Dan Szymborski FanGraphs Chat – 1/11/24

Dan Szymborski: And we are here!

James: You never win trading away Juan Soto. That said, Preller must be thrilled with his haul right now given how the SP FA market has been shaking out, no?

Dan Szymborski: Yeah, though I’m  not thinking much of Padres in 2024; that’s still a lot of performance to replace and they kinda needed pitching *now*

Laxtonto: Does it seem like this offseason is moving slow on the FA front? Or is it just perception of the long offseason?

Dan Szymborski: I dont’ think slower than usual. One thing is that there just aren’t many very good position players ou thtere

Dan Szymborski: Been really lite on stars

Laxtonto: Should I let the 3 year old go outside and build a snowman or stay inside and make hot chocolate?

Dan Szymborski: Build a snowman AND hot chocolate after

aunt bea: happy birthday bethany this is aunt bea Claudine said i can wish you happy birthday in your Facebook is this it hope it works… kisses -aunt “bea” (Beatrice Gerbell)

Dan Szymborski: I don’t quite get the joke, but ok!

Barold Bondsworth: Dan, can you please reassure me that the Giants don’t plan on going into the season with a rotation of Logan Webb, Ross Stripling, and three rookies? I feel like I’m losing my mind

Dan Szymborski: I suspect they’ll make a low key signing.

Dan Szymborski: Maybe just one. Might depend on how optimistic they feel about Cobb and Ray health

Jacstorm: Any thoughts on what Brenton Doyle will produce this year?

Dan Szymborski: Not much offense, tons of defense?

The Q: How are you viewing Brayan Bello’s future? Only 24 years old. Do you perceive he’s a future #1? Or probably more in the 2/3 mix? See anything in the numbers that suggests what he’ll be not so much in 2024, but 2026?

Dan Szymborski: I think 2/3 guy is probably the most likely

Dan Szymborski: He’s kind of got that early Eovaldi problem with punching out guys on two strikes

Watashi Wa Ringodesu: Is it true that there is absolutely zero statistical correlation that has been found between steroids and heightened performance?

Dan Szymborski: That’s a rather loaded question!

Dan Szymborski: Now, I’ve spent like 15 years trying to find any pattern of players busted and any group differences in aging pattern/overperformance/underperformance at any point before and after their usage

Dan Szymborski: And I haven’t found a dang thing

Guest: Dan, if you were Mike Elias, other than the obvious ones–Holliday, Rutschman, and Henderson, who would be on your personal no-trade list and why?

Dan Szymborski: Nobody. You always talk

I am the Apple: Averaged out over the entire MLB population, what age is the average/median peak? 26?

Dan Szymborski: The tricky part is that it changes a bit depending on *how* you evaluate peak and which set of players you include

Dan Szymborski: 26-27ish is the safest bet though

Richard: Is there an easy way on Fangraphs player pages to compare pitchers’ ERA/FIP to the specific league (PCL, International League) averages? For hitters, we have wRC+. If not, is this something Fangraphs is going to eventually add?

Dan Szymborski: It *is* weird we don’t have that for minor leaguers. I’ll bring it up

Padres2024: Any chance we will get an updated ZIPS projection for the Padres.  The first one came out right before their very large roster change!  🙂  Or, at least an updated Ballpark Graphic?

Dan Szymborski: We’re only a few weeks from the pre-spring zips projected standings!

Dan: I find myself wanting Betts Ohtani and Freeman to have historic seasons but also the Dodgers are my 5th favorite team in the NL West. Do  the Padres, Giants, and Dbacks all have 100 win upside that I can dream on?

Dan Szymborski: Depends what percentile you’re willing to go to! Any plausibly .500 team has at least tiny chance of winning 100

laxtonto: Cake or pie?

Dan Szymborski: It depends. I don’t like really sweet things. I’ll take a not sweet cake over pie, but I’ll take pie over anything with, say, American buttercream

Jon: Yo Dan.  Is there a psychology / willingness to learn and adjust factor that changes how players grow or slow a decline over time?  Do teams mine for that info?

