JotCast | Dan Szymborski FanGraphs Chat

Dan Szymborski FanGraphs Chat - 3/31/22

Prev Page View Full Chat Next Page

Dan Szymborski

12:45

Now, there's naturally more *variance* in certain teams.

12:46

But we've ironed out the kinks!

Morbo

12:46

Guess that this is something you might not be able answer but is everything alright at fangraphs? The the pace of articles this spring seems to be...muted vs past years with essentially zero spring training coverage

Dan Szymborski

12:46

The problem is that the timing of the lockout has messed up the schedules for everything.

For example, you'd normally have four articles from me this week on breakouts and busts.

12:48

But the late free agency and the timing of the season - and a number of ZiPS-related tasks that cannot simply be done until a certain part of the spring - has thrown everything into chaos

The changing of timing and less knowledge from what teams are doing with prospects causes problems for Eric and the prospect bunch

And the lack of knowledge about what team WILL actually look like has hurt general preview type stuff

Guest

12:49

does zips take into account draft status for young prospects? you wrote in the zips top 100 introduction that those rankings are only based on pro ball results. I have a hard time believing Marcelo Mayer's FCL performance alone was enough to get him to #23?

Dan Szymborski

12:49

ZiPS does not.

Lonnie

12:49

Blue Jays C with the most ABs this year even if some come as DH?

Dan Szymborski

12:50

The question is - is the label of a player more static or dynamic?

If I say which Yankees great has the most home runs, does not mmean career homers of a player who WAS a yankees great, or only WHILE he was with the Yankees?

Cynthia

12:50

Recently, it seems like the best position players from Japan have struggled more than expected (i.e. Yoshi Tsutsugo, Shogo Akiyama). Since ZiPS uses player comparisons for its projections and given the small number of position players from Japan, do they make a material difference in future projections?

Dan Szymborski

12:51

Not so much simply because there's also a lot of incoming and outgoing on the minor league level

Some players have done better too!

But the failures look worse because the general variance is higher

Erik

12:52

A follow-up to my question on flat rosters: part of the idea there involves not just injury but performance. Projections assume an average number of PAs for everybody and an average level of performance. In reality, the guys who outperform their projections get more PAs and the guys who underperform theirs get less, if you have the depth to do that. That strikes me as harder to account for than injuries.

Dan Szymborski

12:52

That *is* a problem, of course.

There are a lot of soruces for error that you can't do much about

Minor league being a thing helps.

Since the full-timers that struggle will frequently get more time at a lower level

Morbo

12:52

"Evaluating a manager is a very difficult thing to do because data for it generally sucks." Unless it's Mike Matheny, then you just subtract 5 wins ±2 and you're good to go.

Dan Szymborski

12:54

One of my favorite quips is that Matheny's bullpen management is basically throwing a hundred spiders into a kindergarten class

Kyle

12:54

Have you taken a look at the article by Jacob Herlin on FantasyPros about the most accruate projection systems for 2021? ZiPS was 14th out of 17.

Dan Szymborski

12:54

I have not, but I'm surprised it would be *THAT* high. ZiPS is specifically not a playing time predictor, while fantasy projections absolutely *ARE* playing time dependent.

12:55

Ariel Cohen's done work on this with ZiPS which you can see here on this very website!

12:56

12:57

If you draft, say, Langeliers high based on his projection in fantasy, you're not using ZiPS correctly!

12:58

That's why we started offering a ZiPS version that matched DC playing time

Sirras

12:58

A friend of mine described skyline chili as "an acquired taste but good actually." What is the appropriate way to end this friendship?

Dan Szymborski

12:58

I think you should help stricken friends rather than abandon them!

Morbo

12:59

Personally I think Dan's closest comp is a young Brad Pitt (if this doesn't get my comment posted nothing will)

Big Fan

12:59

Why do ZIPS offensive projections change based on a player being at different positions? I noticed this with Trevor Story as a 2B versus as a SS.

Dan Szymborski

12:59

Slightly different group of comps. Not using positional comps *has* made projections worse.

Guest

12:59

The LF rankings suggest that Dom Smith is blocked at DH by Cano. Is that just guesswork at this point?

Dan Szymborski

12:59

You'd have to ask Jason, but I'd guess yes

Philip

12:59

What’s a team that you are confident beats the ZIPS win total and a team you are confident loses more than ZIPS projects? Don’t count tanking teams as obviously they will sell off parts at deadline

Dan Szymborski

1:00

I don't think they'll match last season or come close, but I think the Giants will be create inseason and beat their ZiPS

I think Reds fall short - it's likely they're disappointing any resume dumping this summer

Connecting…