IRR: Card Counting vs. Poker Cash as Investments

Most advantage play discussions compare strategies on “how much you can make.” That’s the wrong frame. If you’re deciding where to allocate 20 hours a week and some amount of capital, the right question is internal rate of return — annualized return on the capital at risk, adjusted for risk and volatility. That’s how every other business is evaluated.

Let’s run the numbers on two classic AP strategies: counting blackjack at a mid-sized casino bankroll, and grinding $2/$5 live cash.

Setting the stakes

We’ll assume both players are skilled — in the 90th percentile of their discipline — and have the bankroll appropriate to their variance.

The counter

• Edge: 1.2% average (a good counter spread-wise, playing shoe games)
• Standard deviation: ~1.15 units per hand (blackjack variance is high)
• Hands/hour: 70 (low because of cover and spread timing)
• Unit size: $50 (typical mid-sized shop, allowing spreads to $400)
• Expected hourly: 0.012 × 70 × $50 × ~0.8 (cover drag) ≈ $34/hr
• Bankroll required: ~$25,000 to keep RoR around 1%
• Annual hours: 1,000 (hard to sustain more — shops heat up, you get backed off)

That’s about $34,000/year gross off $25,000 of bankroll, before travel, comp losses, and the non-trivial operational overhead (rotating venues, managing cover, dealing with heat).

The $2/$5 grinder

• Win rate: 5 BB/100 (strong live player)
• Standard deviation: 110 BB/100
• Hands/hour: 30
• BB value: $5
• Expected hourly: 5 × 30 / 100 × $5 = $7.50/hr

Wait, that can’t be right. Let me sanity check with a different framing: 5 BB/100 × 30 hands/hr = 1.5 BB/hr × $5 = $7.50/hr. Yep.

Hmm. That’s… not great. Let me bump the win rate assumption — a very good live $2/$5 player might run at 8–10 BB/100. At 10 BB/100:

• Expected hourly: 10 × 30/100 × $5 = $15/hr
• Bankroll required: ~$12,000 (1% RoR at that win rate / σ)
• Annual hours: 1,500 (easier to put in hours than counting)

$22,500/year gross off $12,000 of bankroll.

The IRR comparison

Strategy	Bankroll	Annual gross	Hours	Gross IRR	$/hr
Card counter (mid-size)	$25,000	$34,000	1,000	136%	$34
$2/$5 elite grinder	$12,000	$22,500	1,500	187%	$15

The IRR on bankroll actually favors poker. But the hourly favors counting. Which is the right metric?

It depends on what’s scarce in your life.

• If capital is scarce: poker wins. Lower bankroll, higher turnover on the dollars at risk.
• If time is scarce: counting wins. Every hour produces more, and the ceiling on hours is lower anyway.
• If opportunity is scarce: counting has a shelf life — each shop gets tighter over time, and online blackjack isn’t really an AP game. Poker’s TAM is bigger.

The hidden costs

Both numbers are gross. Taxes are the obvious deduction. Less obvious:

• Counter operational overhead: travel, multiple ID setups, rotating shops, cover play. Easily 20% drag on net hourly.
• Poker rake: at $2/$5 10%/$6 cap, your 5–10 BB/100 “win rate” is already after rake. But your downtime, broken games, bad tables — those aren’t in the hourly.
• Health costs: sitting 30 hours a week for years is real. Both disciplines have this, counting marginally more.

What this tells you

Neither strategy is a sane primary income at these stakes. Both are side-hustle-economics: 20–30 hours a week of focused effort for $15–40k annualized, with variance that can make any given quarter look bad.

Scale changes everything. At $5/$10, the poker IRR flatlines (variance scales linearly with stake, win rate in BB compresses, rake matters less). At $25/$50+ counting with bigger spreads and a team, the counter IRR can blow past poker. But neither is accessible to someone sitting down for the first time.

The practical takeaway: pick the discipline whose bottleneck matches your situation, not the one with the bigger absolute hourly number. And be honest about which bottleneck is actually yours.