The Blackjack Advantage Is Real. Wendover's Math Leaves Out the Most Important Part.
Wendover Productions presents card counting as a remarkably reliable way to turn blackjack into a profitable business. The mathematics behind card counting is real. The problem is that the video moves from a mathematical possibility to an apparently achievable income without showing readers enough of the calculations, conditions or risks separating the two.
The distinction matters. A person watching this episode could reasonably walk away believing that keeping a running count, recognizing a favourable shoe and raising their bets is a relatively straightforward route to earning $110 an hour. That is a much stronger promise than the evidence presented can support.
The first big comparison uses an unidentified accounting measure
The video says MGM generates approximately $1,138 per square foot of casino floor. Its cited 2019 annual report tells a more complicated story.
MGM reported 2.543 million square feet of casino space and $6.518 billion in casino revenue for 2019. Its consolidated Adjusted EBITDAR was $3.016 billion that year, compared with $2.838 billion in 2018 and $2.858 billion in 2017.
FinancialFilings+2
There is an important clue: the three-year average Adjusted EBITDAR divided by approximately 2.55 million square feet produces roughly $1,139 per square foot. That almost exactly reproduces Wendover's $1,138 figure.
But Adjusted EBITDAR is not casino revenue. It is a non-GAAP earnings measure that excludes interest, taxes, depreciation, amortization, rent and other expenses. MGM also warns investors against treating it as an alternative to net income or a complete measure of operating performance.
FinancialFilings
More importantly, MGM's domestic resorts generate more than half their net revenue from non-gaming operations, including hotels, food, beverages and entertainment.
FinancialFilings
If the $1,138 figure is indeed the reconstructed Adjusted EBITDAR ratio, it attributes earnings from the wider resort business to the casino-floor square footage.
That is a useful measure of property economics, perhaps. It is not proof that the gambling floor itself produces that amount.
The distinction matters when the video uses the figure to illustrate the extraordinary productivity of casino gambling.
A running count is not the number that tells you what to bet
The video explains Hi-Lo counting correctly at a basic level, but leaves out a critical conversion.
It says a running count of +20 can give a player a greater-than-50% chance of winning the next round.
Pasted text.txt
In a multi-deck game, however, the running count must be divided by the number of decks remaining. This produces the true count, which is the more useful measure of the remaining deck's richness. Wizard of Odds explicitly explains this distinction.
Introduction & Basics
A running count of +20 with five decks remaining produces a true count of +4. The same running count with one deck remaining produces a true count of +20. Those are dramatically different situations.
The episode never supplies the remaining-deck calculation or a probability model supporting its greater-than-50% statement.
There is another distinction: expected profit is not the same as probability of winning a hand. Blackjack counting creates an advantage through several mechanisms, including better standing decisions, insurance, doubling, blackjacks, surrender and splitting. Wizard of Odds estimates that standing accounts for 40% of the advantage and insurance another 34%.
Introduction & Basics
The player does not need to win most hands to make money. The size and timing of wagers, together with blackjack's unequal payouts, matter enormously.
A separate simulation by blackjack mathematician Norm Wattenberger illustrates the practical difference. Using six decks, Hi-Lo, a 1:15 betting spread and favorable rules, it estimates that only 52% of 50-round sessions finish profitably and 56% of 500-round sessions do. Even across 10,000-round trips, the estimated success rate is 75%.
blackjackincolor.com
These figures do not disprove the possibility of a profitable high-count hand. They demonstrate why a positive mathematical advantage should not be presented as a reliable short-term winning probability.
The 1.182% advantage is real, but the conditions matter
To be fair, Wendover did provide a source trail. Its video description links directly to Wizard of Odds for the card-counting mathematics.
AYTA
The published simulation reproduces the video's 1.182% player advantage. But its assumptions are unusually favorable.
The model uses:
- Six decks.
- Five decks dealt before reshuffling.
- A 1:15 betting spread.
- One betting position.
- Full card-counting strategy indices.
- Dealer stands on soft 17.
- Double after splitting, surrender and resplitting under favorable rules.
The simulation also assumes highly accurate play and carefully controlled betting adjustments.
Wizard of Odds
The most important omission is deck penetration.
The Wizard of Odds table identifies 4.5 decks as the normal penetration benchmark for its six-deck comparison. At that depth, the same 1:15 spread produces an advantage of approximately 0.862%, rather than 1.182%. That is roughly 27% less.
Wizard of Odds
Wendover does explain that shuffling after four decks reduces the advantage substantially. But it skips the more informative intermediate comparison: 4.5 decks.
The distinction is not trivial. The headline result depends on allowing the player to exploit a relatively deep shoe.
The mathematics survives. The suggestion that this is a representative casino opportunity does not.
$110 an hour is an expected return, not a wage
The video assumes one round per minute and an average bet of approximately $200, then estimates earnings of $110 per hour.
Pasted text.txt
There is a denominator problem that the narration never resolves.
The Wizard of Odds simulation calculates its 1.182% advantage against total initial wagers. Its average-bet figure, however, includes additional money wagered through doubles and splits. Those are not interchangeable measures.
Wizard of Odds
If $200 means the average initial wager, multiplying by 60 rounds and 1.182% produces $141.84 per hour, not $110. If it means total money wagered after doubles and splits, a different calculation is needed.
