Let It Ride Strategy: When to Keep or Withdraw Your Bets

Optimal Let It Ride strategy means keeping removable bets only when they have positive expected value and withdrawing the rest. In a few cases, the expected value is zero: keeping or withdrawing is equally optimal, but withdrawing reduces your exposure. Even with correct decisions, the game as a whole has negative expected value for the player.

First, confirm your table’s paytable. The rules below assume the standard payouts, including 1000 to 1 for a royal flush and 1 to 1 for a pair of tens or higher. Different payouts, fees or aggregate payout caps can change the calculation.

Here, expected value, or EV, means the expected net profit from one removable bet, expressed in base-bet units. It is not the expected return from the entire hand. Withdrawing that bet has an EV of zero: you get your money back without a profit or loss on that bet.

How the hand works and which bets you can withdraw

You place three equal bets—#1, #2 and #3—but only the first two can be withdrawn. Bet #3 must remain until the hand is settled.

You receive three cards and decide whether to keep or withdraw bet #1. The first community card is then revealed. Using those four known cards, you independently decide whether to keep or withdraw bet #2. The second community card completes your five-card hand, and each remaining bet is settled according to the paytable.

You cannot exchange cards, and there is no conventional comparison against a dealer hand. Withdrawing a bet also does not fold the whole hand: your remaining bets stay in play.

For example, with a base bet of $5, you initially place $15. You might withdraw bet #1 and recover $5, then keep bet #2 after seeing the first community card. Two $5 bets would remain:

• If the final hand does not pay, you lose $10.
• If you finish with a pair of tens or higher paying 1 to 1, your net profit is $10. Including the return of the two stakes, you receive $20.

Your final loss depends on how many bets remain. If you withdraw both removable bets, a nonpaying hand costs only the mandatory $5 bet.

The standard Let It Ride paytable

The standard game pays for a pair of tens or higher and every stronger poker hand. Anything below a pair of tens loses; it is not a push. Two low pairs still pay as two pair.

“To 1” describes profit in addition to the returned stake. If all three $5 bets remain for a royal flush:

3 × $5 × 1000 = $15,000 net profit, plus $15 in returned stakes = $15,015 total paid.

This example assumes no aggregate payout cap. Before playing, check whether there is a cap and whether the posted payouts are stated as “to 1” or “for 1.” A payout stated “for 1” includes the stake rather than adding its return separately.

The probability of a royal flush in a random five-card hand is 1 in 649,740. That does not mean a royal flush becomes “due” after a particular number of hands.

Top payout rows from historical Let It Ride paytable variants in a California Department of Justice document

Original English excerpt from the California Department of Justice rules, Rev. 01/13—January 2013. The historical document shows differences among the top payout rows; it does not verify the paytable at a current venue. See the original PDF, pages 1 and 3–5.

Three-card strategy: when to keep bet #1

Keep bet #1 with an existing paying hand or one of the qualifying suited draws below. Withdraw it in every other case.

A high card in this strategy is a 10, J, Q, K or A. Suited cards share the same suit, and the draw rules require distinct ranks.

To count gaps, put the ranks in order and count the span from the lowest to the highest, including both ends. For this test, use J = 11, Q = 12, K = 13 and A = 14, not 1. The span is highest rank − lowest rank + 1.

The examples show how the conditions work:

• 8♠, 9♠, J♠: the span is 8–9–10–J, covering four ranks. Only the 10 is missing, and the J supplies one high card. Keep the bet.
• 7♠, 10♠, J♠: the span is 7–8–9–10–J, covering five ranks. Two ranks, 8 and 9, are missing, but you have two high cards: 10 and J. Keep the bet.
• 3♠, 4♠, 5♠: these consecutive suited cards qualify even without a high card.
• 2♠, 3♠, 4♠ and A♠, 2♠, 3♠: these are the exceptions. Withdraw the bet.

An ace can be low only as part of the wheel straight, A–2–3–4–5. Ranks do not wrap around: K–A–2 is not a consecutive sequence.

