How big should my bankroll be before I move up in stakes?

For cash games, have at least 30 buy-ins (3,000 big blinds) for the new level. Use 40 or more if you play aggressive, high-variance styles or are in a tough player pool. For MTTs, 100-200 buy-ins is the standard because variance is much higher. Don't base the move on one good week. Also confirm you're a proven winner over a meaningful sample, ideally 20,000-50,000 hands at your current stakes.

What results and skill indicators show I'm ready to move up?

Look for a sustained positive win rate (for example, 3+ bb/100 in 6-max cash) over a large sample, backed by solid fundamentals. These include correct preflop ranges, sound c-bet and barrel frequencies, and the ability to explain your river decisions. Check your database for leaks, such as a poor fold-to-3-bet or over-folding to river bets. Then review hands with a coach or study group. Winning with a clean, stable stat profile is a better signal than a short heater.

How should I adjust my strategy when I move up?

Expect fewer big mistakes and more thoughtful opponents. Tighten your bluffing frequency in spots where players at lower stakes over-folded, and expect more 3-bets, float plays, and check-raises. Rely less on exploitative reads and more on a balanced, fundamentally sound game, then exploit confirmed tendencies. Take notes, use HUD stats, and study solver outputs for common spots. Don't expect the same easy edge. Your win rate will likely drop at first, which is normal.

What is a smart way to move up without risking my whole bankroll?

Use shot-taking with strict limits. Play a limited number of sessions or buy-ins at the higher level (for example, 5-10 buy-ins), while keeping your main bankroll at your regular stakes. If you lose your shot allotment, return to your old stakes without ego. You can also mix levels, playing mostly your regular stakes with some tables at the higher stakes, until your results and confidence justify a full move.

When should I move back down, and how do I handle the mental side of it?

Set a stop-loss rule in advance, such as moving down when your bankroll falls below 25 buy-ins for the higher stakes (or 20 if you're more conservative). Treat it as sound bankroll management, not failure. Downswings are normal, so judge your play on decision quality rather than short-term results. Review hands, avoid tilt, and keep session limits. If you feel scared money or frustration affecting your decisions, step down, rebuild, and retry when the numbers support it.

Moving Up Stakes: The Sharp's Protocol for NL50 and NL100+

TL;DR: Moving up is a capital-allocation decision, not a confidence exercise. Sharps take capped shots with a hard 5 buy-in stop-loss, adjust to pools that 3-bet 10-12% in the blinds, keep tables at 2-4 until decision quality is proven, and split work 3:1 between play and solver study. The solver tells you the baseline. Your reads tell you where to deviate.

Why Do Most Players Blow Up When Moving Up?

Amateur recreational players move up because they feel hot. Sharps move up because the math and the sample agree. The usual failure goes like this: a player crushes NL10 for 40k hands, jumps to NL100 with a thin roll, loses 3 buy-ins in two sessions, tilts, and re-deposits. The skill gap was real, but it was not the cause. The cause was risk of ruin and emotional bleed. Read emotional decisions before your first shot, because a downswing at a new stake is when the leaks show up.

Win rates drop as you climb. A 7 bb/100 micro-stakes crusher is often a 2-4 bb/100 player at NL50-NL100, and 6-max standard deviation runs around 85-100 bb/100. Over 5,000 hands, one standard deviation is roughly 640 bb, or about 6.4 buy-ins. A true 3 bb/100 winner still finishes that sample more than 5 buy-ins down roughly 15% of the time, and more often hits that mark at some point along the way. That is why the stop-loss exists. It is not a verdict on your skill. It is capital protection.

The Systematic Shot-Taking Protocol

Sharps use a written protocol. Squares use a feeling.

Core roll first: keep 25-30 buy-ins for your current stake untouched.
Dedicated shot fund: allocate only surplus profit, enough for 20 buy-ins at the new stake or less. Never fund a shot from the core roll.
The 5 buy-in shot-stop rule: if you lose 5 buy-ins at the new stake, you drop back immediately. No "one more session."
Re-entry condition: rebuild the fund, review 200 hands, and try again.
Success condition: promote to the new stake permanently only when you hold the full bankroll requirement.

