The Reverse Martingale (Paroli) is a reverse martingale strategy that increases bet size only after wins to harvest short winning streaks while limiting damage from losing runs. It does not change the game's edge; it reshapes variance. To use it safely, predefine a small session bankroll, cap progressions, lock profits early, and stop immediately when limits are hit.
Core principles for running a Paroli session

- Keep the paroli betting system bounded: fixed maximum steps and a hard bet ceiling.
- Separate bankroll into session allocations; never "reload" mid-session.
- Only press after a win; after any loss, reset to the base unit.
- Define a profit lock rule (take profit after N consecutive wins) and a stop-loss for the session.
- Use unit sizing that survives variance (one of the best bankroll management betting strategies is to keep the unit small enough to withstand multiple resets).
- Treat it as a tool for betting systems for winning streaks, not a way to force wins.
Why Reverse Martingale targets streaks and how that changes edge
Paroli targets the distribution of outcomes: you accept many small resets (loss then back to base) in exchange for occasional multi-step streaks that create a larger win. The house edge (or your sports bet's true EV) stays the same; Paroli mainly changes when money is won/lost.
- Best fit: you can execute strict stop rules, your game has stable rules/payouts, and you can tolerate frequent small losses.
- Do not use when: you chase losses, you lack hard table/market limits clarity, you are playing very high-variance paytables (many slots), or your bet has changing odds/terms that break the progression (some live sports markets).
Numeric intuition: if your base unit is 100 THB and you cap at 3 wins (100 → 200 → 400), your "big" outcomes depend on streaks of length 3; most sequences will reset earlier, so discipline matters more than prediction.
Setting a risk-aware bankroll baseline and session allocation
You need a bankroll plan that assumes you can hit multiple resets without tilting. In Thailand context (THB units, local sportsbook limits, casino table minima), the constraint is often the minimum bet and the maximum allowed stake.
- Total bankroll (B): money you can lose without affecting essentials.
- Session bankroll (S): a strict slice of B you will not exceed in one session.
- Base unit (u): your starting bet; all progression sizes derive from u.
- Progression cap: maximum consecutive wins you will press (e.g., 2 or 3 steps).
- Tools: notes app or spreadsheet; optionally a reverse martingale calculator (any simple sheet) to precompute bet sizes and worst-case exposure for your chosen cap and ceiling.
Risk-aware setup rule: decide the cap first (e.g., 3 wins), then choose u so that your maximum bet and session stop-loss are both comfortable. Example: if your table max is 5,000 THB and you use 3 steps (u, 2u, 4u), then u must be ≤ 1,250 THB to keep 4u within limits.
Bet-sizing formulas: fixed units, proportional increases, and ceilings
- Risks and hard constraints to accept upfront:
- Paroli does not "beat" negative EV games; it only changes variance.
- Without a cap, a long streak can push you into table/market limits or comfort limits.
- One emotional override (adding steps, increasing u mid-session) defeats the method.
- High payout asymmetry (some slots/props) makes "pressing after wins" less predictable than it looks.
-
Pick a progression model (doubling or proportional)
Common choices are doubling after each win or a milder proportional press. Use one model consistently for the session.
- Doubling: bet after k wins is bk = u × 2k (k = 0,1,2...)
- Proportional (milder): bk = u × rk, where r might be 1.5-1.8 (choose based on comfort and limits)
Example: u = 100 THB, doubling for 3 steps gives 100 → 200 → 400.
-
Set a hard cap on consecutive wins (N) and a bet ceiling (C)
Define N (max presses) and C (maximum stake allowed by you and the venue). If the next bet would exceed C, you must stop pressing or end the cycle.
Rule: only press while (k < N) and (bk ≤ C).
Example: u = 200 THB, N = 3, C = 1,000 THB. Sequence is 200 → 400 → 800, then stop (next would be 1,600 > C).
-
Define the reset condition (always) and profit-lock condition (usually)
Reset is mandatory: after any loss, return to u. Profit-lock is your win trigger: after N consecutive wins (or hitting a profit target), stop the cycle and bank the session profit.
- Reset: if a bet loses at any step, next bet = u.
- Profit lock: after N wins, stop and book profit (end session or restart at u with the same limits).
Example: u = 100 THB, N = 2 (100 → 200). If you win both, you stop the cycle rather than risking the next round immediately.
-
Precompute exposure and write the session stop-loss (L)
Choose a session stop-loss L that ends the session regardless of "being due." A simple conservative approach is to stop after a fixed number of failed attempts (resets).
- Attempt-based stop: stop after A losses (at any step) within the session.