Three offsuit high cards do not qualify on their own. Neither does a low pair such as 8♠, 8♥, Q♦.

Four-card strategy: when to keep bet #2

Keep bet #2 with a paying hand, any four suited cards, or a genuine open-ended straight draw containing at least one high card. Two other draw categories have zero incremental EV, so keeping or withdrawing is equally optimal.

An open-ended straight draw has four distinct consecutive ranks and can complete a straight at either end. For example, 8–9–10–J makes a straight with a 7 or a Q. Eight cards in the remaining deck complete it: eight outs.

An inside straight draw, also called a gutshot, needs an internal missing rank. With 10–J–Q–A, only a K completes the straight. J–Q–K–A also needs just one rank, a 10, but at one end. Both contain four high cards and have zero EV for the removable bet. If all four cards are suited, however, keep the bet under the four-suited-card rule: that makes the decision positive EV.

An inside straight draw with only three high cards does not qualify. For example, offsuit 9–J–Q–K needs a 10, but keeping the bet has negative EV. Withdraw it.

A–2–3–4 is not open-ended. Only a 5 completes a straight; there is no rank below the ace to complete the other end. Another ace would give a paying pair, but those additional wins are not enough to justify keeping the bet unless all four cards are suited.

Four consecutive cards are not yet a made straight. You need five cards. Assuming no other cards are known, there are 48 possible final cards:

These probabilities describe only completion of the named hand. A betting decision must also account for other paying results, such as pairing a high card. Straight outs alone do not determine EV.

Why similar draws can require different decisions

The correct decision depends on the removable bet’s expected net profit, not just its number of outs. Withdrawing gives zero; keeping is justified when the average net result is positive.

With four known cards, you can check all 48 possible final cards. For one base-bet unit, a straight earns +5, a pair of tens or higher earns +1, and a nonpaying hand loses −1.

For 4♠, 5♥, 6♦, 7♣, eight cards complete a straight and the other 40 lose:

Keeping and withdrawing are therefore equally optimal. Withdrawing simply removes that additional bet from settlement.

For 8♠, 9♥, 10♦, J♣, eight cards complete a straight, another six make a pair of tens or jacks, and 34 lose:

At a $5 base bet, keeping bet #2 has an expected net profit of $1.25. This is an average across possible completions, not a guaranteed outcome on the next hand.

For A♠, 2♥, 3♦, 4♣, four fives make a straight, three remaining aces make a paying pair, and 41 cards lose:

Withdraw the bet.

With three known cards, the calculation requires all 1176 unordered pairs of community cards from the remaining 49 cards. Evaluate each resulting five-card hand using the net paytable: +1000 for a royal flush, +200 for a straight flush, and so on, down to +1 for a high pair and −1 for a loss. Add those results and divide by 1176.

An independently coded enumeration using that method gives:

The displayed three-card EVs are rounded to seven decimal places. These calculations assume a uniformly shuffled standard 52-card deck, no other known cards, and no fees, side bets or payout caps.

The main strategy and game statistics are supported by the Wizard of Odds mathematical analysis. The enumerated examples above are separate checks using the stated method, not outputs from the embedded calculator.

House edge: why you see 3.51%, 2.85% and 1.17%

These percentages use different denominators; they do not describe different versions of the game. With the standard paytable and optimal strategy, expected loss per hand is 3.51% of one base bet.

The reference mathematical analysis gives an average of approximately 1.232 base-bet units remaining at settlement. That reference convention keeps bets in zero-EV cases. Relative to the average amount left at settlement, expected loss is approximately 2.85%.

Dividing the same expected loss by all three initially placed units gives 1.17%. This is only an arithmetic normalization. It does not turn withdrawn bets into losing bets or make them part of the amount actually settled.

