Dan Szymborski FanGraphs Chat – 5/30/24

Dan Szymborski: It’s a chat!

Red: hey Dan! Can you put in a good word with the site team to add a date range function for minor league stats? Would be an awesome QOL improvement!

Dan Szymborski: I keep a bit of a list of things to bring up when we talk stuff at staff meetings, I could add it

Dan Szymborski: I’m pretty sure from the raw data we have, we at least have the ability

John M.: Have you ever tried to make any kind of projections about how Negro League players would fare if they were in an integrated MLB? Or how MLB players would fare if the leagues were integrated then?

Dan Szymborski: I’ve tinkered a little, but nothing substantive

Dan Szymborski: Naturally, stats for all players would go down in a much more competitive league.

Dan Szymborski: MLB would never do it because of the players it pushes off the leaderboards, but on a fundamental level, segregation-era records should probably all be considered sketchy, similar to a lot of 19th century stats

Dan Szymborski: My personally preferred leaderboard change — which would have pleased exactly nobody and displeased nearly everybody — would have been for 1961 to be considered the start of the modern era, not 1901

Dan Szymborski: The fact is, segregating players in baseball wasn’t just a horrible evil, it was bad for the baseball ITSELF. That’s a definite secondary thing, but it is a thing; both leagues were basically high level minor leagues because of the dilution of talent

Dan Szymborski: There’s a very real argument that Tony Gwynn is the all-time major league BA leader at .338

Billy: Real talk – are the Royals contenders or not?

Dan Szymborski: In terms of beign really really good? Not really. But they have lots of wins banked and a .500ish team with a bunch of extra wins banked is a contender

Dan Szymborski: Because anyone who makes the playoffs is a contender

Dan Szymborski: If MLB added two playoff spots for the two worst teams in baseball, then the Rockies and White Sox are contenders

Key Flaw: If pitchers and hitters don’t really have an effect on BABiP (to simlify at least), but batting average goes up each time a batter sees a pitcher in a game, and batting average is related to count on the batter, how does that work? Is it that there are definite things that are related to BABiP, but when you sum them all up it is too much noise to predict based on individuals?

Dan Szymborski: There are a whole bunch of key flaws here, largely missing the differences between BABIP and average

Dan Szymborski: BABIP is a *driver* of BA, but there’s other stuff too. Homers are part of batting average. And when talking about BABIP/BA interaction, so is strikeout rate

Dan Szymborski: Let’s given an example from last year

Dan Szymborski: First time through order last year, the league hit .246. The third time through order, .267

Dan Szymborski: But BABIP was only two points higher. from first to third time

Dan Szymborski: But there were 20% more strikeouts, meaning that a whole bunch of .000 BA plays became .290ish plays

Dan Szymborski: Also, that pitchers have a higher BABIP across the board when fatigued doesn’t mean that it’s an ability

Dan Szymborski: When the league hits .292 BABIP, that’s the average for all fatigue states. But if the difference is driven by fatigue, that’s a fatigue ability, not a BABIP

Dan Szymborski: And there still is a *slight* pitcher BABIP, usually involving numbers of this that you can count on your figners

Dan Szymborski: ALSO

Dan Szymborski: Nobody has said hitters don’t have an effect on BABIP. Hitters have a substantial, repeatable BABIP skill

Dan Szymborski: But it’s still quite volatile.

Joe: The Orioles have the fifth lowest team BABIP in baseball but the second highest hard hit rate. Which of these are likely to continue?

Dan Szymborski: Last I checked, the O’s had a zBABIP a good bit higher than their actual BABIP, so the latter

Bres-VERY low: Where does Sale-Grissom rank on the Betts for Verdugo/Wong/Downs scale of worst trades?

Dan Szymborski: Well off. I consider trades based on the typical reasonable results expected at the time.

