Theoretical Loss in Video Poker

Theoretical loss is the casino’s estimate of the average cost attached to a player’s video poker action. It is built from total wagers and an expected house edge; it is not a prediction of what the next hand, hour, or trip will produce. A player can generate positive theoretical loss and still leave ahead, while another player can lose far more than the estimate in a short cold session.

Start with action, not the cash you inserted

The most important input is coin-in: the total value of all wagers made. Cash inserted is only funding. The same $100 can be wagered, partially returned, and wagered again many times.

Suppose you play 800 hands at $1.25 per hand. Your coin-in is $1,000 even if you began with only $100 in the machine. If the exact game and strategy produce a 99.54% return, the corresponding house edge is about 0.46%. The simple theoretical-loss estimate is therefore $4.60 on $1,000 of coin-in.

That small number does not describe the shape of the session. You might be down $150 after a long run without premium hands, hit a four of a kind and recover, or catch a royal flush and finish far ahead. The estimate prices the action in the long run; it does not schedule the outcomes.

For the inputs behind the calculation, review coin-in in video poker, video poker payback percentage, and video poker house edge.

The four numbers that create a theo estimate

In compact form:

Coin-In = Bet Per Hand × Hands Played
House Edge = 1 - RTP
Theoretical Loss = Coin-In × House Edge

Using the 800-hand example:

Coin-In = $1.25 × 800 = $1,000
House Edge = 1 - 0.9954 = 0.0046
Theoretical Loss = $1,000 × 0.0046 = $4.60

The formula is simple. Choosing the right edge is the harder part.

Video poker adds a strategy condition that slots do not

A conventional slot’s mathematical return is not changed by the player’s choice of which symbols to keep. Video poker is different because the player decides what to hold before the draw. The advertised or analyzed return for a paytable normally assumes a specified strategy, often optimal or near-optimal play.

That means two ideas must be kept separate. The machine’s theoretical return comes from its paytable and strategy model. The player’s realized long-run return can be worse if the player repeatedly makes lower-value holds.

Consider a 9/6 Jacks or Better player who frequently breaks paying pairs to chase visually attractive draws or keeps irrelevant kickers with a pair. The cabinet has not changed. The paytable has not changed. The player’s decisions have reduced the value extracted from that paytable.

The Wizard of Odds video poker summary is useful for seeing how returns vary by game and schedule, while the video poker analyzer illustrates why strategy belongs in the return calculation.

One hold decision shows why expected value and result diverge

Imagine a Jacks or Better hand:

K♠ Q♠ J♠ 7♦ 2♣

The attractive part of the hand is K-Q-J suited. A player who repeatedly keeps an extra unrelated card because “four cards feel safer” can reduce the expected value of the draw. Yet even the correct hold can miss, while an inferior hold can occasionally get lucky.

This is why theoretical loss cannot be inferred from memorable hands. Expected value considers every possible draw after a decision and weights each outcome by probability. The actual session records only the sequence that happened.

A royal flush can dominate the player’s memory and the machine’s short-term result. It does not erase the fact that thousands of ordinary wagers created the action around it.

What the casino is trying to estimate

Casinos use theoretical win or theoretical loss as a management estimate of the value of play. Depending on the property and system, inputs can include denomination, average wager, coin-in, time played, game type, an assigned hold percentage, or a combination of those measures.

That estimate can influence:

• player offers and mailers;
• host attention and discretionary comps;
• reinvestment percentages;
• machine and bank performance analysis;
• comparisons between actual and expected hold;
• profitability reviews by game type or denomination.

A player who hits a royal may have a strong actual result but still generate substantial theoretical value through heavy coin-in. Another player can lose quickly with relatively little action. Using only actual loss to value those two customers would produce unstable and sometimes perverse decisions.

Theo is therefore a smoothing tool. It does not verify jackpots, settle disputes, or prove that a machine behaved correctly. Those questions rely on meters, game logs, approved configurations, surveillance evidence, hand-pay records, and technical controls.

Why a $4.60 estimate can coexist with a $250 loss

Video poker paytables distribute return unevenly. A large share of long-run value can sit in relatively rare hands. In a short sample, those hands may not appear at all.

If a game has 99.54% theoretical return, that does not mean $100 of play will reliably come back as $99.54. It means that under the assumptions behind the calculation, the average return over a very large body of correctly played wagers approaches that percentage.

Variance is the reason the session can look wildly different. Theoretical loss answers, “What is this volume of action worth on average?” Variance answers, “How far can the path wander while getting there?”

Those are different questions, and both matter. If bankroll survival is the concern, video poker bankroll risk and the variance simulator are more informative than theo alone.

Comps should be compared with theo, not mistaken for free money

Players sometimes hear that a casino gives back a percentage of theoretical loss and conclude that comps eliminate the edge. That is too simple.

A property may reinvest some estimated value through free play, food, rooms, points, or offers, but the reinvestment rate varies. The player also faces variance, strategy error, eligibility rules, expiry rules, and the temptation to generate extra action just to earn a benefit.

A $20 benefit is not a bargain if it causes $2,000 of additional coin-in that the player would not otherwise have made. The correct comparison is between the value of the benefit and the expected cost of the extra action needed to earn it.

When theoretical loss is useful to a player

Theo is most useful as a planning number. Before a session, you can estimate how much action a given pace and wager will create and translate that action into an expected long-run cost.

For example, 500 hands at $2.50 per hand produce $1,250 coin-in. At a 1% edge, the theoretical loss is $12.50. At a 3% edge, it becomes $37.50. The session outcomes can still swing far beyond either number, but the comparison clearly shows why paytable quality matters.

It also exposes the cost of speed. Doubling hands per hour roughly doubles coin-in if the wager stays unchanged. Moving from single-hand to multi-hand play can multiply action even faster.

Use the expected loss calculator when you want to test several combinations of bet, pace, time, and edge rather than doing the arithmetic by hand.

Errors that make theo misleading

Several mistakes can make a theoretical-loss estimate look precise while using the wrong assumptions:

• treating cash inserted as coin-in;
• using the best-known return for a game name without checking the actual paytable;
• assuming perfect strategy when the player does not use it;
• ignoring multi-hand exposure;
• converting one winning or losing session into a claim about long-run edge;
• counting comps as guaranteed cash value;
• comparing games at different wager sizes without normalizing total action.

The number is only as useful as its inputs.

Keep theoretical loss beside actual results, not in place of them

Theoretical loss is a model of average cost. Actual win or loss is a record of what happened. A serious analysis keeps both.

If actual results repeatedly differ from expectation over a meaningful sample, casino operators investigate whether the explanation is ordinary variance, a change in game mix, jackpot timing, promotion effects, skilled-player concentration, configuration differences, or an operational problem. They do not assume that one day’s deviation proves anything by itself.

Players can use the same discipline. One session is an outcome. A long series of wagers is evidence. Theo belongs in the long-series conversation.

Continue with actual win vs theoretical win for that comparison, then read RTP vs variance to understand why the average and the path can look so different. Technical standards such as GLI gaming-device standards and Nevada technical standards provide background on the regulated environment in which the machine operates; they do not turn a long-run estimate into a session guarantee.