Quad View film, we have synced up angles from four camera angles, as well as FlightScope data overlayed. See example below.
We can see video from 4 different angles and analyze it with data. We can also isolate a singular camera angle on a mobile browser by triple tapping a camera view.
I wanted to build something that can encapsulate as much data and visuals as possible. On the pitching side, we have a ton of data and video at our fingertips to aid in player development. Stay tuned soon for an article on the hitting side of this dashboard.