Dan Szymborski: And there was certainly upside in Sale, but if you looked at every alternate universe where this happened, I’m pretty sure this isn’t the *average* result

Guest: the fact that FIP has a denominator of IP kind of is against the point of the statistic right? If a pitcher has a better defense, they make more outs, and the pitcher pitches more innings

I get no statistic is perfect but this seems like an oversight for the purpose of what FIP is supposed to measure. why not use something similar based on K%, BB%, HR%, etc?

Dan Szymborski: It’s probably not ideal, but I’ve run numbers on this before and there really aren’t consequences of note

Guest: We take it as a given that all teams should go around .500 in 1-run games, but is that really true? For example a hypothetical team that allowed 0 runs all year would be expected to win 100% of their 1-run games, while a team that allows infinite runs would win 0%. Is there a correlation between how few runs a team gives up & how good their record in 1-run games is?

Dan Szymborski: The no-infinite/no-negatives thing is an edge case that doesn’t make a real difference

Nick: Gunnar Henderson is on pace for the 3rd 50 HR season by a SS of all time. Outside of Ernie Banks and Arod there have only been 2 other seasons of even 40 HRs by a SS. Does Gunnar get to 40?! 50?!

Dan Szymborski: I don’t think he gets to 50

Dan Szymborski: I’d coinflip 40

Dan Szymborski: (without looking)

Transactions: Marcell Ozuna to the 15 day IL with back strain from carrying an entire offense for 2 months.

AL Central Casting: Is there any way for ZiPS (or any other system) to project how a player would bat from the opposite side of the plate (i.e., Ohtani batting RH)? What about for a switch-hitter who gives up switch-hitting?

Dan Szymborski: The former? No

Dan Szymborski: The latter, switch-hitters who go one side tend to have split differentials that regress toward typical L/R splits for their new choice

Connor Norby: Is there any scenario where I get playing time in Baltimore? Is my best hope to be traded this deadline? And is my friend (and 2B competitor) Jackson okay?? I’m worried about him! Debut seems to have messed him up, but maybe Elias and his cadre of scientists are just tinkering with their latest highly-touted prospect with luscious locks…

Dan Szymborski: Holliday will be fine

Dan Szymborski: 36 PA is nothing

Dan Szymborski: If the Orioles weren’t in a dogfight to be the AL East winner, Holliday probably doesn’t even get sent down

Dan Szymborski: And if the Orioles had worse depth, he probably doesn’t get sent down

Dan Szymborski: As for Norby, it’s hard to see him fitting in the lineup long term unless there are some trades

Zach: What’s with everyone’s overreaction of the Dodgers recent lull? They’re still on pace for 100 win season and lead baseball in oWAR and wRC+.

Dan Szymborski: Because everyone overreacts to unusually good or unusually poor stretches of play

Dmytro: Volpe is 8th overall by WAR. Is this sustainable, can he really be a top 10 player in baseball going forward?

Dan Szymborski: He’s playing Gold Glove level defense at short and a .800ish OPS line. You can wring out a lot of BABIP and he’s still awesome

Dan Szymborski: Top 10? That’s trickier as the competition is brutal

Dan Szymborski: But per 600 PA, ZIPS has him 14th in WAR for 2025, so it’s not preposterous

Dan Szymborski: (among hitters)

Sig Mejdal: Who will have a better 2025: Jordan Westburg or Jackson Holliday?

Dan Szymborski: I agree with ZiPS that Westburg’s improvement happening in the majors puts him a few runs over Holliday

AL Central Casting: Do you think we’ll ever see fully automated umpiring? Not just for balls and strikes, but also boundary calls and safe/out on the bases?

Dan Szymborski: I think MAYBE? But WAY after balls and strikes

Bufffy, the umpire slayer: How often are ZiPS (update) actually updated? It appears that the underlying numbers in ZiPS (ROS) or ZiPS DC RoS) have not changed. I base his on Jose Abreu’s wRC+ which is still at pre-season levels in the two later Projections but, correctly, much much lower in the ZiPS (updated) version. Also, can there be some type of indicator when these RoS projections are updated?

