JotCast | Dan Szymborski FanGraphs Chat 5/1/25

Return to

Dan Szymborski FanGraphs Chat 5/1/25

Prev Page View Full Chat

Oddball Herrera

1:02

If you believe it's generally possible to create a computer that could simulate the universe, then the odds are you're in a simulation, because simulations would outnumber realities by a considerable margin

Dan Szymborski

1:02

Yup, that's exactly my take

and not just if it's generally possible, but if it is EVER possible, in the countless years until the age of planets is over

Speaking of Video Games...

1:03

In MLBtheshow it is also strangely easy to barrel with JJ Bleday. Like almost every time I make contact.....

Dan Szymborski

1:03

I shoudl check my MLB the show stats after the chat (I'll answer this in twitter)

Opifijikl

1:03

When testing/working on ZiPS variables, have you found any that seem to conflict with each other?

Dan Szymborski

1:04

Not REALLY, though I do find that something new isn't quite as useful as hoped becuase you were capturing it more indirectly elsewhere

like so far, swing speed data hasn't really change things much as ZiPS was already capturing the results of swing speed

Jason

1:04

Where do you get the ideas for your AI posters which you post on X/Twitter and are you getting help for your issues? Just kidding the posters are wild...

Dan Szymborski

1:05

I'd oddly wired, both nature and nurture

Daniel -

1:05

Follow up: I made a little program in R that compares a players preseason projected WAR / PA rate * their actual PA and sees the difference between the war they have in their career and the war they'd have if they played to their projections every year -- and Judge has the largest difference (close to 20 war more than the projections throughout his career). Do you think the projections have been weirdly anti judge, especially over the past ~4 years?

Dan Szymborski

1:05

I think there's a bit of selection bias becuase the best players in the league are generally going to be the ones that beat expectations

1:06

if someone hits .350, you're SUPPOSED to miss really low

Rajen

1:07

Hi Dan, I was wondering if you had some insight into why full-season defensive value across MLB sums to a large negative total (last year around -590 runs). Total offense sums to 17 runs (functionally zero), so I'm wondering why the total defensive value isn't that after. Does that also mean we should think of a player with -2ish DEF (600/30 teams/10 starters per team) as an "average" defensive player?

Dan Szymborski

1:08

It seems to be a relic of the positional adjustmejnt values, which add up to -1.9 or something if you assume equal PA on everyone

1:09

I suspect the result is in there because it's an offensive adjustment to a defensive position and positions with steeper defensive adjustments tend to hit more and thus, get more PAs

But when it comes down to it, I'm a guy who is less concerned what numbers are, than with what numbers *mean*, so it doens't obther me too mcuh

And on that poorly typed note, I'm going to head out for another week

1:10

And sorry archibaldyeats for me being me.

Connecting…

Message History

On Off

Your current subscriptions

None

E-mail Address	Chat Owner	Status	Remove

Enter your email address to be notified when Dan Szymborski hosts a chat