Stake RTP Explained: What Every Player Needs to Know

Published 2026-04-20 · Tags: guides, rtp, fundamentals

RTP is the most important casino number — and the most misunderstood. How Stake's 99% Originals compare to the industry and why your session RTP is a lie.

There's a question I get asked in chat often enough that I should just put the answer here: "is the RTP Stake publishes actually real?"

Short version from our tracker: close enough. We've logged 660,751 sportsbook bets — singles return 96.4% of staked dollars ($66.1M paid on $68.5M staked), which matches a ~3-4% book vig almost exactly. The published math is the real math.

Where it gets interesting is when RTP doesn't match your session. That's what this post is really about.

The definition

RTP is the percentage of wagered money a game returns to players over a theoretically infinite number of bets. A slot with 96% RTP is designed so that, over millions of spins, players collectively get back $96 for every $100 they wager.

The casino keeps the other 4%. That 4% is the house edge (sometimes written HE). RTP and house edge are just two sides of the same number: RTP + HE = 100%.

The "long term" qualifier that breaks every intuition

Here's what people get wrong: your personal session RTP can be anywhere from 0% to 10,000%. Any of those outcomes is consistent with a 96% RTP game. The 96% only becomes true at very large sample sizes (hundreds of thousands of spins).

In any given 100-spin session, you might see:

• 0% RTP (lose every spin, possible on volatile slots)
• 300%+ RTP (hit a big bonus early)
• Anywhere in between

The "96%" is what the game converges to after enough samples that variance smooths out. For high-volatility slots, that can take tens of thousands of spins per player.

This is why "why is my RTP on this game only 60%?" after 500 spins is a meaningless question. You haven't sampled enough. Our Bet Analyzer will give you your actual RTP across your bet history — once you have 5,000+ bets on a single game it starts being informative, below that it's just variance.

Stake's published RTPs

Stake is one of the more transparent operators. They publish the RTP of every game in the settings popup (gear icon → "Provably Fair" or "Fairness" tab, depending on the game).

• Dice (default config): 99.00%, one of the loosest casino games anywhere
• Limbo: 99.00%
• Crash: 99.00%
• Mines: 99.00% (varies slightly by bomb count)
• Plinko: 98.00%–99.00% depending on rows/risk
• Keno: 92.00%–94.00% depending on risk level
• Wheel: 96.00%–99.00% depending on segments/risk
• Slots (third-party): 88%–99% depending on the game, see the individual game info
• Live dealer games: 97%–99% depending on the game

The Originals (Stake's in-house games) sit at 99% RTP for most configurations. This is substantially better than traditional casinos, which are typically 85–95% on slots and 94–98% on table games.

Stake is actually very loose by industry standards

"Loose" in casino jargon means "higher RTP / lower house edge." Stake's Originals at 99% are among the loosest games available anywhere online.

For comparison:

• Las Vegas Strip slots: ~92% average
• Downtown Vegas: ~93% average
• UK online slots: 95%–96% typical, capped by regulation at 97%
• Stake Originals: 99%
• Bitcoin-only casinos with cheap overhead: 97–99%

Stake benefits from lower operational overhead (no land-based cost, no fiat-rail fees) and passes some of that back as higher RTP. That's a real edge vs physical or fiat-primarily casinos.

Where Stake is not the loosest

Third-party slots. Games from Pragmatic Play, Hacksaw, Nolimit City etc. run at the RTP the provider sets. Many have a "low RTP" mode (92%–95%) that operators can toggle. Stake historically uses the high-RTP mode for most popular slots — but not all. You should always check the specific game's RTP in its info panel.

Our Slot Cruncher computes the true RTP of every Stake Engine slot straight from its real lookup table. You can see at a glance which are the loosest and tightest, sort and pick accordingly.

Why 99% Dice still loses most players money

This is the math that trips everyone.

99% RTP means the house keeps 1 cent per dollar wagered on average. But you're not wagering once, you're wagering thousands of times per session on Dice. If you wager $1,000 in total volume (not deposits — total bets summed), you expect to lose $10. If you wager $10,000 in volume, you expect to lose $100. If you wager $1,000,000 in volume (very doable over a year of Dice grinding), you expect to lose $10,000.

Volume kills you, not the per-bet loss. Every click is a fresh sample from the same distribution.

Martingale, chasing, and "just one more spin" all increase your volume, which increases the expected loss. You cannot grind your way out of a negative-EV bet regardless of sizing strategy. We have a full post on why chasing never works that walks through this with simulated data.

Volatility ≠ RTP

Two games can have identical RTP and completely different session experiences. Volatility measures how spread out the outcomes are.

• Low volatility (e.g. Dice at 2× target, Limbo at 1.5×): frequent small wins, small losses. Session variance is tight.
• High volatility (e.g. Keno 10-pick, Plinko high-risk, slots with big-hit features): long dry streaks punctuated by occasional big wins. Session variance is wide.

Same RTP, very different feel. Some people prefer the slow-burn of low-volatility (more session-time per dollar); others chase the dopamine of rare huge hits. Pick intentionally. Our Bet Analyzer computes the stddev of your per-bet P/L so you can see which games are actually high-vol for you.

The practical player playbook

If you're trying to minimize your loss rate:

• Play Originals at 99%. Dice, Limbo, Crash, Mines with a low bomb count.
• Stay low-volatility so you can actually sample the RTP. High-vol games "require" 10,000+ bets to converge.
• Keep your volume down. Sessions with 100 bets have more honest outcomes than sessions with 3,000 bets.
• Avoid low-RTP slots. The 92% RTP slots are dopamine traps; they feel exciting but the house edge is 8× worse than Originals.

If you're here for entertainment:

• Volatility is what creates the dopamine. Play the game that matches your vibe.
• Budget your session as pure entertainment spend. It is not an investment.

Verify the RTP yourself

You don't have to trust Stake's posted RTP. You can:

1. Run many bets at a fixed stake and target
2. Sum your wagered amount
3. Sum your returned amount (stake × payout multiplier on wins, 0 on losses)
4. Compute returned / wagered, that's your realized RTP

Do this over 10,000+ bets and compare to the posted number. It should match within 1–2% (the rest is sampling noise).

Even faster: just export your bet archive from Stake and drop it into our Bet Analyzer. We do this exact calculation per-game over your whole history.

Big picture

RTP is the single most informative number on any Stake game. 99% Originals are among the loosest games online. But RTP only reveals itself over large samples, and the volume you wager is what compounds the house edge into real money loss. Know your RTP, keep volume honest, and don't confuse short-term swings with long-term math.

Want to know your actual realized RTP per game? Open the Bet Analyzer →