I wrote an article exploring SEAGER, trying to understand what makes it tick, and creating a simpler version that doesn't require advanced modeling (r=.72). SEAGER isn't just fancy Z-O Sw%, and I think it predicts ISO by finding hitters who are feared. Link in next tweet.
New Post up on Medium: Some interesting preliminary thoughts on MLB's new public bat tracking data. Inside is some data on swing length and perfomance vs high velocity, the consistency of Giancarlo Stanton's bat speed and more...Link 👇.
@sdut
That's an awful opinion. Growing salaries are a signal a league is popular and doing well. See: NFL, NBA, international football, etc. Sports leagues make tons of money *because fans love them and spend lots of money on them*. Stop schilling for owners who make 100x more.
I’m 44 and was diagnosed yesterday with ADHD, specifically the Inattentive type. It's a bit of a shock, as I have zero hyperactive tendencies, was a great student (until college) and was always [too] well-behaved. But after reading about the symptoms, it fits me to a T… (1/N)
A simple measure of plate discipline is ZoneSw% - OOZSw%. Swinging at strikes is good, swinging at balls is bad. By that measure, the Pirates are blowing everyone else away in YoY improvement from 2022 to 2023. Biggest losers? The White Sox and Royals.
@Sky_Kalkman
@bymichaelrosen
I was finally able to circle back to this tonight. Cleaned it up and threw the code into some functions... all you need is BaseballSavant data for input. Hopefully this is helpful. Let me know if you catch anything that needs to be fixed
I'm fine with MLB teams being mediocre. By definition many teams will be mediocre. It's entertainment, mediocre teams are entertaining. I'm not ok with teams not trying. I'm not ok with teams valuing money over winning. Reward trying, find the entertainment in mediocre baseball.
Yes, I started a Substack. I'm just going to treat it like a (free) blog. Subscribe or don't, I will share anything I write here on Twitter. I welcome your feedback. /cc
@NotTheBobbyOrr
It's 2023. DH Josh Donaldson steps to the plate with two runners on base, his Diamondbacks down by a run. What the hell, you think, I'd much rather watch the pitcher try to lay down a bunt. You turn on season 19 of Grey's Anatomy instead.
@chrisalbon
It's also a risk/reward issue. What's the upside if you're right and the downside if you're wrong and how often do you expect to be right. The time committed to being more certain can often be better sleeve on new analysis, or follow-up after a test.
@bkabak
From another angle, think about how many more murder suspects these two cops could have caught if they weren't just standing around a turnstile on a daily basis.
A look at team offensive performance so far this year, by wRC+ (all hitting events, no sequencing, adjusted for ballpark.) Yellow highlights:
- LOL Rays
- Red Sox! Rays-ization coming to fruition?
- A's offense is not bad!
- Mariners and Astros disappointing
- LOL Guardians
Listen, I will grant that we don't need* more public Stuff+ models, but for the love of god could one of them please be readily available on a pitch-level basis? It ain't public if you're just teasing the aggregated results.
@BrentToderian
@modacitylife
For those asking about bike thefts -- daily bikes are not super valuable and there are a zillion of them (1.3 bikes per person vs 0.3 in US.) Lower motivation for thefts and lower probability of having yours stolen.
Your logical fallacies are:
1. False dichotomy
2. Strawman
3. Appeal to emotion
4. Appeal to authority
5. Personal bias
Points for efficiency, I guess.
Ex-
#Yankees
1B Mark Teixeira on ESPN today: “I would rather make pennies on the dollar and give hope to people and play baseball than not make anything and lose an entire year off their career.”
Stuff+ models correctly account for the context of the ball-strike count. But I wonder if they over-weight it. For example, two-strike counts are high leverage. Whiff pitches get rewarded. But should that one pitch trump others, or should it be de-leverages, like WPA -> WPA/LI?
Just published an article about converting Stuff+ to runs and the ERA and wOBA scales, including reasons you would want to do that even after the creators went through all the hassle of putting it on the Stuff+ scale in the first place. Link in next tweet.
I believe you can calculate Simple Seager from just Zone%, Z-Sw%, and O-Sw%.
YIKES is where the calculations got annoying to write out. (Classic "leave it to the reader" vibes.)
Original explanatory article in the next tweet.
/cc
@TJStats
Y'all are fretting over picking guys who are less popular in the Immaculate Grid, and I'm just trying to remember anyone other than Randy Johnson who played for the Diamondbacks.
@FIDE_chess
So, "we weren't doing enough, good thing the world champion brought the issue into brighter light, and now we will ramp up our anti-cheating efforts."
Again, if the argument for billionaires making so much money is that they carry the risk, then they need to carry the risk. You can't have it both ways.
Mind drifts, sidetracked, difficulty getting started on tasks, lose motivation to finish, unorganized, and feel overwhelmed by complex or multiple tasks. Apparently bright kids can mask their ADHD because they understand and complete work quickly; there’s no *need* for attention.
"Things we observe but can't explain" is one of my favorite genres. Always brings to mind the apocryphal story of arguing to shut down the patent office a hundred years ago because we'd already invented everything.
You're probably seen amazing videos of metronomes getting in sync (there are several movies on YouTube with millions of views). But surprisingly the math of how all this works is still up for grabs. Our latest article on this puzzle is freely available at
Strongly held opinions I don't see that often:
1. Losing the "sanctity of sports" is an easy "price" to pay for social equality of trans people. Like, it's sports.
2. Of course felons should be able to vote. Otherwise you just throw all dissenters in jail and keep power.
How have team components contributed to their overall record? I broke up win probability added by hitting, SP, and RP for each MLB team ordered by best to worst win percentage.
