The Nonzero Year: Poker as a Decision Lab
In 2020 I lost $17,000 in the stock market.
It wasn’t bad luck. It was bad decisions: no process, no sizing rules, no way to tell a good call from a lucky one. I learned to invest properly after that and rebuilt. But the lesson that stuck wasn’t about stocks. It was that I’d never been taught how to decide.
Tesla money
This year I had $20,000 set aside to buy a Tesla in cash.
I’m not buying it. For 2027, that money is tuition for a different kind of education: one year of using poker as a decision lab.
The Bet
My investing is on track, and my long-term plan holds up whether or not this $20,000 sits in it for one more year.
In exchange, I get a skill: better, faster decisions under pressure, with an upside that’s hard to put a ceiling on. If I lost all $20,000, it would be no different from tuition at an overpriced university, except this time I choose the curriculum.
Most of life’s important decisions give slow, noisy feedback. You pick a career, a house, an investment, and find out years later whether it was right, with no clean way to separate your judgment from your luck.
Poker compresses that loop. Every hand is a decision under uncertainty with incomplete information, real money on the line, and an outcome you see in minutes. It forces the skill most people never practice: judging a decision by the quality of the thinking, not by how it turned out.
You can make the right call and lose. You can make the wrong call and win. Learning to tell the difference is the whole game, at the table and everywhere else.
The setup
- Cap: $20,000 for the year, $1,666.67 a month. When a month’s money is gone, I stop until the 1st. No borrowing forward.
- Winnings: chips I win stay in play until I leave. When I clock out, the profit is recorded and swept to a separate account. It never comes back to the table.
- Rules: ten hard rules, including leaving at 3 buy-ins down, no alcohol at the table, and no sessions on short sleep. The full list is on the hub page.
- Ledger: exact dollars, published monthly. Losses posted the same way as wins.
I’m a bartender, not a poker pro. I’m starting with almost no experience. October through December is study and preparation only. Day 1 is January 1, 2027.
My Prior: A 12-Month Distribution
Before I’ve played a single hand, here’s my best guess at how the year shakes out. It’s a prior, not a forecast: a starting estimate I’ll update in public every quarter as real hours, real win rate, and real costs come in. Updating beliefs with evidence is one of the skills I’m here to learn.
The rules shape the distribution. Losses are capped at $1,666.67 a month and $20,000 for the year. Winnings leave the table at the end of every session, so they don’t compound from one session to the next, and stakes only move up if I pass the gates on the rules page. The upside comes from big individual results, not from letting winnings ride.
Distribution Matrix
| Scenario | Probability | Cash at Year End | Net Result | The Reality |
|---|---|---|---|---|
| 1. The Bad End | 7% | $8,000 | -$12,000 | Severe negative variance, cold decks, and coolers. Monthly stop-losses trigger. $12k for a live-fire course in drawdown control. |
| 2. Paying Tuition | 43% | $16,000 | -$4,000 | A mix of coolers, minor leaks, and seat fees. $4k for hundreds of hours of high-pressure reps. |
| 3. Coming Out Ahead | 43% | $26,000 | +$6,000 | The edge shows up. Table selection works, leaks get fixed, and the $20k comes back with $6k on top. |
| 4. The Sun-Run | 7% | $55,000+ | +$35,000+ | A deep tournament run or a long heater. Rare, but tournaments give real right-tail upside even at capped stakes. |
| Total | 100% | — | — | Full probability space accounted for. |
Cash at year end = what’s left in the bankroll plus everything swept to the winnings account.
Mathematical Expectation
Expected value is each scenario’s dollar outcome multiplied by its probability, added together.
1. Expected Cash at Year End
- Bad End: 0.07 × $8,000 = $560
- Paying Tuition: 0.43 × $16,000 = $6,880
- Coming Out Ahead: 0.43 × $26,000 = $11,180
- The Sun-Run: 0.07 × $55,000 = $3,850
Expected cash = $560 + $6,880 + $11,180 + $3,850 = $22,470
2. Net Expected Value
Net EV = $22,470 − $20,000 = +$2,470 (+12.35% on the $20k)
What the Numbers Say (and Don’t)
- Worst case: −$20,000. That’s the hard cap, and it only happens if I lose every session of the year. Every winning session pushes the real floor up. The Bad End is my worst realistic case.
- My prior is mildly positive: +$2,470. Most of that comes from the 7% tail. Across the middle 86% of outcomes, it’s close to a coin flip between paying $4,000 and making $6,000.
- Seat fees are the number to beat. Local clubs charge roughly $12–18 an hour just to sit. My win rate has to clear that before I can call myself a winner.
- I may be overrating myself. Starting from almost no experience, a 43% chance of finishing ahead could be optimistic. If it is, the ledger will show it, and I’ll say so. I give myself an edge because of my people reading skills and prior knowledge of psychology, math, mental make up.
What counts as winning
Not the money. The money gets tracked honestly, but it isn’t the scorecard.
The scorecard is whether I come out of 2027 a measurably better decision-maker: clearer about expected value, calmer under pressure, and more honest about the difference between a good decision and a good outcome. Every week I’ll take one poker lesson and apply it to a real decision outside the game. That transfer is the point.
The downside is capped at $20,000. The upside is open.
Follow along
The rules, the ledger, and links to every platform are at rudygalan.com/nonzero. Every session gets logged within 24 hours. After each month closes, the numbers, my best and worst decision, and one lesson that carried over to life go out by email. The signup is at the bottom of this page.
The Real Comparison: Tesla, Index Fund, or Decision Lab
The honest alternative to buying the car isn’t only poker. It’s also leaving the $20,000 invested. So here are all three.
| Used Tesla ($20k) | Index Fund | Poker Decision Lab | |
|---|---|---|---|
| Expected cash, year 1 | −$2,000 to −$3,000 (depreciation) | +$1,400 to +$2,000 (7–10% long-run average) | +$2,470 (my prior, unproven) |
| A bad year | Repairs, an accident, faster depreciation | −$7,000 or worse (a 2008-style drop) | −$12,000 realistic, −$20,000 cap |
| How reliable is that estimate? | Very | Decades of data | A guess until January |
| What else I get | A year of driving a car I like | Nothing; it just compounds | Hundreds of hours of decisions under pressure |
| Feedback speed | None | Years | Minutes |
| Skill acquired | None | None | Practice holding a plan through drawdowns. That’s the goal; I haven’t proven it yet. |
The Verdict
On cash alone, this isn’t a slam dunk. The index fund’s expected return is smaller than my poker prior, but it’s far more reliable; my number is a guess made before playing a hand. The car is the only option with a near-certain loss, and also the only one I can drive.
What tips it for me is everything that doesn’t show up on a statement. Neither the car nor the fund teaches me anything. The lab gives me a fast, honest feedback loop on how I decide, and I’ll carry that into every investment, career move, and big purchase for the rest of my life. That’s worth paying for, and the cost is capped at $20,000.
I’ll find out whether I priced it right. That’s the point.