- Money-based stop: stop when cumulative P&L ≤ −L.
Example: u = 100 THB, A = 6. If you record 6 losses total (even if interspersed with wins), you end the session.
-
Execute with strict logging (one line per bet)
Track: step k, stake, result (W/L), and session P&L. This is where a spreadsheet or reverse martingale calculator helps prevent "creative" bet sizing.
Example log fields: "k=1, stake=200, W, P&L=+300 THB".
Knowing when to lock in wins: stop rules, caps and session targets
- I stop the cycle after N consecutive wins (no extra press "because it feels hot").
- I stop pressing immediately if the next stake would exceed my bet ceiling C or venue max.
- I end the session when I hit my session stop-loss L (money-based) or A total losses (attempt-based).
- I end the session at a predefined session profit target T (small, realistic, and banked).
- I do not increase u after a loss; u is fixed for the whole session.
- I do not "reload" the session bankroll S if I bust it; the session is over.
- I re-check that the game/market conditions match assumptions (same payout, same rules, same odds type).
- I can state my current step k and next stake without hesitation (if not, I pause and reset).
Tailoring Paroli across games: roulette, blackjack, sports bets and slots
- Roulette: pressing on outside bets can feel "smooth," but the house edge remains; a common error is ignoring table limits that truncate your progression right when a streak arrives.
- Roulette: don't mix bet types mid-cycle (outside to inside) because payout changes break your intended risk profile and profit-lock math.
- Blackjack: Paroli is easiest to implement if you keep rules consistent; a frequent mistake is changing strategy or side bets while also pressing, which makes results harder to control.
- Blackjack: avoid raising bets after "good hands" that were actually high-variance (e.g., doubles/splits) unless your cap and ceiling already account for swings.
- Sports bets: odds changes matter-pressing with different prices can invalidate your planned progression; don't treat a higher-odds next bet as equivalent to the previous step.
- Sports bets: don't press into correlated markets (same match outcomes) thinking it's diversification; correlation can amplify drawdowns within one "session."
- Slots: Paroli rarely maps cleanly because "wins" vary in size; a common error is counting any small payout as a "win step" even when it doesn't cover the stake.
- Slots: if you insist on using it, define a "win" as net profit on the spin (payout minus stake) and expect very frequent resets; otherwise the progression becomes arbitrary.
Stress tests, variance limits and measurable failure modes
Use these as alternatives or guardrails when Paroli's failure modes show up (tilt, limit truncation, inconsistent pricing, or session drift).
- Flat betting with strict session limits
Appropriate when you cannot reliably define "wins" (slots) or odds are inconsistent (some live betting). Rule: constant stake u with the same stop-loss L and profit target T. - Fractional Kelly (sports only, when you have an edge estimate)
Appropriate when you can estimate probability and price. Use a small fraction to reduce model error impact; if you can't quantify edge, don't pretend Paroli substitutes for it. - Fixed-profit staking (target profit per bet)
Appropriate when odds vary and you want consistent profit size rather than a win-streak press. You compute stake from desired profit and odds, with a hard stake ceiling. - Stop-after-one-win micro-sessions
Appropriate if you notice you give profits back after streaks. Rule: after any win, stop the session immediately; it's blunt, but it eliminates "heat-check" overtrading.
Practical clarifications and borderline scenarios
Does the Paroli betting system create a mathematical edge?
No. It changes bet sizing across outcomes; it does not change the underlying expected value of the game or bet.
How many steps should I use for a risk-aware reverse martingale strategy?
Keep it small (commonly 2-3 presses) so table/market limits and your own ceiling don't force bad decisions. More steps increase the chance you hit limits or tilt during a rare long streak.
Can I use a reverse martingale calculator to choose my base unit?
Yes-use it to precompute the sequence (u, 2u, 4u, ...) and verify it stays under your ceiling C and venue limits. Do not use it to justify increasing u mid-session.
What is the cleanest stop-loss rule for intermediates?

Attempt-based stops are simplest: end the session after A total losses (at any step). Money-based stops are fine too, but require cleaner logging.
Should I press after a "win" that barely pays (especially on slots)?
No. Count a win step only when the result is net profit on that spin/bet; otherwise you are pressing after a loss in disguise.
Is Paroli one of the best bankroll management betting strategies?
It can be a disciplined session framework if you cap steps, fix u, and lock profits early. It is not a substitute for bankroll allocation, stop-loss rules, and avoiding negative-EV situations.
What breaks Paroli fastest in sports betting?
Changing odds types and correlation. If your next "step" is at a very different price or on a correlated outcome, your intended risk profile is no longer the one you planned.


