Choosing Your First Stakes: Micro-Stakes Economics, Winrates & Rake Traps (NL2 to NL10)

TL;DR: Start at NL2 or NL5 whatever your bankroll, because the first stakes are tuition for interface, pace, and player-pool knowledge. Micro-stakes rake often costs the average player roughly 8-12 bb/100, so a "small winner" before rake can be a loser after it. Hold 30-50 buy-ins for your level, and take shots at NL10 only with a real sample, solid stats, a defined stop-loss, and money you can afford to lose.

Most new players ask "what's the highest stake I can afford?" Sharps ask a different question: "what's the cheapest place to learn everything I need to know, and where does the math actually let me win?" Those questions lead to very different starting points.

Why Must You Start at NL2 or NL5 Even With Disposable Capital?

I hear this a lot: "I can afford to lose $1,000, so why grind NL2?" Because at the start, you aren't buying profit. You're buying information, and NL2 and NL5 sell it cheaply. Here is what you're learning:

• The interface. Misclicks, mis-sized bets, and timebank panics are expensive at $50 buy-ins and nearly free at $2. Learn your HUD, hotkeys, bet-sizing buttons, and multi-table layout while mistakes cost pennies.
• The pace. Online play deals 80-100 hands per hour per table. Four tables is 350+ hands an hour, so you'll face more decisions in a week than a live player sees in a year. Handling that volume without tilting is a skill you build over time.
• The player pool. Micro-stakes opponents are weirdly passive, call too much, and rarely fold top pair. Without that calibration, you'll bluff too often and misjudge raises. Learn how each pool actually behaves before you pay for lessons at higher stakes.
• Your own leaks. Tilt, impatience, and "just one more session" don't show up in theory. They show up in your database.

A bigger bankroll doesn't speed up this learning. It only raises the price. If you want a framework for matching stakes to skill and roll, see our stake selection guide.

The Micro-Stakes Rake Trap: What Is Eating Your Winrate?

Rake is the main economic fact of NL2-NL10. Most sites take around 4-5% of each pot, with a cap that is almost never reached because micro-stakes pots are small. In effect, the cap doesn't protect you, and the percentage applies to nearly every pot you play.

When you compare this against a typical 100 big blind buy-in, the cost is large. On many sites and tables, the average player pays an effective rake of roughly 8-12 bb/100, depending on site, table size, and promotions. Your numbers will vary, so check the actual figures. We list poker rake benchmarks at PokerMath.org so you can compare sites.

Here is why that matters. A good live-cash player might win 5-10 bb/100 against weak opponents. At NL2, the pool's collective losses are mostly going to the house, not to winning players. A break-even player before rake is a significant loser after it.

What Winrate Do You Need to Beat the Rake?

Run the numbers on a simple example. Suppose you're a competent, tight-aggressive player and your raw, pre-rake winrate is 14 bb/100. If you pay 10 bb/100 in rake, your net is 4 bb/100. That sounds small, but it is a real winner at NL2-NL10.

Now add rakeback. At 30% rakeback on 10 bb/100 of rake paid, you get 3 bb/100 back, taking your net to 7 bb/100. Rakeback often decides whether a micro-stakes player is profitable. For more on how this works, read rakeback explained.

Here's what to expect as a decent beginner:

• Break-even after rake: You're ahead of the pool on skill but not yet clearly profitable. This is normal for the first 20,000-30,000 hands.
• 2-5 bb/100 net: A solid result for a new, disciplined player at NL2-NL5.
• 5-8 bb/100 net: A strong player taking advantage of passive opponents, often with rakeback helping.
• 10+ bb/100 sustained: Rare over large samples. Don't plan your bankroll around it.

Variance will mislead you. Standard deviation in 6-max NL is roughly 80-100 bb/100. Over 50,000 hands, the standard error of your measured winrate is about 4 bb/100, so a 95% confidence range is roughly ±8 bb/100. A "5 bb/100" result after 50k hands could be a true 0 or a true 10. That's why I tell students to trust process more than short-term graphs.

Bankroll Requirements: 30-50 Buy-Ins at Every Level

Because variance is high and edges are small after rake, micro-stakes bankrolls need to be generous. The standard I teach is 30-50 buy-ins (100bb each) for the level you play.

• NL2: $60-$100
• NL5: $150-$250
• NL10: $300-$500

Use the low end (30 buy-ins) if you're disciplined, play a solid tight-aggressive game, have rakeback, and can reload if needed. Use the high end (50) if you're new, play many tables, or have a history of tilt. A 30 buy-in roll is a minimum, not a goal. Poor discipline wipes out more bankrolls than bad luck does.

Also, if you can't comfortably lose the money, you don't have a bankroll. You have a deposit with a time limit. Separate poker funds from rent, bills, and savings.

How Do Sharps Move From NL2 to NL5 to NL10?

Move up on a plan, not a feeling. A good sequence looks like this:

1. NL2 (about 10,000-20,000 hands): Learn the software, build a basic preflop strategy, and review hands. The goal is clean fundamentals, not profit.
2. NL5 (about 30,000+ hands): Play solid ranges, fold more marginal spots, and value-bet thinly against calling stations. You'll start to see more regulars, and the pool is a bit more competent.
3. NL10: Here you meet better regulars, more 3-bets, and more adjustments. Rake is still heavy, but you're closer to games where skill pays more consistently.

Move up only when your bankroll, sample, and stats agree. If they don't, stay put. There's no award for reaching NL10 first.

Shot-Taking at NL10: Precise Criteria

A shot means playing above your bankroll with a strict plan to protect yourself. Before you take one at NL10, check every item below:

• Sample: At least 30,000 hands at NL5 (or 50,000 combined across NL2/NL5) with a non-negative net winrate and a graph that isn't collapsing.
• Stats: Preflop numbers in a sensible 6-max range, roughly 21-25 VPIP, 17-21 PFR, 6-9% 3-bet. WTSD around 26-30% and W$SD above 50% usually indicate sound showdown play.
• Study: You review hands weekly and fix leaks, rather than just playing volume.
• Shot bankroll: At least 20 buy-ins ($200) set aside for the shot. Ideally 30.
• Stop-loss: Pre-commit to dropping back to NL5 after losing 10 buy-ins ($100) at NL10. Write the number down before you sit.
• Emotional readiness: The money isn't needed for anything else, and you aren't "chasing" losses at lower stakes.
• Tables and focus: Reduce your table count for the first sessions so you can observe the regulars.

If you hit the stop-loss, you haven't failed. You've collected data. Rebuild at NL5, review the hands, and try again later. If the shot works and you build a full 30-buy-in roll ($300) at NL10, you've earned the move.

What Is the Bottom Line?

Choose your first stakes for learning, not ego. Start at NL2 or NL5 even if you could afford more, because mistakes there are cheap and the lessons carry forward. Respect the rake: with 8-12 bb/100 going to the house, your edge has to cover it before you see profit, and rakeback matters. Hold 30-50 buy-ins, trust process over short-term results, and take shots at NL10 only after you meet the sample, stats, bankroll, and stop-loss criteria above. Sharps move up when the data says it's time, not when impatience does.