methodology · how we simulate

slots·science is a slot behavior analytics lab. we publish data on online slot games. our data is generated by monte carlo simulation, not by live testing of real-money games. every metric on this site is preceded or followed by methodology. every number is sample-based. we never claim a player has an edge.

how we simulate

each slot in our database has a set of published parameters: return to player (rtp), volatility tier, hit frequency, bonus trigger frequency, max win multiplier, paylines, bet range. these parameters come from the game provider’s certified math sheet or, when unavailable, from independent third-party audits.

we feed those parameters into our monte carlo simulator. the simulator runs 100,000 to 1,000,000 simulated spins per slot, per bet size, per bankroll level. it samples symbol distributions, calculates payouts, triggers bonus rounds at the published frequency, and records every spin’s outcome.

we then aggregate the simulated session data to produce the metrics we publish: dead spin percentage, bankroll survival curves, bonus payout distribution, average session length before bust, payout clustering, and similar statistical summaries.

what we are not doing

we do not run real-money games and report the results. we do not have a back door into provider rng systems. we do not predict the next bonus on any slot. we do not have a strategy that beats the house edge.

slot games have a mathematical house edge built into the rtp. the house edge cannot be overcome by play style, bet sizing, or timing. anyone who claims otherwise is selling something.

what we do is simulate the math the slot already publishes, at sample sizes large enough to reveal session-level behavior the published rtp number obscures. an rtp of 96 percent across one million spins says nothing about whether a $100 bankroll survives 200 spins. our data tries to answer that second question.

sample sizes

our standard simulation runs at 100,000 spins per parameter set. deep analyses on featured slots run at 1,000,000 spins. all published metrics state the underlying sample size in the lab report card.

we report the variance, not just the mean. a slot can pay 96 percent rtp across a million spins and still bankrupt 80 percent of $100 bankrolls in fewer than 100 spins. both numbers are true. we publish both.

data sources

slot parameters come from:

• provider math sheets published with each slot release
• gambling lab certification reports (eCOGRA, GLI, iTech Labs)
• slotcatalog.com for cross-reference
• askgamblers.com for player-reported bonus and bust data

we do not publish a slot until we have at least two independent parameter sources that agree.

what changes our recommendation

if a slot provider issues a math update, we re-simulate. if a player community surfaces a meaningful gap between our simulation and lived behavior, we re-simulate. if a slot is removed from a regulated market, we mark it stale. our methodology and our data are open to challenge.

contact methodology@slotsscience.com if you find an error.