Dan Szymborski: I think there’s some weirdness in wRC+ RoS because of the offensive environment changing

Dan Szymborski: Abreu’s RoS OPS in ZiPS has dropped nearly 50 points, which ought to be more than a wRC+ drop of four points

How Bout Dem O’s: What’s your projection for the *key* year for this Orioles core? Like when would you truly go all in? Maybe next year… Jackson gets his sea legs, one more year of Grayson development, sorting out/improvement of the Heston/Cowser/Mayo jumble. But no Burnes…

Dan Szymborski: 2025 IF they replace Burnes

Dan Szymborski: If not, this might be THE biggest year

Guest: you guys have meetings?

Dan Szymborski: We have Slack meetings on Mondays when we have a lot of issues to bring up

Dan Szymborski: But things tend to run pretty smoothly. FanGraphs has about a dozen full-time employees, each with their own specific role they know well and have been doing for a long time, and we communicate quite a lot on Slack

Dan Szymborski: Now, for all I know, there’s a slack channel in which everyone is pissed at me and trying to get David to fire that goddamn Dan Szymborski, but if there is, I don’t know about it

O’s on the Interstate: As an o’s fan, should I be worried about Mullins and Santander starting slowly? Will/should they get replaced in the lineup?

Dan Szymborski: Not so much. I mean, they’ve already probably SURVIVED the worst stretch of Mullins season

Dan Szymborski: And if Santander struggles, the O’s have lots of depth in corners

Insert Witty Name Here: Im tired of the chilli/beans/hot dog convos. Yay or nay on scooping  mashed potatoes and applesauce together?

Dan Szymborski: OH dear god no

Ners luvver boy: Howdy! I always wonder what the process is for prospect rankings at Fangraphs. No shade, I would just love to see how my beloved mariners stack up and I wonder why the roll out is so gradual. Do the rankings account for current play given we’re well into the minor league season?

Dan Szymborski: You’d have to talk to Eric for the exact process! The rollout is because generally speaking, we don’t consider a team’s overall top prospects to be something that freezes on X day every year. Think of it as an ongoing, rolling covnersation

Plato: Projection systems like ZIPS seem to focus on being not wrong rather than right and are very conservative on estimates. Even 1/3rd into the season, the highest projected win total is 95. I think the median output of a team is different from maybe the realistic/ expected output as a median takes thousands of sims to find the teams expectations, but the season only happens once. Vegas odds for team win totals are higher for the best teams and lower for the best teams. ZIPs still has a lot of value and thank you for the great work, just a random person thinking of how to be right more than necessarilt being wrong less.

Dan Szymborski: But this is by design.

Dan Szymborski: You’ll always project someone flipping coins to have a 50% heads projection

Dan Szymborski: even though you KNOW it’s going to be off a ton

Dan Szymborski: Let’s take the AL East winner, which ZiPS projects WILL have 99 wins, on average

Dan Szymborski: OK, argue that a team winning 108 games is AS likely as 90

Dan Szymborski: Projections SHOULD have a smaller range than the actual results, or you’ve done something horribly wrong

Dan Szymborski: a player’s exact 1st percentile projected BA ought to happen just as often as their exact 50th percentile or 99th percentile BA projection

Dan Szymborski: Type =rand() into 30 cells in excel. Keep hitting F9 until all 30 numbers are between .450 and .550. Stop at that point. You’ll still be hitting F9 at this time next week.

JK: A way to help curb strike outs, redefine the checked swings. Raise the threshold for what qualifies as a swing IE ‘going around’. It would remove some whiffs from the game. Any substantive impact? More trouble than it’s worth?

Dan Szymborski: Interesting at least

Dan Szymborski: I think most crazy ideas should be auditioned

Skenesball: Why didn’t MLB insist on Skenes vs. Skubal during yesterday’s double header instead of letting them be staggered. Give people the SK Bowl!

