Poker Variance & Downswings — How to Survive Bad Runs in 2026

You're a winning player. You know your ranges, you respect pot odds, your preflop charts are solid. And yet you've just lost 15 buy-ins over 30,000 hands and you're starting to wonder if you've forgotten how to play poker. You haven't. You're experiencing variance — the unavoidable statistical reality that separates short-term results from long-term truth. This guide walks through the math behind downswings, the bankroll requirements to survive them, and the mental framework to stay grounded when the cards run cold.

1 What Is Variance in Poker?

In statistics, variance measures how spread out a set of results is around the average. In poker, it quantifies how much your actual results deviate from your theoretical expected win rate over any given sample of hands. The higher your variance, the wider the range of outcomes you can experience — even when playing correctly every single time.

The standard mathematical expression of variance in poker is standard deviation (SD) measured in bb/100 hands. For a typical 6-max cash game player, standard deviation runs between 60 and 120 bb/100, with the exact number depending on playing style. Aggressive, high-volume players who play large pots frequently will have higher standard deviations. Tight, positionally sound players will have lower ones — but nobody escapes it entirely.

Here's the critical insight that every poker player must internalize: a positive expected value outcome does not guarantee a positive result in any individual session, week, month, or even multi-month stretch. The expected value of getting your stack in as a 70/30 favorite is positive — but 30% of the time you lose. Multiply that across hundreds of all-in confrontations and the negative outcomes cluster in time, producing stretches that feel catastrophically bad despite being entirely within statistical expectation.

The Variance Formula

Your expected results over N hands, expressed as a confidence interval:

Expected range = (win_rate × N/100) ± (SD × √(N/100))

For a 3bb/100 winner with 80bb/100 SD over 10,000 hands: expected result = 300bb ± 800bb (1 standard deviation). That means results between -500bb and +1,100bb are all within a single standard deviation — a perfectly normal range of outcomes, despite the 1,100bb spread.

Why Short Samples Lie

Most poker players dramatically underestimate how many hands are required before results become statistically meaningful. At 1 standard deviation of confidence, you need approximately 100,000 hands before your observed win rate narrows to within ±2bb/100 of your true win rate. At 2 standard deviations (95% confidence), the required sample grows to 250,000+ hands.

The practical implication: a 10,000-hand sample — which many players think of as a large dataset — is almost statistically meaningless for determining your true win rate. A player running at -5bb/100 over 10,000 hands might be a breakeven player who ran two standard deviations below expectation. They might also be playing poorly. The number alone cannot tell you which.

Variance by Game Type

Not all poker formats are created equal in terms of variance. Omaha variants produce significantly higher variance than hold'em because players connect with the board more often, pots go to showdown more frequently, and bad beat setups (second-best full houses, non-nut flushes) are structurally more common. Rush Poker (fast-fold) produces higher hand volume but the same per-hand variance, simply compressing the downswing timeline. MTT variance is the most extreme — a single tournament can represent hundreds of hours of expected edge that simply fails to materialize.

2 How Bad Can Downswings Get?

The honest answer is: far worse than most players expect, and far more common than feels statistically fair. Downswing simulators using Monte Carlo methods consistently show that severe downswings are not outliers — they are near-certainties over a long enough poker career.

The table below shows typical maximum downswing expectations by win rate, assuming a 6-max cash game player with 80bb/100 standard deviation. These figures represent the downswing length that approximately 50% of players at each win rate will experience at some point in their career — meaning half of all players at each win rate will experience something this bad or worse.

Win Rate	Typical Max Downswing	Max Buy-Ins Lost	Breakeven Stretch
1 bb/100	60,000–80,000 hands	15–25 buy-ins	100K+ hands possible
2 bb/100	40,000–60,000 hands	10–20 buy-ins	50K–80K hands
3 bb/100	25,000–40,000 hands	8–15 buy-ins	30K–50K hands
5 bb/100	15,000–25,000 hands	6–12 buy-ins	20K–35K hands
8 bb/100	8,000–15,000 hands	4–8 buy-ins	10K–20K hands
10+ bb/100	5,000–10,000 hands	2–6 buy-ins	5K–15K hands

Note: Assumes 100bb effective stacks, 80bb/100 standard deviation (typical 6-max aggressive player), and defines "downswing" as peak-to-trough drawdown in bb. Tournament players face 3–5x these figures due to higher variance structure.