The $110 figure might be reconcilable with a lower average initial wager. The video simply does not show the conversion.
The next number comes from another source altogether.
Wendover cites Blackjack Apprenticeship for its claim that a player with $100,000 can generate $170 per hour while maintaining a 1% risk of ruin.
Pasted text.txt
That source is a commercial card-counting training business. Its published bankroll model assumes perfect blackjack play and a 1:12 betting spread under standard rules. It describes the $100,000 scenario as producing $170 in hourly expected value with mediocre rules. Its higher-limit example uses different rules and a 1:16 spread.
www.blackjackapprenticeship.com
Blackjack Apprenticeship
This is not the same model as the Wizard of Odds simulation using a 1:15 spread and five-deck penetration.
The source also sells $299 bankroll coaching, a $997 Elite membership and proprietary betting software. That does not establish that its calculations are wrong, but it does mean the figures come from a commercially interested training provider rather than an independent audit.
www.blackjackapprenticeship.com
Blackjack Apprenticeship
Most importantly, Blackjack Apprenticeship itself defines expected value as a statistical estimate and explicitly warns that actual results rarely equal expected value in the short run.
www.blackjackapprenticeship.com
Blackjack Apprenticeship
The video uses the phrase "$170 an hour" without giving that distinction sufficient weight.
An expected $170 per hour is not a guaranteed $170 paycheck. A player can lose substantial money while following a profitable strategy.
The distinction between those two statements is the difference between a mathematical model and a business plan.
Legal advantage does not mean a right to keep playing
The video correctly explains that mental card counting is not generally treated as cheating. But it understates the practical limitation.
In Lyons v. State (1989), the Nevada Supreme Court explicitly recognized that card counters are not subject to criminal prosecution merely for counting cards, while also recognizing that casino management may deny them the right to play.
Justia Law
Nevada law separately prohibits players from using electronic, mechanical or software devices designed to track cards, calculate probabilities or obtain a gaming advantage.
CRIMES AND LIABILITIES CONCERNING GAMING
A 2011 federal court decision also confirmed that Nevada casinos may exclude suspected card counters at their discretion.
Justia Law
This is a significant practical distinction.
A professional counter may have a mathematically favorable game, substantial capital and flawless technique, but the casino still controls access to the table.
The ability to count cards is not an entitlement to continue playing.
That makes the business comparison less romantic than the video suggests. A professional counter is operating in an environment where the establishment can terminate the opportunity.
The sponsorship and the actual source of the numbers
The video openly discloses its Brilliant sponsorship, and the narrator says he took Brilliant's Casino Probability course while preparing the episode. I found no evidence that the sponsor influenced the financial calculations.
Pasted text.txt
The relevant financial estimates instead come from Wizard of Odds and Blackjack Apprenticeship, both identified in the video's references.
AYTA
The distinction is important. There is no basis for accusing Brilliant of influencing the results. The more relevant question is whether the commercial bankroll estimates have been independently tested and whether viewers are being shown the conditions under which they apply.
They are not given enough information to reproduce those estimates independently.
The central argument survives: card counting can create a mathematical advantage, and casinos have strong financial reasons to suppress it.
But the video repeatedly moves between different kinds of numbers without giving them the distinction they deserve.
Casino revenue is not Adjusted EBITDAR. A running count is not a true count. Expected value is not the probability of winning. A simulated hourly return is not an actual wage. And a bankroll's risk of ruin depends on the betting model and the period being measured.
These are not minor technicalities. They determine whether the central financial story is realistic.
Wendover has explained how a player can theoretically beat a casino. It has not established that the advertised earnings are a dependable professional opportunity.
The mathematical advantage is real. The apparent simplicity of making a living from it is the misleading part.
References
1. How to Beat the Casino, and How They'll Stop You — Wendover Productions, October 8, 2020. The original episode and its description, which lists the underlying research and sponsorship.
2. MGM Resorts International — 2019 Annual Report — MGM Resorts International, February 27, 2020. Provides casino square footage, gaming revenue, Adjusted EBITDAR and the accounting qualifications behind the $1,138 comparison.
3. Hi-Lo Card Counting Strategy — Wizard of Odds. Provides the 1.182% simulation, betting assumptions, penetration comparisons and volatility measurements.
4. Card Counting in Blackjack — Introduction and Basics — Michael Shackleford, Wizard of Odds. Explains true-count conversion and the different ways card counting produces an advantage.
5. Winning Blackjack Sessions — Norm Wattenberger, Blackjack in Color. Quantifies how frequently profitable card-counting strategies actually produce winning sessions and trips.
6. Recommended Blackjack Bankroll and Money Management — Blackjack Apprenticeship. The commercial source behind the $100,000 bankroll, $170 hourly expected value and 1% risk-of-ruin figures.
7. Lyons v. State — Supreme Court of Nevada, June 1, 1989. Establishes the distinction between lawful mental card counting and a casino's ability to refuse service.
8. Nevada Revised Statutes, Chapter 465 — Nevada Legislature, current through April 15, 2026. Defines gaming-related offenses and prohibits unauthorized devices and software used to obtain a gambling advantage.