Dan Szymborski: There *may* be, but it’s really noisy. I don’t know what teams are *specifically* doing here as my relationships with teams/agents is always at arms’ length.

KrusherKovalev55: Do you expect a guy like Cole Ragans to be someone who ZIPS is likely undervaluing due to small sample size and large velo bump? The projections on him were somewhat disappointing but I get that he is a weird case. Just wondering what you expect from him this year and hoping the answer is something other than just “see ZIPS!” ha

Dan Szymborski: I’m really not sure what to expect for him honestly!

Dan Szymborski: I mean, ZiPS knows velocity

Dan Szymborski: Pitchers are trickier and ZiPS isn’t confident on a guy without a whole lot of IP

TomBruno23: We have the full 2024 NL Central ZiPS at this point…who are you betting on to get to 85 wins and take home the crown? Who is your favorite longer shot? Does it matter?

Dan Szymborski: Cardinals, though I imagine people will be annoyed with me since it’s against the current popular feeling about them

Dan Szymborski: But I wouldn’t be shocked by any team winning the NL Central. Even the Pirates.

Richard: For pitchers who had starts and relief appearances, do you think it makes more sense to project them as “Player X – SP” and “Player X – RP” than mash a combination together? I understand the blend projection is going to be closer to reality but it muddies the talent level when you’re trying to parse the two.

Dan Szymborski: I’m still thinking of better ways to express that

Dan Szymborski: For next year, I’m considering any pitcher with, say, three projected appearances in their non-primary role, to tack on another chart breaking down the ERAs/FIPs

Dan S.: Dylan Cease has had two very different recent years, making it hard to settle debates over his trade value. What does ZiPS think of him going forward?

Dan Szymborski: Don’t take away one of the few reasons to people to look at White Sox ZiPS next week! lol

Drew: Is Grissom really the solution at 2B?  The bat says yes, but the glove seems to scream NO.

Dan Szymborski: He *might* be. Remember, 2B is the easier position and defensive stats are volatile

Richard: If EDLC cut hit K rate to 20%, would he become a 7 WAR player according to ZiPS?

Dan Szymborski: That’s a bit much. It would help of course

J: Any ZiPS projections on Imanaga? Was surprised to see none in the writeup on the signing.

Dan Szymborski: Cubs go tomorrow!

Mark: Tan you equate a team’s total projected WAR to a number of wins?  In other words, does 48 combined WAR equate to 97 Wins, etc?  Related to the fancy projection charts you publish.

Dan Szymborski: Not REALLY since I use a method that attempts to deal with depth issues

TomBruno23: Journey/Def Leppard/Cheap Trick and Billy Joel/Sting as the big summer concerts at Busch fits so perfectly with that Cardinal Rotation.

Pistol Pete: Hi Dan, do you think that Brandon Pfaadt’s growth during the postseason will carry into the 2024 season or is his arsenal too hittable to be trusted?

Dan Szymborski: I remain optimistic

Bob: What is a good way to evaluate the career of a pitcher? Is the bar set too high for WAR for pitchers? WAR is supposed to be an indicator of value over what a replacement player would bring. Pitchers like Drew Smyly and Ryne Nelson had a WAR of zero last year. The 10th best pitching WAR is equal to the 42nd best hitting WAR. Pitchers totaling 10,106 IP out of 43,087 total IP (23%) had a WAR of 0.0 or less last year. The best Team Bat WAR last year was around 40 but the best Team Pit WAR was around 33. And bullpens that accumulate 500-600 IP often have a team bullpen WAR less than of a single starting pitcher. Seems like using WAR as an indicator of overall value undervalues pitchers (especially relievers) and the calculation should be modified.

Dan Szymborski: I’m not quite sure

Dan Szymborski: There’s no rule that says pitchers have to equally be valuable as hitters

Dan Szymborski: Imagine baseball set a rule that pitchers could only pitch three innings in a game

Dan Szymborski: (individual ones)

Dan Szymborski: That would serve to make pitchers less valuable.

Dan Szymborski: And for bullpens, the fact remains that performance there has a much higher replacement level.