Common Mistake: Quitting Too Early

The single most common self-destructive response to a downswing is quitting poker — or drastically changing strategy — before the sample is large enough to distinguish variance from a real win rate problem. A 2bb/100 winner who has lost 10 buy-ins over 20,000 hands is experiencing something that happens to roughly 30% of players at that win rate. They haven't suddenly become a losing player. They're inside a completely normal distribution tail.

Tournament Variance: The Harshest Version

MTT variance makes cash game variance look mild. In a standard 100-player tournament with a top-heavy payout structure, the majority of your EV is concentrated in the top 3 finishes. Going deep but not winning — 4th, 5th, 10th — is statistically the worst outcome from a variance perspective because you invest full time and rake but capture none of the EV-heavy payouts. A winning tournament player with 20% ROI can easily experience 200-tournament losing stretches. That's not a sign of a bad player. It's the math of high-variance structures.

3 Bankroll Management for Variance

Bankroll management is not a conservative strategy for weak players. It is the mathematical framework that determines whether a positive-EV player survives long enough to realize their edge. Without sufficient bankroll, even a winning player can go broke — not because they played badly, but because variance hit before the edge had time to manifest.

Cash Game Bankroll Requirements

The standard recommendation for 6-max cash games is 20–30 buy-ins at your target stake. More precisely:

20 buy-ins: Minimum viable. Risk of ruin at 5bb/100 win rate is approximately 20%. Suitable only if you have income outside poker and can reload.
25 buy-ins: Standard recommendation. Risk of ruin drops to roughly 12% at 5bb/100.
30 buy-ins: Conservative and correct. Risk of ruin below 5% for most win rates. This is the target before moving up stakes.
50 buy-ins: Necessary if your win rate is below 3bb/100, or if you play PLO where variance is 20–40% higher than NLHE.

Risk of Ruin Formula

Risk of ruin (RoR) is the probability that your bankroll reaches zero before your edge prevails. The approximation formula:

RoR = e^(-2 × win_rate × bankroll / SD²)

Where win_rate and SD are both in bb/100, and bankroll is in bb. For a 5bb/100 winner with 80bb/100 SD and a 2,500bb bankroll (25 buy-ins at 100bb): RoR = e^(-2 × 5 × 2500 / 6400) ≈ 9%. That means roughly 1 in 11 players with this exact profile will go broke before recovering, even though they are winning players.

Tournament Bankroll Requirements

Tournament poker requires dramatically larger bankrolls relative to buy-in size, due to the high-variance payout structure. General guidelines:

Regular MTTs (100–500 players): 100–150 buy-ins for a 15–20% ROI player.
Large-field MTTs (500+ players): 150–200 buy-ins. Field size multiplies variance directly.
Satellites: 30–50 entries. Variance is lower due to flat payout structure.
Mixed MTT/Cash schedule: Maintain cash bankroll separately; treat tournament buy-ins as a % of cash roll (e.g., no tournament buy-in exceeds 2% of total bankroll).

Move Down Rules: Non-Negotiable

The most important bankroll discipline is having pre-committed move-down rules before a downswing starts, not during it. Deciding to move down while you're emotionally invested in recovering at your current stake is a decision made under duress. Write the rules in advance:

If my roll drops below 20 buy-ins for my current stake → move down one stake immediately, no exceptions.
Return to current stake when roll rebuilds to 30 buy-ins.
No shot-taking (playing above your stake) unless roll is at 40+ buy-ins for the higher stake.

The Pride Tax

Refusing to move down due to ego or the belief that "I'm too good for this stake" is one of the most expensive decisions in poker. The math does not care about your reputation. A $200NL player running badly who refuses to drop to $100NL risks going broke at $200NL — which then forces them to play $25NL just to rebuild from a much smaller base. The cost of pride during downswings is measured in buy-ins, time, and sometimes the game itself.

