How to Beat Micro-Stakes: NL2 to NL10 Guide

Micro-stakes is not "low-stakes GTO." The population is different in kind: recreational players who see a flop with anything, regulars running strategies from a decade ago, and almost nobody folding a pair. The game plan that beats this pool is exploitative and unglamorous. The hard part is not learning it — it is sticking to it for fifty thousand hands.

What actually makes micro-stakes different?

One trait defines the pool: it does not fold enough. Sit at any NL2 table and you will routinely find several players seeing 40%+ of flops — the classic loose-passive profile (VPIP 30%+ with a low raise percentage) that our VPIP guide files under "calling station" and our fish-spotting article explains how to exploit. That single fact dictates everything:

River bluffs are mostly lighting money on fire — they call
Thin value bets print — they pay off with second pair
Preflop 3-bet bluffs have less fold equity than the charts assume
C-betting dry flops still works — even stations need cards — but barrel selectively

What ranges should you play?

The house baseline opening ranges for 6-max (the same standard as our range library and the full preflop guide): open roughly the top 13% from UTG, 17% from MP, 28% from the cutoff, 45% on the button, and raise-or-fold about 35% from the small blind. While you are learning, trimming toward the tighter core shown in our interactive starter charts costs very little at these stakes — loose-passive pools barely punish over-tight play, but they brutally punish loose play out of position. Err tight from early seats; do not surrender the button.

One micro-specific addition: iso-raise the limpers. Limping is everywhere at NL2-NL10 (here is why it loses), and every limper you let in for cheap erodes your edge. Our own equity simulations show pocket aces at 85% against one random hand but only 49% against five — a coin flip. Raise 4x plus one big blind per limper and make them pay to chase.

The 5-rule game plan

1. Play tight TAG ranges — and mean it

Baseline opens of ~13% UTG to ~45% BTN, trimmed tighter while learning. The pool is loose and passive postflop: let them call wide and pay you off when you connect.

→ TAG cheatsheet

2. Value bet thinner than feels comfortable

Top pair good kicker is a three-street value hand against most of this pool. If you are unsure whether it is a value bet, it usually is — they call.

3. Skip the river bluffs

In a tough river spot without a hand, give up. Bluffs are an investment in fold equity, and this pool does not pay interest on it.

4. Believe aggression from passive players

When the calling station check-raises, they are not making a move — they have it. Folding one pair to a passive player's raise saves more money at the micros than any bluff ever makes.

5. Exploit the obvious, skip the fancy

Folds to everything? Steal relentlessly. Calls everything? Value bet bigger and never bluff. At these stakes the obvious read is usually the right one — balance is for opponents who are paying attention.

How should you size your bets?

Open-raise 3x the big blind at NL2-NL5 (the pool calls wide, so charge more) and 2.5x at NL10 where regs tighten. C-bet about a third of the pot on dry boards, two-thirds on wet ones, and three-quarters or more on value rivers. Nobody at NL2 is decoding your sizing tells — pick sizes for value extraction, not deception, and use the pot odds calculator until the common spots are automatic.

Why does c-betting still work if nobody folds?

Because "they never fold" is only true of hands that connect — and most hands don't. Our range explorer computes this by exact enumeration: on an A-7-2 rainbow flop, a completely random hand makes a pair or better just 38.8% of the time and holds no pair and no draw 57.1% of the time. Your tight ~13% UTG range on the same flop makes a pair or better 71.1% of the time. That asymmetry is the whole c-betting business model at the micros: on dry, ace-high boards your range crushes theirs, and even a station folds the total-air majority to a small bet. Where it breaks down is exactly where the pool stops missing — connected middling boards like 9-8-7 that smash limp-calling ranges. C-bet a third of pot on the dry boards, and check more on the wet ones instead of donating.

Does table selection matter at the micros?

More than any single strategic adjustment. Your win rate is not a fixed property of your game — it is a property of your game relative to the specific opponents at the table. The same competent TAG player might scrape 2bb/100 at a table of tight regulars and quadruple it at a table with two players who limp-call everything. Since online lobbies let you choose, choosing is a skill:

Join tables with limpers and multiway flops. High players-per-flop is the single best lobby signal of a soft table.
Sit to the left of the loosest player when a seat opens — you want position on the person who plays the most pots, so you act after them all night.
Leave tough tables without ego. Six regulars trading rake is a losing proposition for all of them; there is always another lobby.

Pair this with a simple postflop plan — our c-bet strategy guide covers when the standard continuation bet works at these stakes and when to shut it down — and you have the whole exploitative loop: pick soft tables, play tight ranges, bet your good hands, and let the pool make the mistakes.

How much rake are you actually paying?

