When to Move Up in Stakes Without Blowing Up

Move up when your bankroll covers at least 30 buy-ins for the new stake, you have a stop-loss set before you sit down, and your win rate comes from a sample large enough to mean anything. Not when you have run hot for three weeks. Most players take their shot after 10,000 winning hands, and 10,000 hands tells you almost nothing about whether you beat the game you are already in.

What being ready actually means

Readiness is three separate gates, and players who blow up have usually cleared exactly one of them.

Bankroll is the arithmetic gate: enough buy-ins that a normal downswing cannot end your ability to play the stake. Edge is the evidence gate: proof from a real sample that you beat your current game. Process is the behaviour gate: you play the same way at 3am, two buy-ins down, as you do in your first hour.

A hot run satisfies none of them. It only makes all three feel satisfied at once, which is why it triggers so many badly timed shots. It is worth understanding what a heater that flatters your results does to a graph: several weeks of running above expectation is visually identical to improvement while it is happening.

Why 10,000 winning hands proves nothing

Standard deviation in 6-max online cash sits somewhere around 90bb/100 — higher in loose, aggressive games, lower in full ring and live. That number is what turns your results into noise.

Uncertainty shrinks with the square root of your sample, not with your sample.

Standard error = standard deviation ÷ √(hands ÷ 100). Double it for a rough 95% range.

Run that on a player whose true win rate is 5bb/100:

Hands played	Standard error	Rough 95% range around a 5bb/100 result
10,000	9.0bb/100	-13 to +23
25,000	5.7bb/100	-6.4 to +16.4
50,000	4.0bb/100	-3.0 to +13.0
100,000	2.9bb/100	-0.7 to +10.7
250,000	1.8bb/100	+1.4 to +8.6

At 10,000 hands, a 5bb/100 result is statistically consistent with being a losing player. At 50,000 hands you still cannot quite rule it out. Only past 100,000 hands does the range stop containing zero, and that assumes your edge held steady the whole way.

You can run these numbers against your own results and see the sample size your win rate needs before it means anything.

The conclusion is not that you must wait for 100,000 hands. It is that your graph is one weak input, so the other inputs have to carry the weight. Can you defend your decisions when you review them? Are you beating the regulars or only the recreational players? Do you know why your winning sessions won?

The bankroll gate: 30 buy-ins, not 12

Moving from 50NL to 100NL means the buy-in doubles and every downswing doubles with it. A 10 buy-in downswing that cost you $500 now costs $1,000, and 10 buy-ins is an ordinary week.

Thirty buy-ins for a shot, 40 to 50 before you call the stake yours. That means $3,000 to take a shot at 100NL and $4,000 to $5,000 to live there. Standard bankroll management rules will give you the full ladder, including the wider margins tournaments demand.

Rules for the shot itself

Decide all of this before you sit down, because none of it survives contact with a losing session otherwise.

• A stop-loss in buy-ins, not dollars. Two or three buy-ins down at the new stake and the session ends. No last table.
• A fixed length. Five thousand hands, or three sessions, then you evaluate. An open-ended shot becomes a permanent move by default.
• Fewer tables than usual. New stake, harder opponents, more decisions per hand. Cut your table count by a third for the first few sessions.
• The same game selection standards. If you would not sit in that lineup at your normal stake, it does not become acceptable because the blinds are bigger.
• No mid-session promotion. Winning three buy-ins at your regular stake is not a reason to open the next one up.

What actually changes at the next level

The step up is rarely about opponents playing an unrecognisably better game. It is about the specific leaks that stop being free.

Three-bets arrive more often. Your BTN and CO opens get punished, so a defence plan that consisted of folding everything but premiums starts costing real money.

Passive callers thin out. At micros a large share of your win rate is value betting people who cannot fold. Higher up, more of it has to come from not losing money in marginal spots.

Your own win rate drops. Expect to give back 2 to 4bb/100 in the first stretch. That is normal and temporary, but you should know what a realistic number is at each level so you can tell adjustment from failure — what a real win rate looks like at each stake is worth checking before you panic.

Rake as a share of the pot falls. This is the quiet reward for moving up, and over a large sample it is worth more than most players realise.

If any of those weaknesses is already costing you at your current level, the shot magnifies it rather than fixing it. There is a strong argument for beating micro stakes first over a real sample before you go looking for tougher games.

Moving back down early, and without drama

Set the number before you sit. If your bankroll drops back below your threshold, you move down that day. No negotiation, no one more session to win it back.

Moving down early is cheap. Moving down late is what ends bankrolls, because the delay is always powered by the belief that you are owed something.

Treat it as a scheduled event, not a verdict. Plenty of strong players have taken four or five shots at a level before it stuck, and the only difference between those and a blow-up is whether they left on the number they wrote down.

The pre-shot checklist

1. Bankroll: 30+ buy-ins for the new stake, after any money you need for rent.
2. Sample: 50,000 hands minimum at the current stake, recent, at a stable table count.
3. Result: clearly positive after rake and before rakeback, not propped up by one enormous month.
4. Review: you have gone through your own losing sessions and can name the errors.
5. Written rules: stop-loss, shot length, and the move-down number, decided while you were calm.

Clear all five and the shot is a business decision with a known worst case. Clear three of them because you just ran hot and it is a mood.