RTP is the long-run return relative to a stated betting denominator. Consequently, 96.49% is not a universal RTP for all three bets. Using the average amount left at settlement as the denominator gives an RTP of approximately 97.15%. Withdrawing in zero-EV cases can change the average number of retained units, but it does not change expected loss.

For more on why the denominator matters, see our house edge explanation.

How often the final hand pays

A final five-card hand pays in 23.8780% of cases and does not pay in 76.1220%. Those are hand frequencies, not RTP or the proportion of money returned.

Category counts are calculated combinatorially from the 2,598,960 possible five-card hands. Probabilities are rounded to three decimal places. The chart shows hand frequency, not each category’s contribution to payouts.

Calculator: check its recommendations against the tables

The embedded calculator is a limited heuristic with known mistakes, not an exact analyzer. The decision tables above override its recommendations, even if the interface calls a decision optimal.

The tool misses some positive-EV keeps: it pulls suited 3–4–5, a one-gap suited draw with one high card, and a two-gap suited draw with two high cards when it should keep the bet. It also wrongly keeps A–2–3–4 when the four cards are not all of the same suit; keeping is correct when all four are suited. With offsuit 10–J–Q–A, the tool pulls the bet, which is a valid optimal decision at zero EV, but its explanation does not identify that keeping and pulling are equally good. Separately, it can incorrectly label a four-card sequence as a made straight, which requires five cards.

It also has these limitations:

• It does not check whether you entered the same card twice.
• It does not calculate numerical EV or simulate payouts.
• It does not analyze side bets or alternative paytables.
• The ante input does not change its recommendation or calculate profit.

Here, the ante field means one base-bet unit, not the combined total of all three bets. The mandatory third bet is also one such unit.

For a straightforward example, select three-card mode, enter Q♠, J♠, 10♠, and enter 5 in the ante field. The tool recommends keeping the bet, which is correct for this hand.

Three-to-a-royal-flush draw with Q♠, J♠ and 10♠ in the localhost Let It Ride calculator

Screenshot of the English tool running on localhost, not a production-site test. These three suited cards form a royal flush draw, not a completed royal flush. The three-bet payout example above is a separate calculation, not a calculator feature.

To enter a fourth card, turn on the fourth-card toggle, then select its rank and suit in the fields that appear. Compare the result with the four-card decision table; the tool’s explanation is not proof that its decision or hand description is correct.

Next, try 3♠, 4♠, 5♠ in three-card mode. The tool recommends withdrawing, but that recommendation is wrong. The correct decision is to keep: exact enumeration gives positive EV.

Localhost Let It Ride calculator incorrectly recommending withdrawal with 3♠, 4♠ and 5♠

Screenshot of the English tool running on localhost, not a production-site test. Its recommendation for 3♠, 4♠, 5♠ is incorrect. Follow the consecutive suited-card rule and keep the bet.

What play can cost and what to check first

At a $5 base bet over 100 hands, the standard model gives an expected loss of $17.55. That is neither a loss limit nor a measured result from a two-hour session. Your actual result can differ substantially.

Each hand begins with $15 on the table. Under the reference convention of keeping zero-EV bets, the average amount left for settlement is approximately $6.16 per hand. Withdrawals explain the difference between the initial amount placed and the amount ultimately at risk, but they do not remove the house edge.

Before playing, check:

• The exact paytable and whether payouts include or exclude the returned stake.
• Minimum and maximum bets.
• Fees and any aggregate payout cap.
• Separate rules for any side bets.

Side bets are optional. They can use their own three-card, five-card or progressive paytables. Without those rules, they cannot be evaluated, and the main game’s 3.51% house edge does not apply to them.

Set an affordable spending budget and a time limit in advance. Do not increase your bets to chase losses: correct decisions do not guarantee that your bankroll will last or that a session will be profitable.

Do not confuse the poker game with a blackjack progression

This guide covers the standalone Let It Ride poker table game, not the Let It Ride betting progression in blackjack.

Mississippi Stud is a different game with different betting mechanics. Its staged card reveals do not make these decision tables transferable to it.