tangotiger

Tom Tango is the nom de plume. I am TangoTiger

Who did what now? Park factors are right here and they are done properly: https://baseballsavant.mlb.com/leaderboard/statcast-park-factors

Getting the data the way I need it is really the challenge. I don't know what "fair bit" means, but I'd think 2 or 3 runs.

That's the current iteration, yes. The "transfer" or "teamwork" portion of the play is not currently being handled. The DP is one of the big items that I'd like to get done.

We show the data right here based on where they stand on the field, regardless of what their official position is listed on the batting lineup: https://baseballsavant.mlb.com/visuals/statcast-infield-defense?type=Fielder&playerId=543760&startYear=2016&endYear=2023&result=&direction=&normalize=undefined&roles=&esrGT=0&esrLT=1&evGT=0&evLT=125&distGT=0&distLT=200&batside=&viz=intercept_fielder_starting_position_ We can see in this particular case Semien is far more effective standing in the 2B spots than the SS spots.

Right, the bath water position is a political position. We have been very very good in showcasing stats at a pretty granular level. Just because we don't do one or two things doesn't mean you can discard the other three or four things that we DO do. We don't consider batting with runners on base to be different from bases empty. If you want that, that's fine. Use RE24. That's WHY it exists! It's to give folks a choice. For things that we don't have at all, like scooping, feel free to add 2 or 3 runs to one fielder while ALSO subtracting 2 or 3 runs to another fielder. You will find that your ultimate conclusion is not going to change, not unless you have two players who are close to begin with.

I think you can say that for other positions as well to some extent. Relay throws, or anything that requires teamwork. To the extent that there is undervaluation of good fielders related to teamwork, it's +/- 2 or 3 runs.

This is a good question, and it's on my todo. I call this the "transfer process". The way OAA is built, it's based on first-touch. Overwhelmingly, the impact of a fielding play is determined by the first fielder to touch the ball. With 1B, they are often at the mercy of their fielders, whether receiving a throw, or a pitcher getting to the bag. Or even depending on other fielders to make a play in foul territory. That said, the overall impact of this will naturally be 0 at the league level, since all the 1B will be compared to each other. The league average 1B will naturally be 0 above average. The RANGE in this "teamwork" or "synergy" skill will likely be +/- 2 or 3 runs. I can tell you that other researchers have tackled this subject in novel ways and we end up with around that. So, yes, there's a skill, but it's a question of magnitude. The first-touch skill will always be much higher than the synergy skill.

Sure, there's a qualitative thing that happens here, and we can measure it. In the end however, the actual outcomes, the overall value, that is easily quantifiable, and it turns out: it's negligible.

I didn't see a question in that long post, so there is nothing for me to respond to. If you have a terse question, I'd be happy to respond.

We devoted a section of The Book to that. You can read about it in Amazon's Look Inside for free. Again, you asked the very boring question that I will give you a very boring answer to: of course it exists, because humans are involved. NEVER ask "if" it exists. What you must ask is: "to what extent" does it exist. What happens with protection is that the quality of the outcomes are shifted around, but that the overall impact ends up being negligible.

Thank you for the kind words. Because humans are involved, clutch is a skill. HOWEVER, that is a very boring question and a boring answer. Our job is not to say IF something exists. Humans are involved, so naturally it DOES exist. No, our job is to see the DEGREE to which something exists, the MAGNITUDE. The clutch skill, to the extent it exists, is much less than the platoon advantage for example (another human-based trait). It takes a loooong time to find the clutch skill. And by the time you find it, the player's career is half over. (And being human, it's not even clear how persistent it is.) Even if you are able to determine the clutch skill in time, its impact is so muted (like I said, much less than the platoon advantage) that it makes it NOT ACTIONABLE. So, in the end, what we have is: - yes, clutch skill exists - it takes a long time to find it - its magnitude is small - which makes it not actionable

Yes, you could definitely do it by run expectancy. It's a question of how simple or complex you want to make it. After all, you can do the same thing with HR and hits and walks and strikeouts too. Just because you make it more accurate doesn't mean you are making it better. There's definitely an argument to be made both ways.

Thank you kindly

First rule of Chat GPT: we don't acknowledge Chat GPT.