Perry: Is Zips able to make predictions about a pitcher’s performance but with an altered profile? For instance, as a Braves fan, I’m anxious about Strider’s effectiveness in the short term if he loses, say, 3mph on his fastball since he’s basically a two pitch pitcher.
If so, how would that loss in velocity be mitigated by him adding an effective 3rd pitch with X profile and Y usage rate?

Dan Szymborski: Yes, though it depends on the exact question

Devon: When do 3-year ZIPS projections push to player pages? Will that be when all team ZIPS profiles are finished?

Dan Szymborski: Yeah, when everything goes live

Dan Szymborski: I send it to David in one big chunk

Marcos: Why do you like Imanaga for the Cubs so much?

Dan Szymborski: Yeah, he doesn’t have a ton of velocity, but his fastball isn’t a lifeless one

Dan Szymborski: I think he’s very Kodai Senga-y in value

Dan Szymborski: and that contract is a steal for that

laxtonto: Has anyone done any text mining in baseball? It seems to generate less large textual segments (maybe outside of scouting reports) and large scale text topic classification doesn’t seem like a real functional application. Maybe sentiment correlation within scouting reports to determine hedging by scouts….

Dan Szymborski: I’ve actually experimented with this in the past with good results, but I find that just using the data I have access to gets me most of the things that are useful for projections

TomBruno23: Do you (or whoever it is that runs ZiPS still trying to figure that one out…) put any stock in Winter League statistics? Wondering because new Cardinal acquisition Victor Santos is at the least showing he is healthy.

Dan Szymborski: I’ve found them too noisy for my use

Dan Szymborski: and I only use them, like college stats, when I basically *have* to project a dude

Dan Szymborski: But as I’ve said before, there’s always a decent chance that someone has figured out an approach that I have not

Dan Szymborski: there are a lot of people out there who are both smarter than me and look at baseball

Insert Witty Name Here: the rule changes last year I think made the game more enjoyable.  What other changes do you think are needed to keep fan interest going? Maybe not specifically tied to the actual game, but also aesthetics or game presentation?

Dan Szymborski: COMPULSORY BULLPEN CARTS

Dan Szymborski: EVERY STADIUM MUST HAVE AN ORGAN AND FULL-TIME ORGANIST

Dan Szymborski: CB BUCKNOR SHOT OUT FROM THE STADIUM BY TREBUCHET AT ALL-STAR GAME

Dan Szymborski: I also think MLB should go fully crazy on Atlanta League in experimentation

Dan Szymborski: they already do

Dan Szymborski: but be even more ambitious

Dan Szymborski: 88 feet to first base

Dan Szymborski: scoring on grounders with two outs not requiring batter to be safe at first (purely temporal)

Dan Szymborski: Maybe it’ll suck, but I want to try

Dan Szymborski: I think people have a fair point that there’s too much sameness in offenses

Dan Szymborski: this stems on the fact that what makes a good offense is already *solved* in baseball

Dan Szymborski: so UNsolve it!

JT: Zips seemed to love Noelvi Marte before last year. Now it…hates him? After a solid year? I’m confused

Dan Szymborski: League differences

Dan Szymborski: he basically lost 30 wRC+ in promotion

Dan Szymborski: which has reduced the short-term upside somewhat

Dan Szymborski: This isn’t 2005. The International League scored nearly six runs a game and the Southern League was over five

Nervous Flyball Pitcher: Is there an easy way to index a given offensive stat (OPS, wRC, etc.) to the position instead of all of MLB?

Dan Szymborski: Not REALLY

Dan: How long before someone currently in the Minor Leagues is considered the best player in baseball?

Dan Szymborski: I’ll say five years

Greg: Today marks 3 weeks from the solstice, meaning we have survived the darkest 6 weeks of the year

Yo Soy La Manzana: Does ZiPS use WAR figures in its calculations at all? And, if so, does it use bWAR or fWAR?

