JotCast | Ben Clemens FanGraphs Chat

Ben Clemens FanGraphs Chat - 8/25/25

Prev Page View Full Chat Next Page

Ben Clemens

2:13

but obviously people get hot and cold, and not just in 'batted ball bounces' ways

as we speak, I'm looking at hot streaks for a prospective article, believe it or not

Oaktown Blues

2:13

I think Darell Hernaiz is likely to be a better SS defender than Jacob Wilson - any chance the A's would bump Wilson to 3B (where they already have a hole) to fit Hernaiz at SS? Or is that getting too cute?

Ben Clemens

2:14

Probably not. They seem very committed to Wilson there. I think it would make sense if they're pretty confident Hernaiz is part of the future, but.... I don't think they are. And if there's some chance you're optioning him or otherwise taking him off the roster, then keeping Wilson at SS gives you more flexibility

CFH

2:14

Longshot, but if Cal Raleigh hits 60 bombs does that get him the MVP?

Ben Clemens

2:14

surely, right?

Opifijikl

2:15

Do you or do any teams do much Bayesian analysis when it comes to player analysis?

Ben Clemens

2:15

I mean, I'm guessing teams do

2:16

I am coding some Bayesian stuff in my hot streak model right now, for example

I think it's a really powerful way to think about the world

bringbackpologrounds

2:17

FONSECA IS NEXT

Ben Clemens

2:17

exciting

Guest

2:17

Sam Miller has written a lot this year about true talent & variation & these thoughts on Pebble Hunting

Ben Clemens

2:17

Yeah, a good reminder to read Sam Miller if you aren't

Joe

2:17

The Twins the last 2 years have at times alternated Ryan Jeffers and Christian Vazquez day by day. Adley and Basallo different tiers of player, but do you like that in theory for those two players and the Orioles?

Ben Clemens

2:18

I do, yeah

when they both hit well, it seems like a pretty good plan to me

Mariners Fan

2:18

ZIPS has Cal Raleigh hitting 28 homers next year. I’m not expecting 50+ again (or 155 games, for that matter), but that seems way low. That’s less than 2023 and 2024, and surely he’s shown improvement this year. Is this just a case of the models being way too conservative, or do they not buy his improvement?

Ben Clemens

2:18

oh, that's just because our 'zips 3 year projections' run in the preseason and don't update during the year

2:19

but also yeah models being conservative

AL Central Casting

2:19

I'm definitely on the side of labor, but let's say the owners were willing to ensure a specific percentage of revenue goes to players - larger than what they get now - in exchange for a cap/floor system. (This would presumably require verification of the revenue numbers, so set aside skepticism about that and similar issues for now.) What is the argument against it? If the players as a whole get more than they get now, and there is less payroll disparity between teams, is that a net benefit for the league?

Ben Clemens

2:20

I mean, I think that system would be the platonic ideal

I think that revenue verification is very difficult for a sport like baseball with so much local revenue, but if basketball (and their very similar RSN network) was able to make it work, maybe baseball can

2:21

I'd love a cap/floor system, it's just an excellent way to handle salaries in my opinion

Jeremy

2:21

Follow-up to the talent vs. luck/timescale thing: is the distribution of hot/cold streaks for individual players any different than you would expect if everybody's true talent was constant within the season? If in fact "true talent" actually goes up and down on a timescale of weeks or months, then you should see more, and longer, hot and cold streaks then you'd expect in a world where everybody's true talent is constant within a season.

Ben Clemens

2:21

So I've been approaching this from 'hot streak persistence' as in whether players who are 'hot' are just more talented and thus stay hot in the next sample

or like, they're locally more talented, they're on a hot streak!

2:22

the answer is..... i dunno man, still doing the analysis. I have to think more about what the distribution "should" look like if there's no change in talent

Insert Witty Name Here

2:23

I’m a man of reasonable intelligence, and I agree that Bayesian stuff is super powerful how we think about the world, a phrase I heard about our brains is that it’s simply a Bayesian Prediction Machine that never turns off. Curious to know more about how you think of it?

Ben Clemens

2:23

Yeah, I think I agree with that

2:24

if you're just working off the cuff in generalities, obviously you are being Bayesian. If you thought your team sucked yesterday and they played well, you're updating your prior about them today, not starting fresh

I'd more say that I like to give some empirical weight to prior expectations when I model stuff, because I think that when I'm doing predictions it's too easy for me to forget what I thought I knew in the past and get obsessed with 'starting fresh'

Dirt

2:24

Wait so do Barry Bonds' wRC+ numbers not include his IBBs? Or is that just since the new rule?

Ben Clemens

2:25

yeah, correct. Although his WAR is still great b/c he gets credit for those PA's at his seasonal wRC+, which was always comically high at that point

Colton

2:25

At what point does a hitter's passivity at the plate become a detriment? Someone like Nootbaar who is a very solid in zone hitter seems to put an inherent cap on his potential offensive production because he's too passive. Have you ever looked into a way to categorize this type of issue guys will have?

Ben Clemens

2:25

BP's SEAGER method is pretty good at this, I think

it basically just rates swing decisions as good or bad based on count/production

2:27

Nootbaar never does all that well there because he gives up too many good swings in pursuit of avoiding bad swings

so yes, it's definitely quantifiable, I haven't looked into it recently

Todd

2:27

Why are IBB’s excluded from wRC+? I guess the assumption is that they don’t have much predictive value? Feels like only good hitters get IBB’d, thus it’s a “skill”. Not to mention, you can work a 3-0 count before they decide to IBB you, and it still gets recorded as such

Connecting…

Message History

On Off

Your current subscriptions

None

E-mail Address	Chat Owner	Status	Remove

Enter your email address to be notified when Ben Clemens hosts a chat