JotCast | Dan Szymborski FanGraphs Chat

Return to

Dan Szymborski FanGraphs Chat - 6/5/25

Prev Page View Full Chat Next Page

This chat is available to members only. Please log in to participate.

Dan Szymborski

12:44

I'm not at that point yet. I still think of him as a quality inning eater

namenamename

12:44

Do either Coby Mayo or Heston Kjerstad work out for the Os? Seems like Mayo is totally lost, although still quite young, but Kjerstad is a few weeks away from being a decent role player

Dan Szymborski

12:44

I think they'll work out still

DJ Tofu

12:45

What do you think is wrong with LaMonte Wade? What are some players with similar steep and rapid drop-offs in performance?

Dan Szymborski

12:47

The BIG dropoff in hard-hit percentage is concerning

12:48

Even with some BABIP bounceback, which is inevitable, the lack of power this year is supportable

I think the downgrade in the projections have him right

But Wade's problem with the Giants is that he's a role player/borderline starter who HAS to hit to have value

drplantwrench

12:49

do you think fWAR should include xBA type stats?

Dan Szymborski

12:49

Guest

12:49

is ZiPs too hard on Ben Rice defense (-12.6 ROS)? Or is this just a function of DH that I'm misinterpreting ?

Dan Szymborski

12:49

ZiPS is certainly twoo hard on the defense

too

(there's some position mixed in there)

12:50

but he had absolutely oddly poor numbers in ZiPS probabilistic method of evaluating minor league defense PBP

he should have a much sunnier defensive projection this fall

drplantwrench

12:50

i guess i should follow up - why dont you think fWAR shouldnt include xBA type stats?

Dan Szymborski

12:51

Because we're going too far into expected value rather than what actually happened. In a different way than FIP with pitching

People sometimes describe RA9 WAR as what actually happened, but that's a bit different because in that case, it's what happened to the team. The pitcher has to share responsibility with defense.

12:52

There's not the same issue with an individual batter

Now, we do have to figure out batter vs. *opponent* valuation

But batters have a mroe repeatable tendency to perform better/worse than their expected stats than pitchers do in their FIP vs. ER stats

12:53

Which can be seen in how ZiPS *mixes* the expected stats with the actual ones as a better predictor of the future than just the expected

Spider

12:53

Why are there some players who ROS ZiPS has for negative projected WAR but ZiPS Depth Chart has them for positive projected WAR? I understand they have different playing time projections but the rate stats are the same, so wouldn’t they both always be either negative or positive for each individual player?

Dan Szymborski

12:54

Can you give an example? I don't actually run anything for this part as I have little to do with the depth charts (unless both Jason and Jon get ebola or something) and I don't do the FG implementation of the depth charts into a model

Insert Witty Name Here

12:54

Thought experiment. If you watched every single game, every single pitch of one team for the entire year, how close would you be able to predict their WAR?

Dan Szymborski

12:54

I expect very poorly.

12:56

I like to think I'm an insightful observer of things, and I have watched as much baseball as anyone on the planet. But I'm subject to the same cognitive biases as anyone, even if I was in person for every game (violating space-time constraints) and could see stuff that the cameras don't show you that are necessary to evaluate defense

Plus, the human brain is really bad at this kind of thing.

Spider

12:56

For example, Tyler Phillips is projected -0.1 WAR by ZiPS depth chart and 0.3 WAR by ZiPS. Different amount of innings, yes, but how can this be correct with the same ERA/FIP/etc.?

Dan Szymborski

12:56

Hmm

12:57

Ah, it's because the different roles.

FG is still a bit awkward at players in different roles than projected. Still trying to figure out a technological solution for that

drplantwrench

12:57

id argue that xBA is what actually happened to a certain extent, but also completely understand it too. it would be fascinating to see how an expected-based WAR (xWAR) would fare over different sample sizes against the other two WARs. also i just want more types of WARs to confuse people

Dan Szymborski

12:58

I think zWAR is a good middleground between fWAR and bWAR approaches to run prevention (ZiPS WAR gives more credit to the pitcher the longer they out/underperform their ERA relative to team defense expectation). But the *value* of a very slightly more accurate WAR is more than cancelled out, I believe by the negative value of creating more confusion

Smokin' Jim Leyland

12:59

This is probably more of a question for Jay Jaffe...but assuming that the Cubs and Reds are out due to his minimal time spent with those orgs - which cap do you think Nick Castellanos wears at Cooperstown, Tigers or Phillies? And will he go in the "Nicks" wing of the hall, alongside Mick Markakis?

Dan Szymborski

12:59

Phillies

1:00

If he somehow had a Hall of Fame career, it's probably going to have to happen from an explosion of offense coming SOON, so it would be more PHillies-related

And if that explosion doesn't come before 2027, it just wouldn't come - I doubt anyone's giving him a significant contract after 2026

Vacay

1:00

What is your favorite type of vacation? Mountains, beach, fun city, road trip…?

Connecting…