Dan Szymborski: zWAR is kind of in the middle

Dan Szymborski: Basically, ZiPS will evaluate how a pitcher has done relative to their defenses

Dan Szymborski: and as the total batters faced, will credit the pitcher with more and more of their ERA/FIP gap

Dan Szymborski: So if Tom Glavine pitched this year and picked up where he left off, it would be giving a WAR that is closer to RA9-WAR than WAR

Dan Szymborski: But a rookie it’ll be mostly FIP

Dan Szymborski: But those only go in the team rundowns. They’re centered on fWAR when they go on the site

Dan Szymborski: Because the benefits of tinkering are likely outweighed by the cost of making things more confusing

Dan Szymborski: People tend to underrate the value of clarity in metrics

Reed: When you post a table of ZiPS projections in an article (or provide for another author) can you please include an age column?

Dan Szymborski: I could, though problem is, especially with batters, I’m already running against the 600 pixel max

Chris: You tweeted earlier about translating Teoscar’s ZIPs from SEA to LAD needing to take into account the change in ball park. So would the same apply to Sale from BOS to ATL? (I.e., I shouldn’t just take his numbers listed from the red sox projections?)

Dan Szymborski: Generally speaking, it’s only a few runs, so you don’t HAVE to obsess over it

Dan Szymborski: it’s just in the case of Teoscar, in addition to the bump, it made the team’s roster construction make a lot more sense too

Dan Szymborski: It was like a 0.4 WAR projection improvement but the team depth chart went up like 1.1 or something

Dan Szymborski: It’s less than two weeks until these go live, so it’s not long to wait!

No big signings: In your opinion, which orioles starter is most likely to hit his 80th percentile projection?

Dan Szymborski: Grayson

Sleepless near Seattle: The Seattle lineup is really confusing to me. Do you think they are going to go after a big name at 3B or do you expect DiPoto to continue the split shuffle?

Dan Szymborski: I suspect they’re going to keep cobbling stuff together

G4: Are the Brewers overrating the benefits of great defense, or does last year’s record justify its genuine ability to offset poor hitting (presuming a team can’t afford to buy both).

Dan Szymborski: Wel,l the pitching was quite good!

Dan Szymborski: There are limits to how much defense can offset

Dan Szymborski: Because you simply can’t get a 40 run upgrade at defense anywhere

The Q: If you were Breslow would you prefer Snell or Montgomery?

Dan Szymborski: Montgomery

laxtonto: How often do you use nonparmatric approaches vs classical approaches in your modeling? Doesn’t the 40-80 scouting approach in essence try to force a Gausian distribution into the logic of baseball analysis when most extreme talent is probably closer to a Poisson or Exp?

Dan Szymborski: I pretty much exclusively use nonparametric approaches

Refugee: Does ZiPS use pitch modeling stuff like Stuff+ or PitchingBot in its pitcher projections?

Dan Szymborski: It does not. I’m still figuring out how to construct one that results in increased predictive value

TomBruno23: If you were in a competitive eating contest what would you go-to food be for the win?

Dan Szymborski: Milk

Dan Szymborski: I’ve succeeded at gallon of milk challenge two times in three tries.

Dan Szymborski:

Dan Szymborski: The guy on the right, Chris, did not

Insert Witty Name Here: A few years with gambling being legal, the commercials are getting to be a bit much. Especially considering how advertising can modify human behavior (meaning I don’t WANT to gamble, but with the non-stop showing of ads and odds, not everyone has impulse control), do you think it may be in the best interest of the population to ban gambling ads like they did with tobacco ads?

Dan Szymborski: It might be in the best interest of the population

Dan Szymborski: However, I would be against it personally based on my first principles.

Marcos: Do you have a Tom Glavine 2024 projection?

Dan Szymborski: Dear god no

laxtonto: Is there any heuristics that you use that you haven’t been able to figure out why t hey work as a proxy?

Dan Szymborski: I don’t like using *anything* I don’t understand

>this guy<: Dan if you are going to pitch things for the website, how about something that allows you to change ABs or IP to adjust projections for players? I am really thinking about for fantasy and auction calculator/ADP but maybe for everyone?

Dan Szymborski: I could

Dan Szymborski: But if you guys have ideas you want to see, it might be best to DM me as I’ll remember them

Reds trades: They haven’t happened yet but isn’t the deadline still a viable opportunity to rearrange the roster?

