Share-triggered vardiff can leave Bitcoin miners uncredited

Some Bitcoin mining controllers that adjust share difficulty only when a miner submits a share can keep a slowed machine hashing at an outdated, harder target. The miner may continue using power while producing few or no accepted shares, creating a window with little or no pool-side credit.

The issue involves pool-assigned share difficulty, not Bitcoin’s network block difficulty. Pools give each miner connection an easier target than the network block target so they can measure contributed work. Variable-difficulty (vardiff) controllers tune that target to keep shares arriving at a steady rate. When a controller recalculates difficulty only after it observes a submitted share, a sudden drop in a miner’s hashrate can leave the assignment at the previous, harder level. Without a fresh share to trigger an update, the controller can hold the wrong difficulty and accepted shares can become sparse.

An analysis by mining engineer Eric Price published in July described how channels can become trapped under certain decline profiles. The Stratum V2 reference implementation recalculates difficulty on a timer and lowers the assigned target during a share drought. That timer-based behavior is part of the reference code rather than a requirement of the protocol, and implementations that trigger recalculation only on share arrival can recover slowly, especially on long-lived miner-pool channels.

Abrupt curtailment of miners is operationally realistic. Operators sometimes reduce machine power for maintenance, grid stress or weather. A January 2026 winter storm produced a sharp temporary drop in overall network hashrate when miners reduced power. If a curtailed miner continues to hash at reduced power while its assigned difficulty remains high, accepted shares can be very rare until the pool lowers the target.

Payment effects depend on pool accounting rules. Some pools assign extra weight to rare, high-difficulty proofs so a single accepted share can represent more accounted work. Under pay-per-share or proportional accounting, missing accepted shares during a reward window can leave a miner with no payment for that interval or increase the relative share awarded to other participants for the same interval.

Operators can test how upstream pools respond. The MARA Foundation’s open-source shape-proxy can acknowledge shares locally while forwarding a controlled fraction upstream, creating step, ramp or stall profiles that make the pool observe an apparent decline without changing the miner’s physical output. If the pool lowers assigned difficulty in response, the controller has a recovery path; if the target stays pinned, the pool’s controller may be slow or rely on share arrivals to trigger updates. Timer cadence, channel age and random share timing affect each test’s outcome.

At least one deployed pool has been observed operating in a share-triggered mode. How common that behavior is and whether it has caused material losses has not been measured by available analyses. The condition arises when an implementation choice-recalculation on share arrival rather than on a periodic timer-interacts with an abrupt reduction in mining output, producing a finite window in which miners continue to consume power but generate few or no credited shares.