Scott: Dan, thanks for the ZIPS standings update.  One question – it seems a bit odd to me that the Dodgers only have a 73% chance of winning the NL west.  They are up 6.5 games now and clearly have the best team.  At least one sportsbook is giving them -4500 odds.  Why is the division not more “in the bag”?

Dan Szymborski: Well, one important thing to remember is that Vegas is trying to maximize their return

Dan Szymborski: The thing is, the future is very uncertain

Dan Szymborski: and remember

Dan Szymborski:

Salvador Perez: Have you been following the conversation between Tom Tango and Bill James on my framing? Do you think we need to adjust how much value we put on framing in catcher defense?

Dan Szymborski: I think things should be questioned. I’m not convinced that it’s enough to give Sal Perez a huge boost

Dan Szymborski: But there ought to be no such thing as a set belief

Dan Szymborski: EVERYTHING has a nuance

Air Yordan: Please provide 1-10 confidence in a rebound for the following players: Riley, Goldschmidt, Machado, Springer

Dan Szymborski: 9, 3, 7, 5

Bighen: Mets obviously in disarray.  Trade Alonso, JdM, Severino? Get back C level prospects    Same for maybe Manea and Marte (if including $).  What else do they do?

Dan Szymborski: I think they just kinda hang around and do ltitle

Dan Szymborski: but we’ll see

Guest: boom, modern era starts in ’61. Maddux’s 355 wins are still “unbreakable” right?

Dan Szymborski: Under current usage patterns, that is all

Guest: David Fletcher pitched 5 innings last night as a knuckleballer. That’s all.

Dan Szymborski: I wonder if he bet on his inning count

turquoise: Robert Suarez has peripherals 2-3 runs higher than his ERA. Sell high in fantasy or ride til the wheels fall off?

Dan Szymborski: Always depends on what selling high looks like

Random: What team this century has lost the most pre-season projected WAR to injury?

Dan Szymborski: An interesting question, but not one I know the answer too

Dan Szymborski: and would be tricky to value

tseuG: Following up on 1-run games, if we looked at past teams’ seasonal 1-run outcomes, how strong do you think the correlation between bullpen quality and 1-run game record would be? Would you ignore overall record, or would it be illustrative to compare the outcomes of using/ignoring overall record?

Dan Szymborski: Almost none

Dan Szymborski: I’ve actually done the work there

Dan Szymborski:

Dan Szymborski: r^2 of 0.067

Dan Szymborski:

Dan Szymborski: Same results with just looking at closers or best X relievers

Dan Szymborski: Yeah, the years match weirdly, but I did them independently on Twitter

Dan Szymborski: and somehow didn’t get an article out of either

Random Colorado guy: Thoughts on Jordan Walker?  Is there any hope left?

Dan Szymborski: Certainly, but he can’t be in the team’s Plan As until he’s actually hitting Triple-A hitters well

Dan Szymborski: Walker now has 240 Triple-A PA and has a wRC+ of about 85

Dan Szymborski: that’s not a translated wRC+, that’s his ACTUAL wRC+ in the International league

Dan Szymborski: Even his last year of wRC+ 116 at Triple-A is HARDLY want you want to see from an offense-only player

Dan Szymborski: there’s still a lot of upside, but again, it’s speculative

12 to 6: gavin stone is 5-2 with a 3.15 era/3.50 fip and his fastball is consistently in the mid 90s, which is making his 65+ change look even better. any reason to think this isn’t real?

Dan Szymborski: Should actually see his K rate go up. Average contact rate shouldn’t lead to K rates thispoor

TomBruno23: Approximately how many innings of the Regionals do you think you will watch this weekend?

Dan Szymborski: Not a lot probably

Guest: Is there a way to statistically quantify when a player is too passive? If so, is that the Termarr line?

Dan Szymborski: I don’t think there’s a hard and fast rule

Kevin: What, if anything, do you know about what teams are doing or considering using artificial intelligence tools? Any teams known or believed to be at the forefront? Any topics known or imagined to be most ripe for exploration?