Dan Szymborski: It is

Colton: Had a conversation with some friends the other day in regards to defense vs offense. One stood firm in “a run saved is the same as a run earned” and would not come off that notion. Any strong feelings yourself on that ideology? I agree with it in a vacuum but baseball isn’t played in a vacuum and that’s why offense is so coveted over defense in my opinion.

Dan Szymborski: I mean technically, it’s accurate

Dan Szymborski: Though offense is coveted over defense because of initiative

Dan Szymborski: Adding a run more of offense is a more straightforward process than allowing one fewer run of defense, since there are interacting elements (pitcher, defense) at play

Insert Witty Name Here: Please describe in detail the unsuccessful milk challenge try.

Dan Szymborski: I’m sorry, nothing gross was involved

Dan Szymborski: I was just three-quarters of the way and my lower digestive system starting complaining, so I threw in the towel

Greg: “I don’t like using *anything* I don’t understand”

Dan Szymborski: Or at least tinker?

Dan Szymborski: I mean, I couldn’t rebuild a transmission, but I use a car with a transmission

Dan Szymborski: but I wouldn’t tinker with it

Insert Witty Name Here: What are your first principles?

Dan Szymborski: I don’t like getting into politics, but you can probably infer what my general philosophy is

Jeremy: Follow-up question re: rule changes: is there any other way besides sufficiently-frequent rule changes to keep baseball “unsolved”? That is, is there some combination of rules that would make it optimal to be different–to be zigging when others are zagging? I feel like the interplay of NFL offense and defense is kind of like this–that the predominant approaches change frequently, because different approaches have a rock-paper-scissors-type relationship with one another.

Dan Szymborski: The problem is that a lot of offense isn’t so much choice but execution

Dan Szymborski: Choices may be cyclical, but execution isn’t

Dan Szymborski: Though this may only make sense in MY head

Jeremy: Wait, aren’t all players equally likely to hit their 80th percentile projection by definition? Don’t they all have a 20% chance of hitting their 80th percentile projection?

Dan Szymborski: I think I was being asked *about* the 80th percentile projections, not what ZiPS thinks is the most likely

Dan Szymborski: in the latter case, it woudl be 20%

Quist: Of all American writers with the initials RC, which is your favorite?  Raymond Carver, Raymond Chandler, or Robert Crumb

Dan Szymborski: Robert Creeley maybe?

Dan Szymborski: I can’t think of many I’ve read off hand

Dan Szymborski: that have those exact initials lol

Imanaga: When a player like me is underpaid relative to projections, do you generally assume there is something on the medical report the general public isn’t privy to?

Dan Szymborski: No, there are a lot of reasons for things. I don’t like assuming

JB: how do you feel about framing being in /out of WAR ?  It seems odd to give value on an event where the ball isn’t put in play.

Dan Szymborski: It’s a driver of bad/good counts. That has value

Jeremy: No, that makes sense in my head too.

Dan Szymborski: That could just mean you’re as crazy as I am

Greg: My original comment on “I don’t like using *anything* I don’t understand” was “That’s what she said!” so the direction that  it went in was probably better anyway haha

laxtonto: How frequently do you apply clustering algorithms when looking at ways to segment players?

Dan Szymborski: Not *too* often, depends on the task

Frank: Why didn’t the foot soldiers just take a gun and shoot the ninja turtles?

Dan Szymborski: I’m sure that in the lore, the bullets would bounce off their turtle shells?

Marcos: What would you do if MLB gave you unilateral realignment powers,

Dan Szymborski: Two expansion teams and then four divisions of eight

Sir Nerdlington: The RSN situation seems like it’s going to collide with the next CBA negotiation where owners won’t have enough info to decide upon their collective position. Forget lockouts, are we going to see a full civil war over rev sharing amongst owners?

Dan Szymborski: It’s very possible

TomBruno23: Finally got around the MLB doc, “Donny Baseball” man was that an hour well spent. Give me 80s/90s baseball docs all day, please. Thank you.

Colton: Reading through the Reds projections and I audibly cackled when I read the Pagan sentence, what on earth were they thinking signing him!

Dan Szymborski: Wait, did I make a joke unintentionally?

the person who asks the lunch question: what’s for lunch?

