Beyond the GPS tracking issues that others have mentioned, their estimated arrival times strongly underestimate the impact of traffic. Something I explored here: pitchingbot.com/busmodel.htm…

We're looking for a data scientist to join us at the Cleveland Guardians R&D department! I've really enjoyed working here with lots of great people for the past 2+ years. If you're interested in baseball analysis and solving tough problems with data then apply at the link below:

🇬🇧

As I've left the public space, PitchingBot has a new home on FanGraphs. fangraphs.com/leaders.aspx?p… Check out the primer here if you want to know more about what goes on under the hood: library.fangraphs.com/pitchi…

I'm pleased to announce that next week I'll be joining the Cleveland Guardians R&D team! Thanks to everyone who has helped me to develop in the world of baseball analytics over the past few years.

Early Christmas present - submitted my thesis

Fair enough, I thought the graphs above avoided that by changing the samples to still include players on the "bad" side of random variation.

I couldn't locate the wOBA graph, but here are home runs and Ks

I agree the regular season results are mostly selection bias. I redid some of the postseason analysis comparing to the odds ratio wOBA of the batter-pitcher matchup and there seemed to be something real there (didn't publish that though, I'll see if I can dig it out)

Hard to say much. Note that the Phillies game would have come under "Cloudy" using this definition of weather. Fastballs show a bit more of a significant change from clear -> cloudy compared to other pitches.

I filtered all of those out but I agree it would make a good control group.

I didn't include snow in the first graph due to a small sample size, but the effect looks big even with a large error bar.

Pitcher command quality in different temperatures. Sweet spot from 55-70 degrees?

Effect of different weather conditions on pitcher command quality.