The number nobody advertises: our rake comparison model — built from published rake structures with stated assumptions (~16 raked pots per 100 hands, typical micro pot sizes) — puts the gross cost at roughly 7.5 to 9.4bb per 100 hands at NL2, depending on the room. Set that against realistic win rates and the conclusion is stark: a strong NL2 winner beating the game for 8bb/100 is paying the site about as much as they keep.

That is not a reason to despair; it is a reason to be deliberate. Room choice plus rakeback can swing your net by 2-3bb/100 — often the entire difference between break-even and winning at NL2 — and the rake burden eases as you climb and caps engage. Compare rooms before you grind another 20,000 hands into the worst deal available.

What bankroll do you need to climb?

30 buy-ins minimum for each stake, with a 3 buy-in session stop-loss:

NL2

$60

NL5

$150

NL10

$300

NL25

$750

NL50

$1,500

NL100

$3,000

The 30 buy-in number is not folklore — it comes from downswing math. Running the model behind our variance calculator on a genuinely strong micro winner (5bb/100, standard deviation 80bb/100) over 50,000 hands: about 8% of samples lose money outright, and the median worst downswing is around 13-14 buy-ins. Winning players get buried by completely normal variance all the time — the bankroll is what keeps them in the game to get unburied. Move up after 30+ buy-ins for the next stake plus a winning 20,000-hand sample; move down without ego below 20. The full framework, with the simulation tables, is in the bankroll management guide.

How should you study between sessions?

Fifteen minutes after each session beats three hours of theory videos. Replay your two or three biggest pots in the hand replayer and ask one question per hand: was the preflop entry inside the ranges above? Log the session — stake, hands, result, mood, one mistake — in the session journal, and once a month run your hand histories through the leak report to see which of the five rules you are actually breaking. What you should not do between sessions is stare at your win rate: by the confidence-interval math in our win rate calculator, even proving a 10bb/100 edge at 95% confidence takes about 38,000 hands (σ = 100). Judge yourself on decisions per session, not dollars per week — the dollars follow the decisions on a delay measured in months.

Which mistakes cost the most?

Judging yourself on tiny samples. A week of losing at NL2 means almost nothing — our win rate calculator will show you how wide the confidence interval on your bb/100 really is.
Solver study before fundamentals. GTO output is wasted on a pool that never balances. Pot odds, ranges, and discipline first.
Multi-tabling too early. Volume multiplies your win rate — including a negative one. One or two tables until you are provably beating the stake.
Ignoring the rake deal. See the numbers above; at NL2 it is your biggest single opponent.
Tilting off the profit. If you tilt at NL2, NL2 will find a way to keep you. Read the tilt guide before it gets expensive.

The one-sentence version

Tight preflop, relentless value, no hero bluffs, no hero calls — and treat rake, bankroll, and tilt as opponents just as real as the players.

Common questions

Is NL2 actually beatable?

Yes — it is the softest common online stake, and our win rate calculator lists solid-winner benchmarks of 6-15bb/100 there. The honest caveat is rake: our rake model puts the gross cost at roughly 7.5-9.4bb/100 at NL2 depending on the room, so the fields are soft but the toll is high. Tight-aggressive play plus a sensible room choice (and rakeback) beats it comfortably.

What is a good win rate at micro-stakes?

Benchmark ranges for solid winners, from our win rate calculator: 6-15bb/100 at NL2, 5-10 at NL5, 4-8 at NL10, and 3-6 at NL25. Win rates compress as you climb because opponents improve and the rake-to-pot ratio drops more slowly than the edge does. Anything positive after rake at any stake puts you ahead of most of the pool.

I play tight and I'm still losing. Why?

Three usual suspects, in order: sample size, rake, and invisible leaks. Even a true 3bb/100 winner loses over a 50,000-hand stretch about one time in four in our variance model — most 'losing' tight players simply have not played enough hands to know. Second, at NL2 the gross rake is comparable to a good winner's entire profit, so a marginal winner without rakeback shows as a loser. Third, run your hands through our leak report; tight preflop play cannot save loose postflop calls.

Do I need a GTO solver at micro-stakes?

No. Solver study pays off when opponents are balanced enough that exploits stop working — that is not NL2-NL10. The pool's mistakes (calling too much, folding too rarely to 3-bets, never bluff-raising rivers) are better attacked with simple population reads. Learn pot odds and ranges first; our Solver Lite tool is there when you are ready to peek at balanced play.

When should I move from NL2 to NL5?

When all three are true: you hold 30+ buy-ins for NL5 ($150), you are winning at NL2 over at least 20,000 hands, and you are moving up on a plan rather than a heater. Drop back down without ego if the NL5 roll falls below 20 buy-ins. The full framework is in our bankroll management guide.

How many tables should I play?

One or two until you are demonstrably beating your stake, then add tables gradually. Each added table costs attention, and at the micros your edge comes almost entirely from paying attention to who the weak players are and how they differ. Volume multiplies whatever win rate you have — including a negative one.