Abnormally Distributed

285 posts

Abnormally Distributed banner
Abnormally Distributed

Abnormally Distributed

@AbnormallyDist

I'M STILL THAT GUY I'M STILL THAT GUY

เข้าร่วม Ocak 2019
409 กำลังติดตาม7.2K ผู้ติดตาม
Abnormally Distributed
Abnormally Distributed@AbnormallyDist·
easy fix to poly fees. website shows parabola bc they're focusing on $ paid not %. % looks like a linear line and makes it dumb to bet dogs. hurts US poly more where no spreads priced ~50/50. just switch the formula to # shares * 6% * p^2 * (1-p). then ppl can bet dogs again.
Abnormally Distributed tweet mediaAbnormally Distributed tweet media
English
3
0
16
2.1K
Abnormally Distributed
Abnormally Distributed@AbnormallyDist·
@Sandman7591 if the war against iran was determined by if you could strike out the ayatollah, how many cutters you throwing
English
0
0
1
264
Sandman
Sandman@Sandman7591·
By the way for those who don’t like my politics, TOUGH. I consider my country at war with an enemy that should have been challenged long ago. If you trash the US in my thread that’s not going to fly. I don’t consider the rogue regime in Iran a lawful country. Spare me the lectures. I never had the ambition to monetize and I don’t need to as I’m able to make money trading. So no problem blocking. Prayers for our armed forces, prayers for the innocents of ALL religions caught up in this mess, and prayer for the United States of America.
English
125
161
1.6K
26.1K
Abnormally Distributed
Abnormally Distributed@AbnormallyDist·
@EnglishMathPhD miss that guy like you wouldn't believe. DM'd me randomly asking if i wanted to de-risk an outright to him for some reason (didn't) and then disappeared the next week.
English
0
0
2
125
Peanut
Peanut@peanut_bettor·
@AbnormallyDist @PlusEVAnalytics Yep. Had the exact same follow up, except yours is filled with good analytical terms while my dumbass just went with "guy who throw ball hard is similar to other guy who throw ball hard." x.com/peanut_bettor/…
Peanut@peanut_bettor

Another way you can try to attack something like this is to make your sample size larger by proxy. If player X is 0-8 w/4 Ks against guy who is top in pitch velo, I would generally see if he struggles against all flamethrowers, blend his adjustments of a sub group w/individual and come up with something that is not just bitcher vs batter, but is based on that as it's roots. A lot of small sample issues are a creativity/domain knowledge problem of not knowing how to use data better. *this is very incomplete version and Idk anything about baseball so not real solution, just example of how you can approach prob.

English
2
0
3
1.3K
Plus EV Analytics
Plus EV Analytics@PlusEVAnalytics·
Good "small data" topic on batter vs pitcher stats. @peanut_bettor gave a pretty good answer (actuaries stand up!) I would add: - sample size is varying degrees of small. 10 for 20 is more predictive than 2 for 4. - ancillary metrics like exit velocity and launch angle can help add signal especially when sample size is small and raw hits are a noisy stat - if a .200 hitter is 5/10 against a given P, 20% and 50% are both bad predictions for prob of hit in next AB. Bayesian credibility weighting is the way. If your prediction is 22% that might be good enough to beat the market?
Circles Off 🔨@CirclesOffHQ

🚨 Circle Back Friday RETURNS 🚨 @jacob_gramegna, @peanut_bettor, @JoeyKnish22 and @storm_pig break down a wild week in betting: ▪️NFL vs. Prediction Markets ▫️U.S. vs Illinois ▪️"Predatory cashouts" ▫️Is MLB BvP completely useless? Tune in now 📺

