JotCast | Dan Szymborski FanGraphs Chat - 7/24/25

Return to

Dan Szymborski FanGraphs Chat - 7/24/25

powered by

Prev Page View Full Chat Next Page

Devil Ray Jay Johnson

12:42

What would the Rays' best deadline move be?

Dan Szymborski

12:43

New owner Patrick Zalupski is approved by owners and then he rips off his face mask and it's Mark Cuban and he announces that HE'S here to spend money and chew gum and he's all out of gum

laughing

3

Lowe Distribution System

12:43

Anyone gonna trade for Chris Sale? Should be back in time for playoffs, potential playoff difference-maker

Dan Szymborski

12:44

I suspect no

L

12:44

I’ve been thinking this for a few years, but Sasaki falling flat on his face confirmed it for me: Dodgers pitching development is incredibly overrated right now. Most of their best pitchers at this point were wholly developed by other teams. Their supposed specialty, fixing up random cheap scrubs? Every competent team is doing that now. Some teams, like the Padres and the Mets, are even successfully turning them into front-line starters. Meanwhile, the Dodgers have been doing the literal opposite, taking expensive dudes like Scott or Yates and breaking them. They’ve been having trouble fixing their own developed pitchers too (Miller, May, Gonsolin), which is actually a pretty big problem, since none of them have had sustained success. The most frustrating part is that the Dodgers are fully aware that they’ve fallen behind. Why else would they spend so much on pitching?

laughing

1

Dan Szymborski

12:44

I dunno if *Sasaki* confirms it. They didn't really *develop* him after all, and he's injured too

bam bam

12:44

Sebastian Walcott more than doubled his 5-year WAR projection in the new top-100 update. Have his new K-BB% numbers changed how ZiPS views his profile as a hitter, or does it seem more like it's just another years worth of above average production at AA putting weight on the scale?

Dan Szymborski

12:44

ZiPS gets a lot less scared once guys cross the A-AA barrier, which historically seems to be the biggest hurdle

Cat Latos

12:44

How do my cat friends do?

(I saw you missed me the other week)

Dan Szymborski

12:46

Last I was upstairs grabbing something to drink (about 11:15), Cassiopeia was sleeping in the high perch of the cat tree, Constantine was sleeping on a blanket that was left out on the couch, and Mercutio was staring at the empty dry food bowl in the kitchen in meatloaf mode, unable to comprehend why he didn't see the thing he wanted.

Justinian's on my lap at the moment

or more precisely, half on my left desk and half on my left leg while my keyboard is balanced across my right leg and my right desk

12:47

(I sit in a recliner parallel to two desks)

Look At This Sotograph

12:47

Does ZIPS take into account things like pulled fly balls, that we know can make xwOBA underrate a hitter?

Dan Szymborski

12:47

yes. pull/spray tendencies are in there in ZiPS

and I even have dynamic park factors based on pull/spray tendencies, which is why Isaac Paredes always gets a much better projection than his peripherals think about him

I am the Apple

12:47

How do you see the relationship between public-facing analytics (like at FanGraphs) and the proprietary work being done inside MLB front offices? Is the public catching up, or is the information gap widening?

Dan Szymborski

12:47

It's...complicated

12:48

I think the public gets a lot of things wrong about how teams use data

the public seems to think that teams spend a lot of time doing stuff like their "own" WAR or other normal performance metrics

12:49

The edge that individual teams have is that they have special access to their own players, and lots of *micro*level ways to evaluate the talent

Any kind of data that the whole league can use, everybody is going to knwo about and everybody is going to get to use, sooner or later

Jeremy

12:49

For how long, and by how much, would the Brewers have to consistently beat their projections before you'd consider it a systemic error on the part of the projection system? Evidence that the Brewers have some secret sauce that the projections don't account for? After all, if you were to replay (say) 7 consecutive seasons a million times, I'm guessing that in some decent fraction of those million replays there'd be a team that beats its projected win total 7 consecutive times.

Dan Szymborski

12:50

As I noted above, I think the problem with the ZiPS projections has been the Dan part where I have to determine who plays rather than the ZiPS part

which isn't super suprising because ZiPS is pretty good and Dan is pretty dumb

I am the Apple

12:51

Which player in history do you most wish you could see their Statcast outputs?

Dan Szymborski

12:51

I want to see those Babe Ruth exit velocities

or get some pitch data on a spitballer

maybe Burleigh Grimes is the answer

Bob Myrick

12:51

Do you know specifically how the Mets have caused the pitching velocity in the minors to spike so much this year? Are these methods known to everybody and some orgs are just better at teaching them?

Dan Szymborski

12:51

*That* is something I do *not* know

Bob Myrick

12:51

How much would you offer to the Padres for Cease if you were a team with a realistic chance at a World Series this year? Should the Mets be giving up 4.5 years of Mark Vientos for him?

Dan Szymborski

12:52

That would be tough for me as a rental

Look At This Sotograph

12:52

The Orioles' front office hates trading it's young guys, yet also hates locking them up in deals. What's going on here?

Dan Szymborski

12:52

Conservatism

Jeremy

12:53

Re: not just relying on cross-validation, why not? I think I know the answer, but I'm not 100% sure.

Dan Szymborski

12:53

A few reasons. One is just my excessive conservatism.

12:54

And maybe not excessive - the gain from a marginal improvement of ZiPS is much smaller than the consequences of breaking ZiPS

But also, cross-validation of an older model assumes that things that *are* true over a certain time period may remain so

Connecting…

Register

Message History

On Off

On Off

On Off

On Off

Your current subscriptions

None

Register now to continue managing your subscriptions

E-mail Address	Chat Owner	Status	Remove

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