On Player Observations

This weekend I had the privileged to attend the Saber Seminar at Boston University. The conference is held as a fund raiser for the Jimmy Fund, a cancer research charity associated with the Dana-Farber Cancer Institute. It brought together a lot of great minds, guys from baseball operations in multiple franchises, analysts and players, saber-maticians and college students. Needless to say, it was a blast, and rather inspiring.

As I mull over the thoughts and ideas from the conference, I hope to share some new findings here; but first, I wanted to share something not so related to saber metrics at all.

To my surprise, a great portion of the presentations focused on scouting and player observation. Guest speakers focused on amateur and professional scouting, pitching mechanics, and defense evaluations (which is a little more stat oriented). All of the talks were aimed at bringing guys out of the numbers, and back into the field of play. It was illuminating.

Player Observation

Among many of the traditional stat minds, one guy's convicting words stood out. ESPN prospects guru Keith Law spoke off-the-cuff and with clarity on Saturday, and he stressed the importance of compartmentalizing your "old" or preconceived notions of players when making observations.

When evaluating hitters, Law recommended attending batting practice, and even using video to capture multiple easy swings to to see ideal mechanics. You can then leverage practice footage against live footage from the game, and compare differences in the swing or approach in a high leverage situation. (He made the caveat that infield practice is not beneficial in the same ways.)

In terms of evaluating one specific player live, he encouraged sitting in the section of the stadium that would give you the best view, and again, bring a video camera; to use the naked eye for some basic observation, but to never trust it, and re-watch players on tape to confirm your suspicions.

With the naked eye, you can identify approach to the game, player reactions, and even judge "instincts" (which sounds intangible, and more so when it's referred to as "intuition" or "feel" for the game, but is a plausible observation). Following the mantra, "When there's smoke, there's fire," if your brain recognizes something happen, then something probably happened. The trick is to defend your observation with rational arguments.

This makes intuitive sense. Clear your mind, and watch the play at hand. When you're evaluating in the moment, don't think "This guy is gonna suck at hitting a change up," because you'll fall into confirmation bias and/or look for the wrong things. Watch the player, see how he does. Then go back to the stats and confirm what you saw was right or wrong. Let that re-inform your notions. And because our eyes can fails us, he would not shut up about his video camera. It's a cool idea, even if you might look silly.

I hope you found this insight as compelling as I did.