Dan Szymborski: I don’t actually know; it’s not a question I’ve ever really asked anyone and nobody’s brought it up to me

Dan Szymborski: I suspect unless you’re talking things like more traditional statistical analysis using neural networks, AI is not really important

Ben: If tomorrow you were made a GM, what is one thing you’d do that would go against conventional attitudes most?

Dan Szymborski: Argue on Twitter that it was a really poor idea to make me a GM

Dan Szymborski: I’d probably invest more in scouting than most quant types

Dan Szymborski: Especially of players in system

Dan Szymborski: Because even if it’s uncertain, the knowledge you have of YOUR players is where you have a built-in advantage over other teams, so you want to leverage that as much as possible

Dan Szymborski: So even if the data are something subjective and messy and volatile, it’s also where you can discover things about your players and not have someone else do so

Dmx: I like the pitch clock. I just don’t want to see it on the broadcasts. It adds no drama to see a countdown in the scoring bug between every pitch.

Dan Szymborski: Yeah, I don’t need to see it

J: I think it’s also worth noting that median projection may miss out on 103 win teams but it hits very close on a lot of 70-90 win teams every year. It’s just that people latch onto the big misses for outlier teams.

Pat: Speaking of Chris Sale, how much has his resurgence changed his future outlook? Obviously, health is a big concern, but, he’s shown that when healthy, he still has “it”. What percentile of future projections gets him into the HOF?

Dan Szymborski: Yes, and I might write about that soon

Bernie: Has anyone increased their ZIPs wRC+ projection more than William Contreras (113 to 133)? I guess, Joey Ortiz has gone from 82 to 104

Dan Szymborski: Looks like biggest gaienrs are Fry, Pillar, Profar, Ozuna, Sosa, Ohtani, Pederson, ORTIZ, Sal perez, Wade, Stallings, Willson, WILLIAM

Dan Szymborski: (Obviously Willson had something very bad happen which may prevent that)

Dan Szymborski: I really think he should be Wiliam Contreras

Dan Szymborski: Because of Willson

Dan Szymborski: And if you have a Willson and a William, you’ve damaged the balance of L’s in the universe

Mike Trout: Are the headlines wrong that there are new single season record holders with Negro League stats? Don’t they not meet the minimum requirements for at bats or innings?

Dan Szymborski: At bats/innings are league based

Dan Szymborski: So yes, someone with a lead in 40 games or something has to be contextualized

Dan Szymborski: but all leaderboards should be

Dan Szymborski: That someone has a 1.50 ERA but never faced a black player is a contextualized thing too

J: Hey if you just report r of 0.25ish instead of r^2 you can pretend like that’s meaningful.

Sad Mets Fan: can you offer one good ZiPs projection for a Met

Dan Szymborski: Vientos proejctiosn have nudged up a bit

J: I think the conventional attitude would in fact be to argue on Twitter that it was a poor idea to make you gm

Dan Szymborski: No, but for ME the GM to argue it would be unusual

seth: is zips biased towards odd numbers

Dan Szymborski: Well, yeah, because of A-Rod’s alw

Dan Szymborski: law

Question Asker: Most underrated phrase of speech in your opinion?

Dan Szymborski: The proof of the pudding is in the eating makes a lot more sense than it seems when you actually say the damn phrase properly

Dan Szymborski: Not really phrases, but I like words with consonants that give them a different feel

Dan Szymborski: Like jejune is such a wonderful little word

Dan Szymborski: People never really think of the tone color of words

Dan Szymborski: I think partially because English is a low context language

Dan Szymborski: and I love little used words that are evocative

Dan Szymborski: like shabby and tatterdemalion FEEL very different when you read it

Dan Szymborski: It’s not different than music

Dan Szymborski: Like in Dvorak’s B minor cello concerto, the descending theme that the cello plays after the big orchestral moment in the 3-4 minute range does not work the same if it doesn’t have the clarinet giving counterpoint

Dan Szymborski: