时地方
This commit is contained in:
@@ -0,0 +1,138 @@
|
||||
---
|
||||
title: "Math Over Gambling: How Quants Make Millions on Sports Betting"
|
||||
source: "https://x.com/goatyishere/status/2044003377474867481"
|
||||
author:
|
||||
- "[[@goatyishere]]"
|
||||
published: 2026-04-14
|
||||
created: 2026-04-26
|
||||
description: "90% of people bet on sports like it’s a casino, and they loseThe other 10% are playing a completely different game.They don’t guess, they ca..."
|
||||
tags:
|
||||
- "clippings"
|
||||
---
|
||||

|
||||
|
||||
90% of people bet on sports like it’s a casino, and they lose
|
||||
|
||||
The other 10% are playing a completely different game.They don’t guess, they calculate
|
||||
|
||||
Probability. Edge. Expected Value
|
||||
|
||||
That’s how millions are made
|
||||
|
||||
One of the people breaking this down is Alex Monahan (Stanford, ex-trader at Susquehanna International Group, founder of OddsJam, sold for [$160M](https://x.com/search?q=%24160M&src=cashtag_click))
|
||||
|
||||
His lecture isn’t about sports. It’s about decision-making through math
|
||||
|
||||
## The Core Idea
|
||||
|
||||
Sports betting = true probability vs implied probability
|
||||
|
||||
If you don’t understand the difference, you’re paying the market
|
||||
|
||||
## 1/ The Book Always Wins
|
||||
|
||||
Odds are never “fair”.
|
||||
|
||||
Sportsbooks bake in **vig** (their edge)
|
||||
|
||||
Example: 50/50 event → -110 / -110
|
||||
|
||||
That means:
|
||||
|
||||
- each side ≈ 52.4%
|
||||
- total >100% → the difference = bookmaker profit
|
||||
|
||||
You’re playing against a system that’s already winning
|
||||
|
||||
## 2/ Where the Money Is
|
||||
|
||||
Your job is to remove the vig (de-vig) and compare:
|
||||
|
||||
- if your probability > implied → **+EV bet**
|
||||
- if not → skip
|
||||
|
||||
Example: -150 = ~60% implied if your model says 64% → that’s edge
|
||||
|
||||
Every bet = a small mathematical advantage
|
||||
|
||||
## 3/ Arbitrage = Risk-Free Profit
|
||||
|
||||
Different books = different prices
|
||||
|
||||
You bet all outcomes → lock in profit
|
||||
|
||||
Typical ROI: 5-10% in minutes
|
||||
|
||||
But:
|
||||
|
||||
- opportunities are rare
|
||||
- they disappear fast
|
||||
- books limit winners
|
||||
|
||||
This is trading, not gambling
|
||||
|
||||
## 4/ How the Market Actually Works
|
||||
|
||||
Sportsbooks = market makers
|
||||
|
||||
- balance money on both sides
|
||||
- move lines when there’s pressure
|
||||
- react to sharp bettors
|
||||
|
||||
Sharps move the market Public pays for emotions
|
||||
|
||||
This behaves exactly like financial markets
|
||||
|
||||
## 5/ +EV Is Everything
|
||||
|
||||
You don’t need to know sports better than everyone
|
||||
|
||||
You need to:
|
||||
|
||||
- find mispriced odds
|
||||
- compare across books
|
||||
- bet only when EV is positive
|
||||
|
||||
Strategies:
|
||||
|
||||
- **Middles** (win both sides in certain outcomes)
|
||||
- slow-moving lines
|
||||
- cross-book discrepancies
|
||||
|
||||
## 6/ Why Most People Lose
|
||||
|
||||
- vig eats your bankroll
|
||||
- parlays amplify losses
|
||||
- prop bets often have 20%+ house edge
|
||||
|
||||
You can be right often and still lose money
|
||||
|
||||
## 7/ The Real Skill
|
||||
|
||||
**Decision-making under uncertainty**
|
||||
|
||||
You never know the outcome But you can make +EV decisions
|
||||
|
||||
This applies everywhere:
|
||||
|
||||
- trading
|
||||
- business
|
||||
- life
|
||||
|
||||
## 8/ Final Take
|
||||
|
||||
- 95% of bettors lose
|
||||
- not because of luck, because of no edge
|
||||
- pros play probability, not sports
|
||||
|
||||
Betting stops being gambling once you understand the math
|
||||
|
||||
## Full Lecture (Highly Recommended)
|
||||
|
||||
Massachusetts Institute of Technology lecture by Alex Monahan
|
||||
|
||||
🎓 [https://www.youtube.com/watch?v=6mTvAjmF2-A](https://www.youtube.com/watch?v=6mTvAjmF2-A)
|
||||
|
||||
Save this
|
||||
|
||||
In the long run, the winner isn’t the one who guesses right, it’s the one who has edge
|
||||
Reference in New Issue
Block a user