JotCast | Dan Szymborski FanGraphs Chat 4/3/25

Dan Szymborski FanGraphs Chat 4/3/25

Prev Page View Full Chat Next Page

Dan Szymborski

1:00

Gibson

Not because he's the best, but he's extremely durable

Guest

1:00

at 6', 10 years ago i went from 215lb to 185, but then at some point I got back up to 225. Then lost two chunks of 30, one over 5 years (more exercise and fewer sweets), the next over about 5 mos. (concerted effort and calorie counting). been holding about 165 for 11 months now. you got this!

Guest

1:01

Dan this may be a noob question but what rate does wOBA represent? I understand the math that goes into it and I know that it’s an accurate indicator of how good a player is at producing runs, but it’s a rate stat. The denominator reflects the number of plate appearances (ABs?) that a player gets. Is it the rate of “run” creation per plate appearance?

Dan Szymborski

1:01

think of it as a wRC rate state scaled to look like it's OBP

There's no real unit there

zips enjoyer

1:02

you don't think limiting yourself to only what you can replicate (maybe less effectively in places) hurts the project, even if it maintains some sort of "purity".I guess then is the goal of the project to be maximally yours or maximally accurate

Dan Szymborski

1:03

We're talkign slivers of accuracy. I just don't like profiting on the work of others

and I wager that most things that anyone can put in a statistical projection, I get there sooner or later

Because doing this also answers WHY questions, that I wouldn't get

1:04

I want data to CAPTURE WHY someone is a 55 FV prospect in a scout's eyes

and every bit of new data we get gets us slightly closer to that goal

1:05

If I just threw Eno's work into ZiPS, I believe ZiPS would be *slightly* more accurate, but not in a meaningful sense, because we're measuring breadcrumbs with rulers in improvements now

And when a model requires the inputs of others to function...what happens when you lose that input? Or it becomes less accurate in a way you cannot detect?

Tiger Fan

1:05

I fully respect wanting to know the why. Too many kids be cheating through homeworks without knowing why their stuff works 🫡

Dan Szymborski

1:07

There's a reason I GENERALLY prefer supervised to unsupervised learning in modeling. Know the WHY is extremely important to me and a transparent why that has slight accuracy losses from on opaque one is a tradeoff that I'll take

1:09

I think this compulsion comes from the way I was raised. There's an almost pathological knowledge über alles vein in my family, no doubt from my maternal grandfather's influence. Learning stuff was always more important growing up than being nice, or being successful, or anything else. I read very early not because anyone in my family pushed me, but because even as a little kid, I had this programmed compulsion to know EVERYTHING before I die

1:11

Doing projections wasn't some plan for commercial success or even a career. I just wanted to KNOW things. I have a lot of weird veins of knowledge in my head; I subscribe to five academic journals, I've read every SCOTUS decision since 1973, I still read about 150 books a year.

So there's probably some screw loose inside my brain!

Guest

1:11

can any of these learning principles help me with my basketball jump shot lol

War2d2

1:11

Catching up to the beginning of the chat, re: the Q about best player to be traded: Do you think Sandy Alcantara won’t be traded, or do you think Brandon Lowe is/will be better than Alcantara?

Dan Szymborski

1:11

actually totally forgot Alcantara, who would be the correct answer lol

Logan

1:12

As a lawyer, I have to say that subjecting yourself to full-text SCOTUS opinions is a character red flag, Dan...

Dan Szymborski

1:12

Lots of my family has been lawyers. I have way far too much knowledge that I will do absolutely nothing useful with lol

Guest

1:12

If you had to make a living gambling using only ZiPS and baseball could you do it?

Dan Szymborski

1:13

I suspect so, but you have to eke out a lot of small gains over a LOT of bets becuase the house take is a pretty big deal

1:14

If I had to make a living gambling, I'd want to do it on the side of the house!

1:15

is it just me or are we having some real catcher batting success early this season? it's more fun when they are fresh and less so when their knees grind to dust and they're flailing around in september :(

1:15

Everyone understandably has been on Devers' case in this small sample, but your being down on Casas looks awfully prescient in the early going.

Dan Szymborski

1:16

I don't remember if I talked about it, but I find Fenway a poor environment for Casas

1:17

A lefty slugger has to either absolutely CRUSH pitches or be metahuman and dinking it around to the RF jut

1:18

And playing in Fenway incentivizes contact hitting. It's a really solid BABIP park

cat latos

1:19

What do my cat friends to?

Dan Szymborski

1:19

Justinian is sitting on my office couch half asleep. Mercutio and Constantine were awkwardly sharing the heating pad upstairs when I came back downstairs for chat

1:21

Cassiopeia/Cassie is in the little cat cave that I picked up last night, which has pretty exclusively been her base of operations.