Dan Szymborski: Pastitsio

laxtonto: How rare is a 115-loss team? How frequently does the system project an extreme outlier in losses? What was the most extreme you remember?

Dan Szymborski: It’s pretty rare because the underlying ability implied by the spread of results suggests that pretty much everyone has an underlying prob of .400 to .600

Dan Szymborski: The worst ZIPS team projected ever was 2019 Marlins at 56-106

Dan Szymborski: So basically, you need a team like that and they have to also be unfortunate relative to expectations

Behind Enemy Lines: Adam Ottavino last night said Imanaga might have the “hoppiest” fastball in MLB, meaning perceived rise. Is this a thing? And, can we get an actual rhythm section with organ instead of canned samples?

Dan Szymborski: Maybe we should call his fastball the pale ale?

Colton: In my quick research of that offense vs defense topic, I ran across someone (I think on Reddit) who mentioned the idea of having a team of Adam Dunn’s vs a team of Kei Igawa’s and run a 100 game simulation of them against each other to see who wins and it’s all I’ve been able to think about since.

Dan Szymborski: Pitching too?

Marshall: I think I saw that you and Jay teamed up to do some JAWS projections. Is there anywhere I could see a list of players who project to clear the JAWS standards at their position (say, based on a median projection)? If not, that would be a really fun series of articles.

Dan Szymborski: Generally, I give them to Jay as Jay needs them

JB: I notice most analytic writers have moved to using fWAR vs bWAR.  Is this because most people are writing about what might happen next as opposed to what has happened?  When should you use one vs the other ?

Dan Szymborski: Here’s the problem: when it comes to individual pitchers, there really *is* no “what happened”

Dan Szymborski: runs allowed happened to a team, not to the individual pitcher

Dan Szymborski: it’s just that someone in the 19th century decided “if it happens while the pitcher is on the mound, it’s the pitchers fault”

Dan Szymborski: People act like RA9-WAR is a what happened WAR

Dan Szymborski: But the decision to assign *everything* to the pitcher is just as much a theoretical construct as assigning *some things* to the pitcher, whether by isolated pitcher-specific events or applying a general defensive metric

Dan: how does Chris feel about you calling him out in a internal forum like this?

Dan Szymborski: Hey, I didn’t say his last name!

Colton: Just reading “Emilio Pagan was the worst RP addition a team has made” gave me a good laugh cause I thought the exact same thing when it happened lol

Colton: They inferred having a league average pitching staff for the simulation.

Dan Szymborski: Oh that’s not fun

Dan: offense is better because you can’t have negative runs so the closer you get to allowing zero runs the less better defense helps but on offense the better you are at offense the more at bars your best players get and the more you face weaker pitchers so adding offense has an increasing effect.

Dan Szymborski: True, but it doens’t really come into play when talking about assembling a roster for a season

Dan Szymborski: individual games, of course

JB: Baseball trade values is a pretty cool site.  Fangraphs led the way on this valuation logic. Have yiu or FG considered  using zips to create a public version of trade values ? (I assume you use it privately)?

Dan Szymborski: I actually have my own thing I’ve tinkered with in ZiPS

Dan Szymborski: I don’t know whether anyone’s considered putting one together on site. If they have, I wasn’t in the conversation

Dan Szymborski: I actually have a pitcher vs. hitter individual matchup tool that I use myself that may pop up on FG at some point

Milk Toast: Does Chef Boyardee still visit you in your dreams?

Dan Szymborski: If he is, I have no memory of it

Dan Szymborski: For those who are confused, some years ago, I woke up to find I had scrawled this on the notepad on my nightstand at some point during the night.

Dan Szymborski:

Oddball Herrera: Per the tracker, Sheffield is actually over the HOF threshold.  If he and Ortiz end up in the hall and other PED linked (or even speculated) guys don’t, I give up on trying to figure out what the standard is

Dan Szymborski: I think the issue is simply you have hundreds of different standards

Dan Szymborski: On that note, it’s time for me to head out for another week.

Dan Szymborski: Got some ZiPS to finish ZiPSing