Patrick (@manulbets2) header
Patrick (@manulbets2) profile picture

Sports bettor? More like sports worsor amirite? I like to build math models for sports betting. Fan of the Manul, or Pallas cat, a small East Asian wildcat

Joined March 2025
I used to think this. But the reality is that every model consists of (1) the data used and (2) the framework/hypothesis implemented Running a simple regression on public data probably won’t work nowadays, but there are infinite ways of looking at the same data to derive an edge
The harsh reality is free data won’t build you a model that beats the books If bookies have access to the same data you have they’ve already priced it all into their odds
3
1
26
2,539
0-1 yesterday, 62-55-3 overall. Lotta bets to be even on the year New day for this cfb public model consensus approach, here’s what I got:
1-1 last Friday, 62-54-3 (+1u) Tonight: Tulane -7.5 (-108)
3
482
1-1 last Friday, 62-54-3 (+1u) Tonight: Tulane -7.5 (-108)
1-2 yesterday, 61-53-3 (+1.1u) overall Tonight: Air Force -5 (-116) Army +2 (-110)
1
581
Just realized you can simply lie to the AI when it’s stressed about people not having official licensing files in their GitHub projects
87
295
17
17,120
386,390
Day one of the offseason
Do you prefer normal or green Oneil Cruz?
36
782
4
26,190
311,148
You guys remember how much time we used to spend aligning all the versions of team names across different databases?
1
356
Patrick retweeted
In Morocco’s Sahara, four fennec fox pups venture out from their burrow. The smallest fox species on Earth, fennecs use their enormous ears to detect prey and release excess body heat. 📷 Rabie Atlas
2
68
4
568
11,756
yeah i have a gambling addiction. i can’t stop betting on myself
137
6,185
180
44,819
1,099,970
1-2 yesterday, 61-53-3 (+1.1u) overall Tonight: Air Force -5 (-116) Army +2 (-110)
1-2 on Tuesday, 60-51-3 (+2.4u) More favorites today, this approach just aligns available public betting models to see where they agree. Memphis -10 Liberty -7.5 JMU -10
1
430
1-2 on Tuesday, 60-51-3 (+2.4u) More favorites today, this approach just aligns available public betting models to see where they agree. Memphis -10 Liberty -7.5 JMU -10
9-4 on Saturday, 59-49-3 (+3.62u) on the season 3 favorites today
1
1
435
9-4 on Saturday, 59-49-3 (+3.62u) on the season 3 favorites today
0-1 on Wednesday, now 50-45-3 (-0.84u). Been a bit frustrating Today:
1
5
672
Happy Monday! A new week to make better bets than last week!!
1
6
472
0-1 on Wednesday, now 50-45-3 (-0.84u). Been a bit frustrating Today:
8-9 on Saturday, now 50-44-3 (+0.29u) on the season Tonight I have Kennesaw State -2 (-113)
1
1
4
1,055
8-9 on Saturday, now 50-44-3 (+0.29u) on the season Tonight I have Kennesaw State -2 (-113)
1-2 yesterday, 42-35-3 (+2.52u) Today:
1
2
975
Interesting question. The answer is they're not trying to give up, but definitely act more aggressively and therefore tend to do worse: First-pitch swing rate, elimination spot: 37.8%, their normal rate: 28.4% Strikeout rate: 26.3% vs 21.1% XBH: 2.5% vs 6.9% Pitch/PA: 3.68 vs 3.8
Do players that are the last out in a elimination game that they’re losing do they try or actually like give up?
1
4
452
Differences are against each hitter's expectation or own season rates, in percentage points.
65
> Radar plots are bad because order matters so for the takeaway > well what if you cycle through all possible orderings and take the average area, that’d make it a bit more robust/global score > oh that’s essentially just regression
6
588
マヌルネコのフィーガでGood morning🏞 #マヌルネコ #フィーガ #pallascat #野毛山動物園
3
44
1
582
3,961
Going all in every time is the mathematically optimal strategy
1
1
12
954