What is Mining Luck and
Why Your Earnings Fluctuate
The mathematics behind pool luck, why your daily earnings swing wildly, and why switching pools during an unlucky streak is exactly the wrong move.
Mining is a random process governed by probability, not skill or timing.
- Luck = Expected Shares / Actual Shares × 100% — 100% is average, lower is luckier
- Each block attempt is independent — past bad luck does not make future good luck more likely
- Larger pools have less variance — same expected earnings, smoother day-to-day income
- Over thousands of blocks, luck converges to 100% — trust the math, not the feeling
- Switching pools during bad luck does not help — it is the Gambler's Fallacy
What is Pool Luck?
Every block on a proof-of-work blockchain is found by submitting hashes until one meets the network difficulty target. The probability of any individual hash being a solution is astronomically small. A mining pool combines the hashing power of many miners to find blocks more frequently, but the when of finding each block is still random.
Pool luck measures how many shares (units of work) the pool needed to find a block compared to how many it was expected to need based on the network difficulty. It is the ratio of theoretical work to actual work:
Luck % = (Expected Shares / Actual Shares Submitted) × 100
// Example 1: Pool expected 1,000,000 shares, actually needed only 500,000
Luck: (1,000,000 / 500,000) × 100 = 200% // Very lucky!
// Example 2: Pool expected 1,000,000 shares, actually needed 2,000,000
Luck: (1,000,000 / 2,000,000) × 100 = 50% // Very unlucky
// Example 3: Pool expected 1,000,000 shares, needed exactly 1,000,000
Luck: (1,000,000 / 1,000,000) × 100 = 100% // Perfectly averageSome pools display luck inversely: 50% = lucky (found in half the expected shares) and 200% = unlucky (took twice as long). Other pools use the convention above where higher is luckier. Always check what convention your pool uses. In this article, we use the convention where higher = luckier (200% = found in half the expected shares).
Think of mining like rolling a million-sided die. You need to roll a "1" to find a block. On average, you expect to roll the die about 1 million times. Sometimes you get lucky and roll a "1" after only 300,000 attempts. Sometimes you are unlucky and it takes 3 million rolls. Over millions of blocks, the average approaches exactly what probability predicts.
The Mathematics of Mining Luck
Mining is governed by two key probability distributions. Understanding them helps explain why luck varies so wildly.
Geometric Distribution: Time to Find One Block
Each hash has a fixed probability p of being a valid block solution. The number of attempts needed to find the first success follows a geometric distribution. The key properties are:
Probability of success per hash: p = 1 / Network Difficulty
Expected hashes to find a block: E[X] = 1 / p = Network Difficulty
// Key insight: the geometric distribution is "memoryless"
// If you've been hashing for 2x the expected time without finding a block,
// the REMAINING expected time is still the same as when you started.
// The network doesn't "owe" you a block.The memoryless property is counterintuitive but critical. It means that after mining for 10 hours without finding a block, the expected time until the next block is exactly the same as it was when you started. Past hashing work provides zero information about when the next block will arrive.
Poisson Distribution: Blocks Over Time
When looking at the number of blocks a pool finds over a fixed time period, the count follows a Poisson distribution. If the pool expects to find an average of λ (lambda) blocks per hour:
P(finding exactly k blocks) = (e^(-λ) × λ^k) / k!
// Example: Pool expects 6 blocks per hour (λ = 6)
P(0 blocks): e^(-6) × 6^0 / 0! = 0.25% // Very unlikely but possible
P(3 blocks): e^(-6) × 6^3 / 3! = 8.9% // Unlucky but common
P(6 blocks): e^(-6) × 6^6 / 6! = 16.1% // Most likely single outcome
P(9 blocks): e^(-6) × 6^9 / 9! = 6.9% // Lucky
P(12 blocks): e^(-6) × 6^12 / 12! = 0.5% // Very lucky
// Standard deviation of Poisson: σ = √λ
σ: √6 = 2.45 blocks
// So finding 4-8 blocks in an hour is "normal" (within 1σ)The standard deviation of a Poisson distribution is the square root of the mean. This is why larger pools (higher λ) have proportionally less variance: if a pool expects 100 blocks/day, the standard deviation is only 10 (10% of the mean), but if a pool expects 4 blocks/day, the standard deviation is 2 (50% of the mean).
Visualizing Luck: 50 Blocks Example
Here is what luck looks like for a real sequence of 50 blocks. Each row shows the luck for one block. Notice how wildly it swings:
| Block # | Luck | Visual | Assessment |
|---|---|---|---|
| 1 | 312% | Found in 1/3 expected time | Very lucky |
| 2 | 89% | Slightly below average | Normal |
| 3 | 445% | Found almost immediately | Extremely lucky |
| 4 | 34% | Took nearly 3x expected time | Very unlucky |
| 5 | 112% | Slightly above average | Normal |
| 6 | 198% | Found in half expected time | Lucky |
| 7 | 52% | Took nearly twice as long | Unlucky |
| 8 | 76% | Below average | Slightly unlucky |
| 9 | 267% | Found very quickly | Very lucky |
| 10 | 41% | Long round | Unlucky |
Even with just 10 blocks shown, the luck ranges from 34% to 445% — a 13x spread. This is completely normal. If someone showed you only block #4 (34% luck) in isolation, you might panic. But zoom out to the full 50 blocks and the average is very close to 100%.