Looks like what I'd guess to be a histogram of all run scoring. 1st inning is top of lineup. 5/6/7 is tired SP, third time vs SP, or worst RPs. 2/3/4 is bottom of lineup and prime SP. 8/9 is best RPs.
I wish more people would write *about* the implications of their Stuff+ models. There's an article or two about the model development, and then tons of content using its results on the player level. But there's cool stuff in between, about what makes for good/bad pitches.
Ranger Suarez has "struggled" his past two starts, against two suspect offenses in Miami and Washington. But those two teams are actually good lineups against lefties. Miami is 4th best with a 123 wRC+ this year, about as good as Tampa Bay, and Wash has a 107 wRC+.
Teams are hitting just 1.03 HRs per game this season. The main reason is that the ball isn't flying as far.
BPP predicts the result of every ball hit during the year - controlling for weather and park effects. We're now seeing ~20% fewer HRs than contact quality would imply.
The Economist has Biden at 98% to win the the most votes in the next presidential election. That there's doubt he won't be the next president is just absurd. Fix our democracy asap.
Super interesting, uh, stuff!
1. Multiple fastball use is up. Primary assigned by handedness.
2. Cutters more valuable, sweepers and 4S less valuable
3. Fastballs have non-linear gains 96+ mph
4. 2S > FF until 97 mph
26.5 hours of cycling/running in January. No alcohol or caffeine. Eating well. Cut job hours by 20%. Started with a new therapist. Yeah, feeling absolutely great, let's keep this going.
Ok, let's do some "greatest living ballplayer" hot takes...
If it's Bonds, it was Bonds before Willie passed away, too.
If it's a pitcher, it's Randy. Then Maddux.
Nolan Ryan was a compiler. Koufax was an in-season compiler. Blyleven's case is just as good.
It might be Trout.
Mike Pence and Secretary Azar have said that only 3% or 4% of counties in the U.S. are seeing a surge in coronavirus cases.
But they and other officials are skirting a key fact: More than 20% of Americans live in those relatively few counties.
Luis Gil changeup Stuff+ scores this year:
tjStuff: 96
enoStuff: 91
StuffPro: 0.9
PitchProfiler: 100
Mostly below avg. And then botStuff says it's a 73, a ++ pitch. (And might be the closest to right?)
In light of
@MLB
's stance to undermine election integrity laws, I have instructed my staff to begin drafting legislation to remove Major League Baseball's federal antitrust exception.
So let's say Yadier Molina's pre-game prep, pitcher training, pitch calling, in-game tweaks, etc. turned every Cardinals' starting pitcher from a
#4
into a
#2
. Not only is he an inner circle Hall of Famer, but he's worth $50M+ a year as a coach.
Two keys to advancing at and enjoying work:
1. Identify, leverage and thank the hell out of those people who are awesome at their jobs and helpful to anyone to who asks.
2. Become one of those people.
Here's a AAA line: .316/.357/.624 -- that's Mickey Moniak at Salt Lake, good for a... 122 wRC+. Similarly, Jo Adell has 12 HRs and a .300/.395/.709 line for a... 147 wRC+. I'm curious if minor league teams in extreme environments limit prospect development.
It's been a while since I was interested in writing more than a few tweets about baseball, and I really enjoyed it. Thanks to the fun community here on Twitter, especially the Stuff+ crowd that I've been harassing over the past year. I like asking questions.
This is the stat I've been waiting to see. Judge has been absurd, so much better than anyone else. Depending how you treat quality of competition, he *might* have had the best three-season hitting stretch ever.
@Sky_Kalkman
Here's the top 10 by excess Called Strikes, based on the PLV Called Strike probabilities. Langford has the 2nd highest false positive rate of the group.
@chess24com
Can someone explain to me what he's risking? (Honest question.) I would have thought it's the other way around -- worth the risk of losing ratings pts on the chance that he makes the Candidates...?
Sharkman & I are in agreement:
We’re ready to vomit over this “sweeper” craze in MLB broadcasting.
Who died & gave a few bored weenies the power to create a new pitch last winter?
It’s a slider, dammit.
Almost done writing an actual baseball article with actual research. First in a long time, enjoying it. Tease: includes a simple version of Seager that correlates at r=.72.
Connor Joe is among the most improved in swinging at Meatballs while letting OOZ pitches go. Suwinski and a couple other Pirates, too. Organizational effort in the off-season?
So my guess is that bat speed data has significantly shortened the amount of time needed to discern a hitter's platoon splits. The priors on a player like this are not league avg platoon ratios.
Did you know that "SAT" doesn't stand for anything anymore? It was Aptitude then Assessment, but turns out the SAT doesn't assess scholastic aptitude very well. It mostly just measures socioeconomic status, which, well, makes this opinion piece pretty wrong.
In Opinion
“Lots of kids use the SAT or ACT to get a boost out of a bad school,” writes Ross Douthat.
The “children of the upper class could be beneficiaries of the SAT’s decline, while children trying to climb could lose a crucial ladder.”
Life tip: it's ok to give people gifts they've asked for. They will like the gift, mostly because they told you they would like it. You don't lose points for them telling you they'd like it. (I've learned this the long way, you shouldn't.)
@RealRyanWhorton
Sorry, but it's true.. the older generation of homosexuals were more comfortable with their sexuality they were proud of being gay they weren't forcing people to accept them they weren't crying victim. Actually, the older generation had more respect for themselves.
I've had a gut feeling that depth charts overrate how many innings a "reliable" starting pitcher will get, so I decided to run some numbers. I looked at seasons from 2010 through 2019 (avoiding the 2020 season). Here's what I see... (1/x)
I get nostalgic about the Beyond the Box Score crew circa 2010 (?) like every six months, but now seems like a decent time for that, too. Cool people, cool work, cool time of my life.