SHOT FRAMEWORK (NL25 -> NL50 -> NL100)
Core roll (current stake)  : 25-30 BI
Shot fund (new stake)      : up to 20 BI, funded from profit
Stop-loss                  : -5 BI  => drop back
NL50 buy-in = $50   -> stop-loss  = $250
NL100 buy-in = $100 -> stop-loss  = $500
Min. sample before judging : 5,000 hands (variance, not verdict)

For the theory behind these numbers, see the PokerMath Kelly Criterion and bankroll math breakdown. Full Kelly is far too aggressive for a poker player with an uncertain win rate. Sharps use a fraction of it, which is why the shot fund is capped.

Key Concept

Your win rate is an estimate with a wide confidence interval. At 5,000 hands, the error bar around your measured bb/100 is larger than the win rate itself. The stop-loss caps the damage of being wrong about your own edge.

How Do Player Pools Change at NL50 and NL100+?

Below NL25, the average opponent under-3-bets and over-calls. From NL50 up, regs fight back. Expect blind 3-bet frequencies of 10-12% against the button and cutoff, and 4-bet bluffs that exploit your folds. If you keep opening 45% on the button and folding to every 3-bet, you are a donation account.

Here is the scenario. You open A♦5♦ on the button to 2.5 bb. The small blind 3-bets to 10 bb. Your math: you call 7.5 bb to win a pot of 13.5 bb, so you need about 36% equity. Against a 12% 3-bet range you do not have it, but A5s works as a 4-bet bluff because of the ace blocker. A 4-bet to 22 bb risks 19.5 more to win 13.5. That needs roughly 59% folds if you have zero equity when called, which is why the hands with blockers and some equity do the job.

The PokerMath 1,326 Combinations Monte Carlo study shows how unevenly equity is distributed across starting hands, which is why blockers and playability matter more than raw rank when you build 3-bet and 4-bet ranges.

Polarized vs. Linear 3-Betting

A linear range 3-bets the top of your continuing range: QQ+, AK, AQs, TT-JJ as value and little else. It works when the opener folds rarely and calls with worse hands, as in position against a loose button open. A polarized range uses strong value (QQ+, AK) plus bluffs like A5s-A2s, K5s and suited connectors, and folds the middle. It works out of position, where flatting is punished, and against opponents who 4-bet little.

SPOT                     | 3-BET FREQ | STRUCTURE
SB vs BTN open           | 12-14%     | Polarized (rarely flat)
BB vs BTN open           | 10-12%     | Mixed; flat the middle
BB vs CO open            |  9-11%     | Mostly linear + a few bluffs
BTN vs CO open (in pos.) |  8-10%     | Linear-leaning
Value combos (QQ+, AK)   | 34 combos  | AA 6, KK 6, QQ 6, AKs 4, AKo 12
Bluff combos (A5s-A2s)   | 16 combos  | 4 per hand, ace blockers

Use the PokerMath combinatorics and range analysis method to count them. The principle is that holding an ace removes 25% of the AA/AK combos your opponent can use to 4-bet you. Build your ranges visually with the preflop range chart, then pressure-test your bluff-to-value ratios in the EV calculator.

Sharp Tip: Tag every opponent who folds more than 65% to a 3-bet, and over-bluff them. Tag everyone who 4-bets more than 10%, and cut your light 3-bets in favor of calling in position. The exploit lives in the HUD, not in the textbook range.

How Many Tables Can You Play Without Eroding Your Edge?

Mass multi-tabling works when your edge per decision is tiny and consistent. At NL50+, edge comes from hard decisions in 3-bet pots, and those need attention. The sweet spot for most players is 2-4 tables, around 200-500 hands per hour. Beyond that, your time-bank decisions turn into autopilot clicks.

TABLES | HANDS/HR (6-max) | DECISION QUALITY | USE CASE
  1-2  | 120-200          | Maximum          | Learning a new stake
  3-4  | 250-500          | High             | Standard grind
  5-6  | 450-700          | Medium           | Only if win rate holds
  8+   | 700+             | Low/Erosion      | Micros, rakeback-driven

Test it empirically. Record your bb/100 at 4 tables and at 6 tables over 10,000 hands each. If 6 tables gives more dollars per hour at an equal win rate, scale up. Almost always, the win rate drops faster than the volume rises.

Manage the time bank on purpose: burn it on river decisions facing 30%+ pot bets and on 3-bet/4-bet pots, and auto-fold trash without thinking. If you catch yourself clicking "call" without computing, drop a table immediately.

How Should You Split Play and Solver Study?

