Understanding House Edge: Where Casinos Win and Players Can Optimize
By Casey Morgan, gambling math writer • Last updated: 17 Aug 2026
It is a Friday night. Two friends walk into the same casino with the same $200. One sits at a single‑zero roulette wheel. He bets small, takes breaks, and chats with the dealer. The other drops into a video poker bank. He finds a 9/6 Jacks or Better screen and brings a simple chart. After one hour, both feel they had fun. But the first friend is down a steady amount. The second is almost even.
That is not luck alone. That is the quiet pull of house edge. Once you see it, you can make better choices. You will not beat the house in the long run. But you can lose less, play longer, and feel in control.
- The short version
- What drains a bankroll
- House edge in plain words
- Quick table by game and bet
- Why casinos win: the quiet multipliers
- The optimization playbook
- A realistic hour: loss and comps
- Common myths, retired
- Where to learn more (and a useful link)
- Quick checklist
- FAQs
- About, notes, and help
The short version (for the quick reader)
- House edge is the built‑in hold a casino has on each bet over time. It is small on some games, large on others.
- Casinos win with a simple mix: edge × speed of play × hours × bet size. That is the whole model.
- You can optimize. Pick lower‑edge bets and better rule sets. Learn basic strategy where it helps. Slow down. Set a firm budget and time cap. Treat comps as a small rebate, not profit.
Field notes from the floor: what actually drains a bankroll
High edges do hurt. But that is not the full story. Fast games chew through your cash even with a fair edge. Slots can run 300–600 spins per hour. Roulette runs about 40–50 spins per hour. Blackjack deals 60–70 hands per hour at a full table, and even faster heads‑up.
Side bets look fun. Many are the worst bets in the room. Rule tweaks matter. Blackjack that pays 6:5 on a blackjack turns a good table into a bad one. Fatigue also costs money. We make weak choices when we are tired. We press bets after losses. We chase when we should stop.
Comps feel like free money. They are not free. They are based on a “theo” loss number. If you overvalue a buffet, you may tip the scale toward bad play. A nice meal is not worth hours on a high‑edge game.
House edge in plain English
House edge is the long‑run cost of a game. If a bet has a 2% edge, you can expect to lose $2 for each $100 bet in the long run. That is an average over many trials. In the short run, anything can happen. You can win today and still face the same edge tomorrow. For a clear definition with examples, see this overview of house advantage.
RTP (return to player) is the flip side of edge. If edge is 2%, RTP is about 98%. Online rules in many markets ask for fair games and clear rules. See how regulators require fair games and clear rules in the UK.
Edge is not the same as volatility (also called variance). Volatility is how wild the results swing around that long‑run line. Slots often have higher swings. Baccarat is more steady. Both edge and volatility shape how your session feels.
Numbers you can use: quick table by game and bet
Game rules change by place. The numbers below are typical. For context on real‑world holds by game, see historical hold by game in Nevada. Read the notes column. Small rule lines can move the edge a lot.
| Roulette (single‑zero / European) | ~2.70% | Medium | No | Prefer single‑zero. Avoid special bets with extra edge. |
| Roulette (double‑zero / American) | ~5.26% | Medium | No | Common in the US. Costs almost double vs single‑zero. |
| Blackjack (good 3:2 rules) | ~0.3–0.6% with perfect play; ~1–2% for most players | Medium | Yes (basic strategy) | 6:5 payout can raise edge to 1.5–2%+. |
| Baccarat (Banker) | ~1.06% | Low | No | Player bet ~1.24%. Tie is often 14%+ edge. Avoid Tie. |
| Craps (Pass / Don’t Pass) | 1.41% / 1.36% | Low–Medium | Low | Taking true odds has 0% edge. Mix lowers total edge. |
| Craps (one‑roll props) | 5–16%+ | High | No | Fun but costly. Keep them small or skip them. |
| Video Poker (9/6 Jacks or Better) | ~0.46% with perfect play | Medium | Yes | 8/5 paytable jumps to ~2.7% edge. Paytables matter a lot. |
| Slots (regulated online) | ~3–6% (varies) | Med–High | No | Many sites list RTP. Studio sheets help. Check before you spin. |
| Slots (land‑based) | ~4–12% (varies) | Med–High | No | Edge can change by state and denom. |
| Keno | 20–35% | High | No | Very high edge. Play for fun, not value. |
| Sic Bo (common bets) | ~2.78–33% | High | No | Some bets fair, many not. Know which is which. |
Why casinos win: the quiet multipliers
Casinos do not need big edges to win. They use scale. Here is the model: edge × speed × bet × time. They also nudge pace with lights, sounds, and easy “repeat bet” buttons. The industry’s data and standards show how steady this model is.
Comps help keep you in the seat. They are based on your “theo,” which is an estimate of your long‑run loss. If your theo is $100, a 25% comp rate gives you $25 in value. The rest still goes to the house. It feels good, but the math still rules.
Our heads trick us. We see patterns in noise. We think a number is “due” on the wheel. We feel a “hot” hand on dice. This is human, but wrong. See why the gambler’s fallacy sticks. Random streaks do not have memory.
The optimization playbook (real steps you can use)
- Pick better games. Single‑zero roulette over double‑zero. Banker in baccarat. Pass/Don’t with odds in craps. Skip Tie and high‑edge props.
- Read rules and paytables. For blackjack, find 3:2 payout, dealer stands on soft 17 if you can, and allow double after split. For video poker, hunt for 9/6 Jacks or Better or other strong tables.
- Use basic strategy where it matters. Bring a small card or app if allowed. Small errors add up fast. For a light math refresher, try short, free lessons on probability and expected value.
