My Favorite Craps Systems: Strategy Framework by Sam
Sam’s framework · Bet selection · Exposure · Collect and regress
My preferred craps “system” is not a formula for beating the house. It is a framework for keeping the layout understandable: use bets whose cost I know, decide the maximum exposure in advance, collect before escalating, and never treat a recent streak as evidence about the next roll.
What changed after years at the tables
I began with Pass and Come bets, experimented with Place bets, Don’t Come entries, regressions, and progressions, and tracked how the plans behaved when the dice did not cooperate. The scar-tissue lesson was simple: a clever layout can organize risk, but it cannot predict the hand or repair an oversized bankroll decision.
When I first learned the game, my regular approach was a Pass Line bet followed by two Come bets. That is still a defensible, comparatively low-edge way to get several numbers working, especially when the Odds are kept affordable. But after rough sessions I began switching between Pass, Don’t Come, and Place bets according to whether the table felt hot or cold. I bought strategy books, watched other players, and tried to adapt faster than the dice.
Years of playing in more than 65 brick-and-mortar casinos changed how I interpret that experience. Switching sides may change the shape of a session, but it does not make a past streak predictive. Right-side and wrong-side bets can also lose during the same shooter’s hand. The useful adaptation was not “reading” the table; it was learning to reduce exposure, record the full result, and stop treating one memorable recovery as proof of a system.
The rules I now put ahead of any named system
- Know every wager. I want the payout, house edge, working status, and worst immediate loss before placing chips.
- Cap the full layout. A $5 label can hide $50 or $100 of simultaneous exposure.
- Collect real money. A payout belongs to the player; it is not “house money” that becomes harmless to press.
- Do not chase temperature. Hot and cold describe past results, not the next independent roll.
- Make stopping mechanical. Time and loss limits are written before play, not negotiated after a losing streak.
Two styles I find easier to audit
The two $22 Inside versions I used
Both versions begin with $5 each on 5 and 9 and $6 each on 6 and 8. After the first inside number hits, collect its payout and remove that winning bet. Do the same after a second working inside number hits. The result is a smaller exposed layout and two collected payouts—but not a guaranteed profit, because 7 can arrive before either collection.
| Version | What happens after two removals | What I now watch |
|---|---|---|
| Regression and rebuild | Use later wins from the remaining inside bets to restore the removed numbers one at a time | The rebuild puts collected value back at risk; record rack and layout separately |
| Regression and Don’t Come | Use later payouts to fund one Don’t Come entry at a time | This is not a free hedge: the Place and Don’t Come bets can overlap, both carry cost, and the DC may travel behind a number |
I prefer the rebuild version when the purpose is simply to make the chip flow easy to follow. The Don’t Come version adds more states and more ways to misread the exposure. The complete worked regression-and-rebuild cash flow is in the $22 Inside guide.
Reno Reach as a collect-and-reset idea
I came up with the Reno Reach after a fast losing day at the tables in Reno. The original version began with $6 Place bets on both 6 and 8. After either number hit, I collected its $7 payout and removed both bets. On the next point cycle, the prior winning number increased by one proper $6 Place unit while the other returned at $6.
That reset rule did something useful: it forced me to collect and wait instead of covering more numbers impulsively. It did not “stem the tide” mathematically. The total starting exposure rises from $12 to $18, then $24 and higher if the same side keeps winning, while an early 7 loses the current layout. A successful cycle’s payout is already the player’s money, so the progression needs a maximum unit and a restart rule.
Beliefs I no longer use as betting signals
| Old table idea | What the record taught me | Current rule |
|---|---|---|
| Wait for the table to turn hot | A streak is visible only after it has happened | Choose exposure without predicting the next roll |
| Switch to the Don’t side on a cold table | The next roll’s combinations do not remember prior short hands | Use Don’t bets only because their mechanics fit the plan |
| Bet both sides to survive | Opposing-looking bets do not automatically cancel; several can lose in one sequence | Map the net outcome before calling anything a hedge |
| Increase after wins because only profit is exposed | Paid chips belong to the player and can be banked | Count all working chips as current exposure |
Systems I treat more cautiously
| Method | Attraction | Main problem |
|---|---|---|
| No 4/No 10 after recent rolls | 7 has twice as many combinations as 4 or 10 | The lay bet already prices that fact; recent rolls add no signal, and commission matters |
| Mixed Place and Field layouts | Many totals appear to be covered | Coverage is not an advantage; one 7 can defeat several positions at once |
| Lay bets plus across bets | Some outcomes pay one side of the hedge | Both sides carry cost and the full liability can be much larger than the system’s nickname |
| Loss progressions | A later win may recover earlier losses | Required stakes grow faster than most session limits |
How I choose a session plan
If I want the lowest expected cost, I look first at Pass or Don’t Pass and an affordable amount of true Odds. If I want more box-number action, I accept that Place bets add decisions and exposure. If I cannot explain the full layout on paper—including a first-roll 7—I do not call it ready.
The craps strategy guide compares those choices, while the systems hub separates repeatable rules from winning promises. For session boundaries, use the win/loss limits guide.
I use hit and run Martingale on several table games. My craps system, that I have been using daily for over a year, is my most successful. I will tell you how I bet and would like your opinion. My buy in is $1746 and I quit for the day when I am up $100. I place the 6 and 8, always working on the come out. I use a 6 level press up. If win at any level my next bet is level 1. If lose I bet the level up. I play 5 different popups daily and win $500 every day.
1 – $3
2 – 9
3 – 25
4 – 75
5 – 200
6 – 560 –
Here is why level 6 is $560
I play on the popup machines in Vegas. The limit is $240 to place, 200 to buy, and 200 on the big 6-8. I must keep the payout below $1200 or the casino will W-2 me. $560 bet pays out $630. On average I win the $100 in less that 15 min.
Editorial clarification — September 7, 2026: This comment describes one player’s reported sessions; it is not evidence of a guaranteed daily return. A Martingale or other loss progression changes bet size and the timing of wins and losses, but it does not remove the house edge. Many small winning sessions can be offset by a less frequent loss near the top of the progression, so bankroll limits and machine limits matter. The $1,200 statement is also not a universal craps tax rule. The IRS explains that Form W-2G reporting depends on the type of gambling, the amount won and, generally, the ratio of winnings to the wager. Gambling winnings can still be taxable even when no W-2G is issued. Review the current IRS W-2G instructions and Topic 419, or consult a qualified tax professional for personal advice. The original comment is preserved as a reader’s experience, not an editorial endorsement.