JotCast | Dan Szymborski FanGraphs Chat 1/16/25

Dan Szymborski FanGraphs Chat 1/16/25

Prev Page View Full Chat Next Page

Dan Szymborski

12:42

In-season stuff

PECOTA came out slightly before ZiPS did, but I had studied in-season/preseason projection interactions before basically anyone in public

Ghost of Eric Davis

12:43

Is there an injury history database that you can (or do) plug into ZiPS as an additional data point? Is it complete enough?

Dan Szymborski

12:43

I do have a database of generalized injuries

Biscuit

12:43

Who is the most interesting ZiPS lover affair so far? Combination of the biggest difference from other projections and also just overall visibility?

Dan Szymborski

12:43

Well, this year, the Cubs

I forget the biggest difference last year

The year before, ZiPS was in love with the Diamondbacks

Guest

12:43

any more team zips profiles this week?

Dan Szymborski

12:43

One more tomorrow!

Then we'll be down to just five teams

Dave

12:44

Do you publish percentile outcomes anywhere?

Dan Szymborski

12:44

Not systematically outside 80/50/20

Nate

12:44

is it you or the cats that decides ZiPS release order

Dan Szymborski

12:44

=rand()

12:45

Dave

12:46

Do you generally just add code ontop of code or do you rebuild Zips from the ground up?

Dan Szymborski

12:46

It's a math shanty town

Gram

12:47

Zips has a decent projection for Jerar Encarnacion. Does it pretty much just ignore his Marlins time?

Dan Szymborski

12:47

Shouldn't be ignoring anything!

Idiotic Failson

12:47

Hang on a minute - you just said you didn't find anything useful in height/weight data. Does that mean that there might actually be some sort of 'market inefficiency' for the wrong looking prospects (too short/skinny/heavy/whatever)? And does that work for pitchers and hitters?

Dan Szymborski

12:47

It could be

the larger problem is the data just frigging sucks historically

12:48

BR Kruk is smaller than Ichiro

and that's not shade at BR - the data just sucks

Projections

12:48

Are there any players that consistently outperformed or underperformed their ZIPS projections throughout their career? Excluding injuries

Dan Szymborski

12:48

I don't have the EXACT players off hand

I mostly look for systematic errors, which are really hard to find

the year-to-year miss r^2 is very low (under 0.01)

Burnes'd

12:49

Jackson Merrill seemed to come out of nowhere last year. Was that peak-level performance from him, or did he demonstrate that he's better than projections initially thought>

Dan Szymborski

12:49

Maybe not peak

but as a group, one shoudl expect last year's ROY candidates to take a step back

not because of any mythical sophomore slump

it's just that the top achieving of any population are likely the ones that had the highest percentil eoutcomes

12:50

Like a 110 win is far more likely to have overperformed than underperformed

https://twitter.com/DSzymborski/status/1876409772372861383

Dave

12:52

Which of the Jacksons does Zips prefer over their careers - Chourio Holiday Jobe or Merrill?