Across the 10 blocks shown above, the average luck is ~163%. This is slightly above average because of the two extremely lucky blocks (445% and 312%). Over 50 blocks, the average is ~108%. Over 500 blocks, it would be within a few percent of 100%.
This is the law of large numbers in action: short-term chaos, long-term convergence.
How Pool Size Affects Variance
All pools have the same expected luck (100% average). But they experience different levels of variance. The key factor is how many blocks the pool finds per day.
| Pool Size | Blocks/Day | Std Dev (daily) | Variance Feel |
|---|---|---|---|
| Tiny pool | 0.5 | ± 0.7 blocks (±141%) | Extreme — days without blocks, then clusters |
| Small pool | 4 | ± 2 blocks (±50%) | High — 2–6 blocks on any given day |
| Medium pool | 25 | ± 5 blocks (±20%) | Moderate — 20–30 blocks most days |
| Large pool | 100 | ± 10 blocks (±10%) | Low — 90–110 blocks consistently |
| Huge pool | 500 | ± 22 blocks (±4.5%) | Very low — almost flat daily earnings |
// The variance formula for blocks per day:
Standard deviation = √(expected blocks per day)
Relative variance = 1 / √(expected blocks per day)
// Example: Pool finding 4 blocks/day
σ = √4 = 2 blocks
// On 68% of days: 2-6 blocks (within 1σ)
// On 95% of days: 0-8 blocks (within 2σ)
// That means ~5% of days: 0-1 blocks OR 8+ blocks
// Example: Pool finding 100 blocks/day
σ = √100 = 10 blocks
// On 68% of days: 90-110 blocks (within 1σ)
// On 95% of days: 80-120 blocks (within 2σ)
// Much smoother! Relative variance is only 10%Imagine flipping a coin 10 times versus 10,000 times. With 10 flips, getting 7 heads (70%) is not unusual at all. With 10,000 flips, getting 7,000 heads (70%) would be essentially impossible. The same principle applies to mining: more blocks means the luck percentage stays closer to 100%.
Short-Term vs Long-Term Luck
The crucial insight about mining luck is the difference between what you observe in a day versus what you observe over months:
This convergence happens because each new block adds data to the average. After 10 blocks, one extremely lucky block heavily skews the average. After 1,000 blocks, that same lucky block is just a rounding error.
Real-World Example: 100 Blocks
Consider a medium-sized pool that found 100 blocks over two weeks. Here is a realistic distribution of what those luck values might look like:
| Luck Range | Expected Count | Description |
|---|---|---|
| 500%+ (extremely lucky) | 5–8 | Found in < 20% of expected shares |
| 200–500% | 12–18 | Found well below expected shares |
| 100–200% | 20–28 | Better than average |
| 50–100% | 22–30 | Slightly below to average |
| 25–50% | 12–18 | Took 2–4x expected shares |
| < 25% (extremely unlucky) | 5–10 | Took 4x+ expected shares |
Notice the distribution is skewed right — you can get infinitely lucky (finding a block on the first hash) but there is no lower bound on how unlucky you can be (a round could theoretically take forever). This asymmetry means that the median round is actually faster than the mean round. Most blocks are found faster than average, but occasional very long rounds pull the average up.
Common Misconceptions
"The pool has been unlucky for 3 hours, so it must be due for a lucky block soon."
This is false. Each block attempt is completely independent of all previous attempts. The probability of finding a block on the next hash is exactly the same whether the pool has been mining for 5 minutes or 5 hours since the last block. Mining has no memory. The network does not track how long you have been searching or feel sorry for unlucky pools.
The Gambler's Fallacy is the same error that makes people believe a roulette wheel is "due" for red after a streak of blacks. In both cases, each event is independent.
"My pool has been unlucky for two days. I should switch to a luckier pool."
Switching pools during an unlucky streak does not improve your odds. The new pool has the exact same probability of finding blocks as the old one. Worse, if you are on a PPLNS pool, switching resets your share accumulation in the payout window, meaning you earn less on the new pool until your shares build back up.
There is one legitimate reason to switch pools: if you believe the pool has less hashrate than reported (meaning the "bad luck" is actually a sign of the pool shrinking). But if the pool hashrate is genuine, the luck will converge.
"Small pools have bad luck. Big pools are always luckier."
Both have the same expected luck of 100%. The difference is variance. A small pool swings between 20% and 400% luck regularly, while a large pool stays between 80% and 120%. Over months, both average out to ~100%. Your total earnings are the same in expectation — the small pool just delivers them in lumpier chunks.
