JotCast | Dan Szymborski FanGraphs Chat

Dan Szymborski FanGraphs Chat - 1/7/2021

View Full Chat Next Page

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

Dan Szymborski

12:02

And welcome to 2021!

IT STARTS NOW, EVERYTHING ELSE WAS 2020.2

Tony Schiavone

12:02

Welcome to one of the greatest chats in the history of our sport!

Dan Szymborski

12:02

I am in the top 100% of all chats.

See! You! Later!

12:02

What do you think is the biggest area for potential improvement to ZiPS? What are you focusing on improving with the system?

Dan Szymborski

12:03

Injuries, though that's harder to actually *work* with

Mostly adding more percentile stuff (hitters will be upt his year)

improving model for playing time as in projecting when a player will lose time

doing more Gameday mining, which has paid off well in defensive data for minor leagerus

Stevil

12:03

How many GMs tried to sneak players through waivers yesterday amid the chaos in DC?

Dan Szymborski

12:04

BenZ

12:04

Do you think we get a 162-game schedule this year?

Dan Szymborski

12:04

I'm cautiously optimistic

Zach

12:04

Who's you're number one player that you expect will absolutely outperform their ZiPS projection

Dan Szymborski

12:04

Gotta wait for the preseason piece!

Join Dark Order

12:05

Did you see the responses to the passing of Jon Huber? I thought that AEW and even WWE (the company itself and the individual wrestlers' personal in-ring references) did a nice job. I hope the Dark Order, for kayfabe reasons and for real life reasons, gets some time off to regroup before returning to action.

Dan Szymborski

12:05

Yeah, it seems they did a good job

12:06

Was nice to see the belt itself retired

12:07

I'm torn on storyline changes because I can see an argument both ways

12:08

For example, as a counterargument though I didn't *know* Lee, I'd imagine he wouldn't want his death to hurt company momentum. I wouldn't want FanGraphs to temporarily retire in-season projections for 2021 if something happened to me!

dunno what the right answer is

Drew

12:08

why do you keep on using my Netflix account after I asked you to stop?

Dan Szymborski

12:08

So that all your recommendations are Young Sheldon.

Scotty

12:08

ZIPs has Kyle Hendricks and Kwang Hyun Kim with same expected WAR in 2021... that just doesn't pass the smell test for me. Kim is older, has far less MLB experience, had multiple injuries in 2020, pitched less innings in 2020, etc. yet they are projected to be worth the same? I don't get it.

Dan Szymborski

12:09

Less MLB experience affects the error bars not the projection itself.

And there are some bad indicators with Hendricks as well

12:12

A *dip* in K rate as league increases for example, isn't a great sign. And ZiPS sees upside in Kim's K rate -- his hit from KBO was a bigger drop than you'd expect

In any case, I don't change projections, period.

Matt

12:12

The Blake Treinen signing isn't the biggest deal in the world, and he's a good pitcher, but it definitely feels a little bit weird, right? Given the RP market, Brad Hand clearing waivers, Archie Bradley's non-tender, etc., I'm confused why the Dodgers made this move.

Dan Szymborski

12:12

Well, their competition has been more aggressive since Hand was on waivers

BirthdayBoy

12:12

Clearly MLB's "hot" stove has been affected by climate change...

Rick

12:12

How many (and which) Rays prospects do you expect to see in the majors in 2021?

Dan Szymborski

12:12

Too early to say without knowing twhat the season looks like

Derek

12:12

Dan, can you use ZIPS to predict two-strike performance? Juan Soto's approach with two strikes is dramatically different from his approach with zero or one strikes. I'm thinking about how to answer the question "should he use the two-strike approach in more counts, jettison it altogether, or keep doing what he's doing?" I realize the answer is probably not to mess with what is already so good. But I think it's an interesting question, and you can't just look at Soto's two-strike performance. As a starting point, I think you need to compare his actual two-strike performance with what you would expect him to do with two-strikes based on a host of inputs (ergo, ZIPS). What do you think?

Dan Szymborski

12:13

ZiPS isn't really set up for this exactly

BirthdayBoy

12:13

Is this the year that ZiPS decides to initiate the AI takeover?

Dan Szymborski

12:13

I'm not smart enough

druidiful

12:14

Dan! just wanted to say "Hi" since I haven't been in one of your chats in a while.