- Online fairness: in many markets, games are checked by labs. See how independent lab testing for RNGs and RTP works. For dispute help and seals, look at eCOGRA.
- Slow the pace. Take breaks. Use a smaller denomination. Fewer decisions per hour lowers expected loss.
- Set a hard budget and a time limit. Stop when you hit either. Wins are a reason to book out, too.
- Value comps as a rebate only. If your theo is $20 per hour and you get back 30%, that is $6 in value. Nice, but not a win.
- Optional, advanced: bet sizing ideas like Kelly need an edge in your favor. In casino games with a house edge, do not use Kelly. It will speed up losses.
A realistic hour: expected loss and comps, side by side
Here is simple math you can check. Numbers are rounded.
- Double‑zero roulette: $5 average bet × 50 spins × 5.26% edge ≈ $13.15 expected loss per hour. If your comp rate is 20–30% of theo, value ≈ $2.60–$3.95.
- Blackjack (typical player, 1% edge against you): $10 average bet × 60 hands × 1% ≈ $6 per hour. Comps at 30% ≈ $1.80 value.
- 9/6 Jacks or Better (perfect play): $1.25 per hand (quarters × 5) × 600 hands × 0.46% ≈ $3.45 per hour. Miss holds? Your edge rises. Slow down to cut volume.
Note: This is not a bill you must pay each hour. It is the long‑run average. In short runs, your line bounces up and down. But with more hours, the math pulls you to that line. For a deeper look at rating and theo, see how casinos rate play and compute “theo”.
Common myths, politely retired
- “Systems beat the wheel.” No system can fix a built‑in edge. It only changes the ride, not the line.
- “Past results predict the next spin.” No. A wheel and a fair RNG do not have memory.
- “Higher denom always pays more.” Sometimes, but not a rule. Check posted RTP or real paytables where listed.
- “Online RNGs are rigged.” In regulated markets, RNGs are tested. Labs use randomness testing frameworks. Pick licensed sites that use approved labs.
Where to learn more and how to vet casinos (plus one useful link)
Before you sit down, check rules and RTP for the place you plan to play. Does blackjack pay 3:2? Is roulette single‑zero? Is the video poker table 9/6? Look for posted sheets, floor signs, or the help menu online. If you care about safe play and fair games, stick to licensed sites and floors that use approved labs and have clear dispute paths.
If you also mix in sports, this clear guide shows how live betting works for Kenyan sports punters. It breaks down live odds, pace, and risk in simple terms. The same math ideas apply: small edges, fast play, and strict limits.
Want to read more from neutral sources? Try the links in this article. They point to regulators, testing labs, and industry bodies, not sales pages.
Quick checklist to print or save
- Pick games with edges under ~1.5% when you can.
- Blackjack: 3:2 or skip. Video poker: strong paytables only.
- Craps: Pass/Don’t with odds. Avoid high‑edge props.
- Roulette: choose single‑zero if offered.
- Check RTP/paytables, especially online.
- Set time and money limits. Stop at either.
- Play slower. Take breaks. Drink water.
- Treat comps as a small rebate, not profit.
FAQs
What is house edge and how is it calculated?
House edge is the long‑run share of each bet the casino keeps. It comes from the rules and odds of the game. If a bet has a 2% edge, the long‑run loss is $2 per $100 bet on average.
Is RTP the same as house edge?
They are two sides of one coin. RTP is the share paid back to players over time. Edge = 100% − RTP. If RTP is 98%, edge is about 2%.
Which casino game has the lowest house edge?
With good rules and correct play, blackjack and strong video poker tables can be under 1%. Baccarat Banker is near 1%. Craps with odds also cuts your blended edge.
Can basic strategy in blackjack really help?
Yes. It can drop the edge from 1–2% for guess play to about 0.3–0.6% at good tables. Use a chart. A few small fixes make a big change over time.
Are online casino games fair and truly random?
In licensed markets, yes. Games are tested by labs and approved by regulators. Look for lab seals and licenses. We linked to GLI and eCOGRA above.
Do comps and bonuses beat the house?
No. They give back a slice of your long‑run loss. They can improve value but do not flip the edge.
About the author, sources, and safety
Author: Casey Morgan writes about casino math and risk. Casey studies how rules change edges and how pace affects loss. This guide cites regulators, labs, and education sites for clarity.
Sources in this guide:
- Definition and examples of house advantage
- UKGC remote technical standards (RTP and fairness)
- Nevada Gaming Control Board (holds by game)
- American Gaming Association (industry and RG)
- Britannica: Gambler’s fallacy
- Gaming Laboratories International (RNG/RTP testing)
- eCOGRA (dispute help and player safety)
- Khan Academy: probability and EV
- UNLV Center for Gaming Research (theo and rating)
- NIST: randomness test suite
Region note: This guide is for regulated markets. Rules and RTP vary by place. Always check local law.
Risk note: Gambling has risk. Never play with money you need for bills or goals. Set limits. Take breaks. If play stops being fun, step away.
Get help if you need it: National Council on Problem Gambling (US) • BeGambleAware (UK) • Gamblers Anonymous (global)
Disclosure: This page may include links to partners or helpful resources. We aim to keep advice neutral and based on math and public data.
Appendix: How to read the table (a quick guide)
“House edge” is the long‑run share the casino holds. “Volatility” is how swingy the game feels. “Skill needed” tells you if learning a chart can help. “Notes” warn you where rules can change the edge. If two tables look the same but pay different on key hands (like 3:2 vs 6:5 in blackjack), pick the better one. Small lines on a sign can save you real money.