There is a practical concern though: if a small pool goes through a very unlucky streak and you need regular payouts for electricity costs, the variance can be a real problem. Choosing the right coin to mine with sufficient pool activity helps even if the long-term math is identical.
"That pool operator must be doing something right — they always have good luck."
Luck is pure randomness. No pool operator can influence it. If a pool appears consistently "lucky," it is either statistical coincidence (you are looking at a short time window), or the pool is reporting inaccurate statistics. A pool's actual value comes from low fees, reliable infrastructure, fast share processing, and good connectivity — not luck.
How PPLNS Handles Luck vs PPS
The payout system your pool uses determines how luck affects your earnings:
| Scenario | PPS Earnings | PPLNS Earnings |
|---|---|---|
| Lucky day (150% luck) | Normal (unchanged) | ~50% above normal |
| Average day (100% luck) | Normal | Normal |
| Unlucky day (60% luck) | Normal (unchanged) | ~40% below normal |
| Over 1 year | Stable but lower (higher fees) | Variable but slightly higher (lower fees) |
PPS (Pay Per Share) eliminates luck entirely from the miner's perspective. The pool absorbs all variance — paying you the same rate per share whether the pool finds 10 blocks or 0 blocks that hour. In exchange, PPS pools charge higher fees (typically 2–5%) to build a reserve fund that covers unlucky periods.
PPLNS (Pay Per Last N Shares) passes luck through to the miners. When the pool is lucky, you earn more. When the pool is unlucky, you earn less. But because PPLNS pools do not absorb risk, they charge lower fees (typically 0.5–2%), resulting in slightly higher long-term average earnings.
PPS is an insurance policy against bad luck. Like all insurance, you pay a premium (higher fees) for the protection. PPLNS is going uninsured — you accept the variance in exchange for keeping more of your earnings on average. Neither is objectively better; it depends on your risk tolerance and whether you need predictable daily income.
Variance in Your Personal Earnings
Even on a perfectly average-luck pool, your personal daily earnings will fluctuate. Understanding how mining payouts work helps explain why. Your individual share submissions also follow a random distribution. Your earnings depend on:
- Pool luck: How many blocks the pool finds (Poisson variance)
- Your share timing: Whether your shares fell inside or outside the PPLNS window when blocks were found
- Hashrate fluctuations: GPU temperature throttling, restarts, driver updates
- Share difficulty variance: The time between your individual shares also varies randomly
All these sources of randomness compound. The result is that daily earnings can easily swing 30–50% above or below your "expected" rate, even when the pool's luck is perfectly normal.
// Your personal earnings variance (simplified)
Expected daily earnings: 10 coins
Standard deviation: ~3 coins (depends on pool size and payout system)
// What you might actually see over a week:
Monday: 12.4 coins (+24%)
Tuesday: 7.1 coins (-29%)
Wednesday: 11.8 coins (+18%)
Thursday: 6.3 coins (-37%) // unlucky day
Friday: 14.2 coins (+42%) // lucky day
Saturday: 9.5 coins (-5%)
Sunday: 8.7 coins (-13%)
Weekly avg: 10.0 coins // right on targetThis is why judging a pool's quality by one day — or even one week — of earnings is statistically meaningless. You need at least several hundred blocks of data before the luck signal emerges from the noise.
Practical Advice for Miners
Never evaluate your mining earnings over less than a week. Ideally, look at monthly averages. A single bad day — or even a bad week — is statistically insignificant. The pool dashboard might show alarming red numbers today, but check the 30-day average and it will almost certainly be close to expected.
Switching pools to escape bad luck is counterproductive. You lose your PPLNS share history, pay transaction fees moving coins, and the new pool has exactly the same odds. The only reasons to switch pools are: lower fees, better uptime, lower latency, or a genuine decline in the pool's hashrate. Use a profitability calculator to compare actual earnings across pools over time.
If you check your earnings hourly and stress about bad days, use a larger pool — the variance will be smaller. If you are comfortable with lumpy payouts and do not need daily predictability, a smaller pool gives you the same expected earnings with the bonus of supporting decentralization.
If your mining operation must cover fixed costs (electricity, hosting) and you cannot afford extended unlucky streaks, PPS removes luck from the equation entirely. The higher fee is worth it if variance creates actual financial risk for you.
Bottom Line
Luck is pure randomness. There is no pattern, no cycle, no "system." Each block attempt is independent of every previous attempt. Accepting this is the single most important thing a miner can internalize.
Short-term variance is enormous. Individual blocks range from 10% to 1000%+ luck. Daily averages swing 50–150%. Weekly averages swing 70–140%. Only monthly and longer averages approach the true 100%.
Larger pools are not luckier — they are less volatile. Same expected earnings, smaller swings. Choose pool size based on your patience and payout needs, not on perceived "luck."
The Gambler's Fallacy will cost you money. Switching pools during an unlucky streak loses your PPLNS history and gains you nothing. The math does not care about your feelings. Trust it, and your earnings will average out exactly as probability predicts.