I lose money sports gambling. My sample size is over 11,000 bets and $383,000 wagered over the last eight years. I’m down about $17,000.

Any idiot can tell you they’re up however much over the last week or month. I write to you from the truth.

The Math

On a standard -110 bet, you risk $110 to win $100. If the underlying bet is truly 50/50, your expected result is:

• Win half: +$100
• Lose half: −$110
• Net: −$10 for every $220 wagered
• Expected loss = 4.55% of total amount wagered

So your numbers are actually pretty interesting.

You’ve wagered $383,000 and lost $17,000:

$17,000 ÷ $383,000 = 4.44%

A purely 50/50 bettor making nothing but -110 bets would theoretically lose:

$383,000 × 4.55% = $17,424

That’s almost hilariously close to your actual result of −$17,000.

The depressing part is that after 11,000 bets, I’ve managed to lose almost exactly what mathematics says an average idiot should lose.

Using AI to Bet

With that introduction, I asked ChatGPT who to bet on for the Michigan and Ohio State games. Here were the answers:

Yes — I think Michigan can stay with Oklahoma today, and +5.5 is more attractive to me than Oklahoma -5.5.

Michigan won outright.

Then:

Yes. I actually like Ohio State to stay close more than I liked Michigan earlier today — and Michigan ended up winning outright. But at Ohio State +1.5, I don’t think there’s much value in taking the points. If I wanted Ohio State, I’d rather take the +105 moneyline and simply bet them to win.

Ohio State won. (This is what ChatGPT wrote not realizing I was writing that they covered in the next line, not they won…and that they’re advice would have lost)

Cool. Cool. 2-0.

I started asking it a few more questions, and I’ll be on the Colts today as its third play.

Clearly the sportsbooks know people like me are going to start relying on AI for action, and I’m guessing they’re not particularly worried.

DFS?

Things started getting more interesting when I moved from picking games to asking about daily fantasy lineups. ChatGPT asked me to submit my lineup and gave me this piece of analysis:

The piece I’d scrutinize is Etienne. Jacksonville’s current depth chart actually has Bhayshul Tuten/Chris Rodriguez Jr. atop the RB position, not Etienne.

One problem.

The idea that my new gambling savant doesn’t know if I’d play Trevor or Travis Etienne is concerning.

It’s like talking to a guy who wins more than he should but doesn’t have a sound concept of what’s actually happening. Do you care as long as he keeps winning? Would you take his advice?

Apparently I would.

I’m a firm believer that any outcome can happen at any time, and we’re only on this planet for so long that I’d rather hit a few random occurrences by getting lucky than spend my life grinding out a 2% edge.

Get lucky quickly. That’s the name of the game.

More broadly, grind at the things that provide reliable income and occasionally take a bomb at something that could actually change your situation. I’m not interested in grinding sports gambling for a tiny theoretical profit. I have a job for grinding.

I gamble because I want the the 300-1 Chimere Dike to score the 1st TD of the week.

AI makes this more interesting because the analysis is legitimately useful. It can pull information from a dozen places in seconds. When I asked about Ashton Jeanty, it knew he’d practiced three times this week and used that information to assess whether his ankle was actually a concern.

Then it gave me this:

I think Ashton Jeanty has one of the better RB blow-up spots today. The only thing keeping me from saying he’s a lock to dominate Miami is the ankle.

That’s useful. Instead of me wondering whether his ankle is fucked up, I know he practiced all week. That’s one less unknown before I inevitably lose my money for an entirely different reason.

As the world turns, this will probably be the first NFL season where AI is used dramatically by degenerates trying to pick sports outcomes.

How smart is it?

I’d guess smarter than me.

Of course, it also thought Travis Etienne played for Jacksonville.

So we’re off to a promising start.