4 Signs You're Running Bad vs Playing Bad

The most intellectually honest question to ask during a downswing is: "Is this variance, or have I developed leaks?" Both are possible. Both require different responses. Running bad while thinking you're playing fine is dangerous because it prevents necessary self-correction. Playing badly while blaming variance is equally dangerous because it perpetuates the leaks. The distinction matters enormously.

Check Your HUD Statistics

The most reliable diagnostic tool is your own HUD data compared against your historical baseline. Pull your stats from a period you know you were playing well and compare them to the current downswing period. Significant deviations in key metrics are early indicators of tilt-induced play degradation:

VPIP

Voluntarily Put Money In Pot

If your VPIP is 3–5% higher than your baseline, you're playing too many hands — likely widening preflop ranges under frustration. This is one of the earliest tilt signals.

PFR

Preflop Raise %

PFR significantly lower than baseline (VPIP–PFR gap widening) indicates passive tilt — limping and calling more, raising less. PFR significantly higher can indicate aggro-tilt: over-bluffing and reckless 3-betting.

3B%

3-Bet Frequency

3-bet % drifting more than 2% from your baseline in either direction is notable. Down suggests fear of getting it in; up suggests you're bluffing recklessly trying to "force" results.

Aggression Factor

AF dropping substantially below baseline is a major red flag. It means you're betting and raising far less relative to calling — checking down rivers you should be betting, calling rather than raising strong hands, generally playing smaller and more passively.

The All-In EV Graph

If your poker tracker shows an All-In EV (expected value) graph, this is the most direct measure of whether you're running bad or playing bad. The All-In EV line shows what your results would have been if every all-in confrontation had resolved at its exact equity percentage — removing luck from those spots entirely. If your actual results are significantly below your All-In EV line over a substantial sample, you are genuinely running bad. If your results are tracking close to or above the All-In EV line but you're still losing, the problem is not variance — it's your non-all-in play.

Behavioral Tilt Indicators

Beyond statistics, honest self-observation reveals tilt before the data can. Warning signs that you're playing below your standard:

Sessions consistently running 20–30% longer than your intended stop time
Irritation or anxiety between hands (refreshing the lobby obsessively, physical tension)
Internal monologue that uses words like "deserve," "unfair," or "rigged"
Making thin hero calls to "catch bluffs" more than once per session
Sizing up mid-session to recoup losses faster
Reduced hand-reading — acting on gut feeling rather than ranging opponents

5 Mental Strategies for Surviving Downswings

Surviving a downswing is a mental endurance event as much as a technical one. The math will eventually resolve in your favor — your job is to stay solvent, stay skilled, and stay sane until it does. Here are the strategies that separate players who weather downswings from those who break under them.

1. Process-Oriented Thinking

The most durable mental shift is redefining what "winning" means on a session-by-session basis. A winning session is one where you made correct decisions with the information available — regardless of what the cards did. A losing session where you played your A-game is, by this definition, a success. A winning session where you ran hot while playing sloppily is, by this definition, a partial failure. This reframe removes the emotional power of short-term results and replaces it with something you can actually control: the quality of your decision-making.

The practical tool for this is a brief post-session self-rating. Before you even look at the dollar result, rate your decision quality on a scale of 1–10. Did you stick to your ranges? Did you respect your stop-loss? Did you play your best poker? Write it down. Over time, the correlation between high decision-quality ratings and positive results (over sufficient sample) becomes visible in your own data — reinforcing the process-outcome connection at a personal level.

2. Pre-Set Session Limits

Decide your stop-loss and session length before you sit down, when you're not emotionally invested. Standard guidelines:

Stop-loss: Quit after losing 2–3 buy-ins in a single session. This is not weakness — it is recognizing that consecutive losses in one session often indicate either a tilt spiral or an unusually bad table dynamic, neither of which is worth persisting through.
Time limit: Set a maximum session length of 2–4 hours for most players. Fatigue degrades decision quality significantly after hour 3–4, and the EV of a tired session is negative regardless of the cards.
Win goal: Optional but useful. Some players benefit from quitting a fixed amount above their session start — preventing winner's tilt from eroding a good result in the session's final hours.