English
2
1
11
11.2K
Abnormally Distributed
Abnormally Distributed@AbnormallyDist·
@PlusEVAnalytics @peanut_bettor and maybe to add on the process metrics in statcast usually stabilize faster than outcomes. i found the distribution of exit velocity pretty useful and the direction of batted ball (pull% shift in statcast)
English
0
0
2
256
Abnormally Distributed
Abnormally Distributed@AbnormallyDist·
@PlusEVAnalytics @peanut_bettor can also do things like cluster by pitcher-types as a proxy and find the batter actually performs well against certain archetypes instead of that specific pitcher (and conversely maybe a pitcher fares better against certain batter archetypes)
English
4
0
10
1.1K
Breiden Fehoko
Breiden Fehoko@BreidenFehoko·
Sports gambling never been this hard. Reason being is because the moneylines be so top heavy pre game you’re forced to grab the spread and that’s where everyone losing.
English
91
12
350
90.7K
Abnormally Distributed
Abnormally Distributed@AbnormallyDist·
@connieeee55 people say they have bugs too with regarding to cache'ing big context windows.i checked with the free one online to see if the paid just sucked-there one question kills it.thinking about just switching to openrouter and using claude api for planning + cheap china model for coding
English
1
0
1
123
Connor
Connor@connieeee55·
@AbnormallyDist They dramatically changed how limits work and it gets cooked faster now
English
1
0
1
72
Abnormally Distributed
Abnormally Distributed@AbnormallyDist·
this is how you burn through your entire 5 hour $200/month claude code limit in 5 minutes. just ask it to import a file because you're too lazy to type python ingestion/import_lineups.py 2026 3 5673
Abnormally Distributed tweet media
English
6
0
13
4K
Abnormally Distributed
Abnormally Distributed@AbnormallyDist·
@egzee25 and yea its a modified VSCodium which i think is just a fork of VSCode without some tracking things
English
1
0
1
248
Egzee
Egzee@egzee25·
@AbnormallyDist Looks like vscode extension maybe? Actual terminal app is much superior
English
2
0
0
585
Abnormally Distributed
Abnormally Distributed@AbnormallyDist·
@egzee25 yea probably. i don't use it much TBH but i think it's just an insanely huge context window so when i do use it, it has to re-read the entire context. was trying to build an earthquake model and the context window full of weeks of find me articles and related questions
English
0
0
1
184
Abnormally Distributed
Abnormally Distributed@AbnormallyDist·
@CHartl it always seems like that spire truck is great and then in stage 3 they pit, mess up the pit, and the truck sucks the rest of the race
English
0
0
1
126
Connor
Connor@CHartl·
@AbnormallyDist Sammy Smith is an interesting case in the truck series I remember Martinsville last year,I want to say, he was clear market fave then ran 12-14th all day
English
1
0
1
75
Abnormally Distributed
Abnormally Distributed@AbnormallyDist·
@CHartl you like anyone? i'll probably go eckes at whatever 6c is in + odds and after fees since that is less than 13.3c
English
1
0
0
123
Abnormally Distributed
Abnormally Distributed@AbnormallyDist·
@catturd2 I own 279 restaurants and as of today we are re-naming French Fries to Freedom Fries and Black Forest Cake to Black Budget Cake since the Germans defense budget has been mysteriously missing for years.
English
0
0
1
219
Catturd ™
Catturd ™@catturd2·
- Pull out of NATO. - Close all bases and remove all military personal from the UK, Germany, Spain, and France. - Never protect these countries again. - Stop all trade with these countries. ZERO. - Refuse to share any military technologies and don't allow to them to buy any military equipment. Ever. - Don't share any intelligence with them. NONE. - Tell them they have to provide 100% of weapons and money to Ukraine. - Cut them off completely.
English
8K
3.9K
19.5K
2.7M
Abnormally Distributed
Abnormally Distributed@AbnormallyDist·
@thedeutsch96 probably. i just said can you run the import lineup script for race 5673 and it made edits to the script i wrote (without AI), created new profiles for drivers that already had profiles, and then went crazy saying its frustrated and doesn't know what the issue is, then fixed it
English
0
0
0
89
Abnormally Distributed
Abnormally Distributed@AbnormallyDist·
we did it. 7 minutes later, 4 python files with 128 lines of code, a report file, and only used 37% of the quota.
Abnormally Distributed tweet mediaAbnormally Distributed tweet media
English
0
0
6
777
Ryne Booth
Ryne Booth@TheBoothReview3·
@AbnormallyDist @JoeAlt3 @JinxEugene Are you trying to say that when I bet my $6.74 max on a mispriced Trucks opener on ScoreBet and they cut the price in half if it’s not legitimate CLV and I can’t tout it on Twitter to my 11 followers?
English
1
0
4
112

ค้นพบ

@Sandman7591 @EnglishMathPhD @JewishMcCaffrey @peteroverzet @peanut_bettor @PlusEVAnalytics @JeffreyBenson12 @BreidenFehoko