As I’ve mentioned ad-nauseum, at the end of August I bought a Flightscope Mevo launch monitor, and I love it.
The one major thing the Mevo lacks is a display of long-term trend data. The only data view available is for each individual “session”. This makes the Mevo a great tactical tool on the range, but a rather mediocre strategic tool on its own.
Step 1 was to get all of the data that Mevo’s collected over the past five months into a more workable format.
Flightscope’s Myflightscope portal allows you to download individual range sessions to a CSV (comma-separated value) data file. Unfortunately, it gives each downloaded file the same unhelpful filename (stats.csv). The files themselves contain no reference to which session they contain, so turning them into historical trend data required some work.
First, I created a directory for each year I’ve had the Mevo, and subdirectories for each month, and each day that I used it. I then downloaded the CSVs for each session manually (about 60 of them), one at a time, and put them in their respective folders. This was a tedious process, and hopefully Flighscope will add better titles soon to make this easier, but if they don’t, there are a few possible ways to automate this.
Next, I wrote a script (available for free here) in Node.js to parse all of the CSVs, create columns to identify which file and date each shot came from, and create a CSV with all of the shot data merged.
The output of this script was a merged CSV of thousands of golf shots.
The next step was to turn that data into meaningful. I’m not a statistician (or even skilled at Excel), so figuring out how to do that will be an ongoing process.
I started by importing my CSV into Google sheets.
This allowed me to make:
- Scatter plots measuing the dispersion of carry, smash factor, and other statistics for a given club over time.
- Trend line charts for the medians of those same statistics. I was initially looking at means, but I concluded that it’s more important to see the range I’m likely to hit within for a given statistic, rather than the average of my best and worst shots.
This was a good start, but the data you see above is flawed. The aggregate statistics ocasionally rely on small sample sizes and don’t filter for outliers, leading to a few wacky data points. Google Sheets isn’t super beginner friendly, and importing new data into this sheet, in the current way, would be a manual process.
Luckily, I’m in the market for a new hobby, and this seems like a fun one, so I’m going to try to build some tools to aggregate and analyze Mevo and ArcCos data over these next few months (and hopefully improve my golf swing along the way).
If you’re interested in working with me to develop better systems for aggregating and analyzing Mevo data, you have thoughts, or you’ve already implemented a solution of your own, please leave a comment.