3. Strategic Breaks

During a downswing, the temptation is to play more — to grind through it, accumulate hands faster, get back to breakeven sooner. This is almost always the wrong instinct. Playing more while tilted accelerates losses. Instead, build strategic breaks into your routine:

After a bad session: Take at least 24 hours before your next session, regardless of how badly you want to "get it back."
Weekly off day: Designate one full day per week as a no-poker day during downswings. The mental reset it provides is worth more than the volume lost.
Extended break threshold: If you've lost 10+ buy-ins without a win, a 3–7 day complete break often resets perspective more effectively than any amount of "pushing through."

4. Study During Downswings

A downswing is paradoxically the best time to study — not the worst. Here is why: you have more emotional motivation to understand what's happening, you're more receptive to identifying leaks than during a heater when everything feels fine, and studying fills the psychological need to "do something about it" in a productive rather than destructive way. Redirect the energy that would go into extended tilt sessions into solver review, hand history analysis, and range work instead. Players who emerge from downswings as better technical players than they entered are the ones who treated the downswing as forced study time.

5. Perspective: The Long-Run Is Real

One of the most effective mental anchors is a simple mathematical visualization. If you have a confirmed 3bb/100 win rate over 200,000 hands, your current 30,000-hand downswing is a blip on that graph — invisible at full scale. Pull up your lifetime graph and zoom out. The downswing that feels catastrophic at 1:1 scale often looks like a minor wrinkle against the backdrop of your full sample. When you cannot zoom out (because your sample is still small), remind yourself that you are playing a probabilistic game and the expected value of every correct decision remains positive regardless of what happened in the last session.

The Variance Reframe

Variance works in both directions. The same mathematical process that produces your worst 20,000-hand downswing is exactly what produces your best 20,000-hand heater. The player who is mentally stable enough to stay in action — properly bankrolled, sticking to their stop-loss rules, continuing to study — will experience both. The player who quits or goes broke during the downswing never experiences the heater that was statistically due.

6 Using Provably Fair to Verify Your Hands During a Downswing

There is a thought that surfaces in the mind of nearly every online poker player during a brutal downswing: "Is this site rigged?" It is a natural psychological response to a pattern of losses that feels statistically impossible. And on every online poker platform in existence — except one — that thought cannot be conclusively answered. You have to trust the site's word that their RNG is fair.

PlasmaPoker is the only platform where you don't have to trust anyone. Every hand generates a unique SHA-256 audit hash before the cards are dealt. The hash is a cryptographic commitment: the server cannot alter the deck after committing to the hash without the alteration being mathematically detectable. After the hand completes, you receive the server seed, your client seed, and the nonce — and you can verify independently that the deck was generated honestly.

How to Verify a Hand

Accessing the verification tool is straightforward:

Open your Hand History from the lobby or from any table. Select the hand you want to verify.
Click "Verify Fairness" — this opens the Provably Fair Verifier with the hand's seeds and hash pre-populated.
Run the verification — the tool applies the SHA-256 algorithm using the same logic as the server and confirms whether the deck output matches the pre-committed hash.
The result is mathematical truth. If the hash matches, the deck was generated fairly. If it doesn't match, something was tampered with — but that cannot happen without being detected.

For players in the depths of a 15 buy-in downswing, this tool provides something genuinely valuable: certainty. Not reassurance, not a customer support ticket, not a "we take fairness seriously" boilerplate response. Mathematical, cryptographic, independently verifiable certainty that the deck was not manipulated. When you know beyond doubt that the cards were fair, the "rigged" thought loses its power — and you can redirect that mental energy toward the actual problem, which is surviving and playing through statistical variance.

"Don't Trust Us. Verify Us."

This is PlasmaPoker's commitment to every player. No competitor — not Global Poker, not ClubGG, not PokerStars Play, not WSOP — offers cryptographic hand verification. They cannot answer the "rigged" question. We built the system so you never have to ask it.