Use a 3:1 play-to-study ratio. For every three hours at the tables, spend one hour in a solver. Play teaches you what happens, and study teaches you what should happen. Skip either and you plateau.

Your study hour should be structured. Pick the three costliest spots from your database (for example, BB vs BTN 3-bet pots or turn probes), solve a simplified tree, and extract one rule per session. Don't memorize frequencies. Understand why the solver bets small on A-high boards and why it checks back K -high boards. Ask what the solver is protecting against, then ask whether your opponent pool actually punishes that weakness.

How Do You Find Recreational Nodes?

A recreational node is a spot where the population deviates from solver play by more than 10 percentage points, and where you can exploit it without a big theoretical cost. Run a filter in your tracker for flop c-bet, fold to turn barrel, and river fold to a large bet. Sort by sample size (at least 300 hands) and by bb/100 lost relative to your average. The biggest gaps usually appear in three places:

Under-folding to turn barrels in single-raised pots, where recreational players call flop with any pair and any draw but give up on the river.
Over-folding to river overbets after passive lines, where capped ranges can't continue.
Passive blind defense, where recreational players check-call too wide and rarely check-raise.

The sharp approach is to solve the baseline, then shift one lever in the direction of the leak. If a player folds too much, bluff more. If a player calls too much, value bet thinner and cut the bluffs. The square approach is to copy solver frequencies against a player who is nowhere near equilibrium. That gives up the edge you moved up to collect.

What Does the Full Move-Up Protocol Look Like?

Everything above works only if the pieces run together. Here is the checklist we use with students.

Bankroll gate. Take a shot at NL50 or NL100+ only with a full buy-in cushion at your current level, and fix the size of the shot in advance.
5 BI shot-stop rule. If you lose 5 buy-ins at the new stake, move back down immediately. Five buy-ins is well within normal variance at these limits, so the stop-loss protects your bankroll and your decision quality, not your ego. Re-enter only after the next 10,000 hands of study and review.
3-bet adjustment. At NL50 and above you will often face 3-bet pools of 10-12%. Out of position, expect a more linear 3-bet range from regs and a more polarized one from aggressive players. Tighten your light calls, widen your 4-bet bluffs with blockers (A5s, A4s, and some KQo and suited wheel hands), and fold the weak offsuit broadways that look like bluff-catchers but have bad blockers.
Volume discipline. Stay at 2-4 tables while you learn the new stake. Mass volume pays only when your decisions are automatic and your win rate holds at the higher table count.
Review loop. Play three hours, study one, and write down one rule.

Why Do Blockers and Combinations Matter Here?

A 4-bet bluff is a combinatorics decision. Holding an ace removes premium combinations from the opponent's continuing range and shifts the balance toward folds. If you want to check the arithmetic, the PokerMath 1,326 Combinations Study shows how starting hands distribute across the full deck, and the PokerMath Range Analysis walks through counting combos to build and attack ranges. A hand like A5s blocks AA, AK and AQ. Those are the hands that continue against a 4-bet, so folds rise while you keep decent equity when called.

How Many Tables Are Too Many?

Decision fatigue is measurable. Most players see their error rate climb after about 90 minutes of continuous multi-tabling, and it climbs faster with each added table. Use this test: review 30 close decisions from the end of a session and compare them to 30 from the start. If the late ones are noticeably worse, cap your tables or shorten your sessions. A 4-table player making clean decisions will usually beat a 10-table player on autopilot at stakes where the regs adjust.

The Bottom Line

Moving up stakes is a process, not a gamble. The sharp player sets the shot size before the first hand, honors the 5 buy-in stop-loss without negotiation, and treats a downswing as data, not an insult. The square player moves up on a hot week, refuses to move back down, and calls the resulting losses "variance" when the real cause was decisions made out of position, on too many tables, against ranges he never studied.

The same contrast holds in every area we covered. Against 10-12% 3-bet pools, the sharp player counts combinations, picks blockers deliberately, and 4-bets with a plan. The square player calls out of position and hopes. In volume, the sharp player earns more tables by proving his decision quality holds up, while the square player adds tables to chase hourly rate and pays for it in leaks. In study, the sharp player keeps a 3:1 play-to-study ratio, extracts one rule per session, and uses it to exploit recreational nodes. The square player memorizes frequencies and applies them without asking who is sitting across the table.

Build the protocol, follow it for a full sample, and let the numbers decide when you are ready for the next level. The stakes will still be there when you are.