Try the Provably Fair Verifier →

The Psychological Value of Certainty

Research on decision-making under uncertainty consistently shows that ambiguity is more psychologically distressing than known-bad outcomes. Players who can definitively answer "was I cheated?" — even when the answer confirms they simply ran badly — report faster emotional recovery from downswings than players left with unresolved suspicion. The provably fair system resolves that ambiguity completely. Your downswing was variance. The math says so. Now you can focus on playing through it rather than ruminating on whether the game was honest.

PlasmaPoker also includes built-in coaching tips for players who are running badly — surfacing relevant strategy reminders and variance perspective notes directly at the table after extended losing streaks. Combined with the free built-in HUD (VPIP, PFR, AF, 3Bet%, CBet%) for tracking your own stat deviations, PlasmaPoker gives you every tool needed to distinguish variance from a genuine leak — without paying for a single third-party subscription.

? Frequently Asked Questions

What is variance in poker?

Variance in poker is the statistical measure of how much your results deviate from your expected win rate over any given sample of hands. Even a player with a positive long-run win rate will experience substantial negative stretches due to the inherent randomness of card distribution. Variance is typically expressed as standard deviation in bb/100, and for most 6-max cash players this ranges between 60 and 120 bb/100 depending on playing style. Understanding variance mathematically — not just conceptually — is essential to staying grounded during bad runs.

How long do poker downswings last?

Far longer than most players expect. A player winning at 2bb/100 can statistically experience downswings exceeding 50,000 hands before fully recovering. A strong 5bb/100 winner can still face 20,000-hand breakeven stretches as a normal statistical event. Tournament players experience even more extreme variance, with losing stretches of 200+ tournaments being statistically expected at certain ROI levels. Having 100,000+ hands in your sample is the minimum before results become statistically meaningful for determining your true win rate.

How many buy-ins do I need for poker bankroll management?

For 6-max cash games, 20–30 buy-ins is the standard range, with 25 being the typical recommendation. At 20 buy-ins, a player winning at 5bb/100 still faces roughly 20% risk of ruin before recovering — meaning 1 in 5 players with that profile goes broke. At 30 buy-ins, risk of ruin drops to under 5%. Tournament players need 100–200 buy-ins depending on field size and payout structure. PLO and PLO5/PLO6/PLO7 variants require 30–50% more bankroll than equivalent NLHE stakes due to higher structural variance.

How do I know if I'm running bad or playing bad?

The most reliable diagnostic is comparing your current HUD statistics against your historical baseline from a period you were playing well. Increases in VPIP, decreases in PFR-to-VPIP ratio, dropping aggression factor, and unusual 3-bet deviations all suggest tilt-induced play degradation. If your All-In EV line (in supported trackers) is significantly above your actual results over a 20,000+ hand sample, variance is the primary culprit. If actual results and All-In EV track closely but you're losing, the issue is your non-all-in play — betting, calling, and folding decisions — and requires technical review.

Should I move down stakes during a downswing?

Yes — if your bankroll falls below 20 buy-ins for your current stake, moving down is not optional. It is mathematically correct bankroll management. Staying at your current stake below this threshold significantly increases your risk of ruin. Move down to the stake where your remaining bankroll equals 25–30 buy-ins, grind back, and return when you've rebuilt to 30 buy-ins at the original level. Pride is the most expensive thing you own at the poker table during a downswing.

Can provably fair poker really eliminate rigged-deck concerns?

Yes — definitively. Provably fair poker generates a SHA-256 cryptographic hash of every deck before cards are dealt. The hash is a mathematical commitment that cannot be altered without detection. After the hand, you receive the server seed, client seed, and nonce — all the inputs needed to independently reproduce the hash and confirm the deck was honest. This is not a trust-based claim; it is a mathematical proof. PlasmaPoker is currently the only free-to-play poker platform offering this level of cryptographic transparency for every hand.

Poker Variance & DownswingsHow to Survive Bad Runs in 2026