By Samuel Thomas.

Michael.Dalton’s Encyclopedia of Blackjack builds its rule-variation chart on a six-deck game in which the dealer stands on soft 17 and a natural pays 3 to 2. Against a basic strategy player, that game gives the house an advantage of about 0.56%.

Casino table reports usually show hold in the teens or higher, and both numbers are right. The chart divides the house’s expected win by the money bet, while a table report divides the actual win by the cash and credit dropped at the table, a far smaller base.

Private operators have taken bets in Alberta’s regulated online market since July 13, 2026, and the province has yet to publish results from it. Its first figures will describe online play, which the industry measures a different way again.

Six-To-Five Costs Almost Seven Times What Soft 17 Does

The same chart prices each rule change. Paying a natural at 6 to 5 instead of 3 to 2 adds 1.36 percentage points to the house advantage, and letting the dealer hit soft 17 adds 0.20.

A pencil confirms the bigger figure. From a six-deck shoe, a player’s first two cards form a natural on about 4.75% of hands. The dealer also holds one on about 4.56% of those hands, forcing a push, so a paid natural turns up on roughly 4.53% of all hands.

At 6 to 5, each of those naturals pays 1.2 times the bet instead of 1.5. Giving up 0.3 of a bet on 4.53% of hands averages out to 1.36% of each bet, just as the chart says. Hitting soft 17 costs far less because it only changes how the dealer plays one total.

Stacked together, the two rules lift the base game from 0.56% to about 2.12%, close to four times the starting advantage.

Rows add the chart’s adjustments to its base game, which the chart itself calls estimates.

A Drop Box Counts Buy-Ins, Not Bets

Nobody at a live table writes down each wager. A player hands over cash, the dealer drops it into the table’s drop box and pushes chips across, and that box, rather than the stream of bets, is what a land-based casino counts.

Table hold is the win measured against that drop. Writing in Global Gaming Business in 2024, Jeff Hwang called table hold “a completely different statistic” from slot hold, one with “relatively little relation to the house advantage.”

The drop is not one fixed quantity either. Wynn Resorts’ quarterly filings define it three ways: cash plus net markers in Las Vegas, cash plus gross markers at Encore Boston Harbor, and cash in the box plus cash chips bought at the cage in Macau. A marker draws on casino credit, so the way a casino counts credit shapes the base.

Chips Get Bet Again And Again Before Anyone Counts Them

Drop records a buy-in once. The chips from that buy-in then move through hand after hand, and every one of those bets carries the house advantage while the drop records nothing new.

Take a player who buys in for $200 at a six-deck table and bets $2,000 in total before leaving, as winning hands refill the stack. At 0.56%, the expected loss on those bets is $11.20, or 5.6% of the drop.

Switch that table to 6:5 with the dealer hitting soft 17, and the expected loss on the same $2,000 rises to $42.40, or 21.2% of the drop. A 1.56-point change on the chart has become a 15.6-point change in hold, because each dollar of drop was bet ten times.

That multiple moves on its own. It rises when players stay longer or bet their winnings back, and it falls when someone buys in big, plays a few hands and leaves. A table’s hold can drift for those reasons while every rule on the felt stays the same.

Online Tables Divide By Every Wager

An online blackjack game has no drop box. The platform logs each bet as it settles, so online market reports can set what operators kept against the full amount wagered, the base the house advantage uses.

That base yields far smaller percentages. iGaming Ontario’s December 2025 figures, as Yogonet reported them, showed C$320.5 million in online casino revenue on nearly C$8.3 billion in casino wagers, just under 4%, in a category that mixes slots with table games.

Land casinos use a wager-style base where they can track the chips. Wynn measures its rolling-chip play in Macau against turnover, which its latest quarterly report defines as the sum of all losing rolling-chip wagers, and it notes that this larger base yields a lower win percentage than mass-market play.

In the second quarter of 2026, Wynn Palace’s rolling-chip tables kept 2.97% of turnover while its mass-market tables kept 29.7% of drop. Different players stand behind the two figures, but a tenfold gap inside one resort shows how much the choice of base matters.

Find the base before comparing anything. An online figure of 3% does not mean friendlier tables than a land figure of 20%, and it may describe tougher ones.

Alberta’s Record Lists Net Sales Without Wagers

Before private operators arrived, Alberta’s online gambling record came from PlayAlberta.ca, the platform Alberta Gaming, Liquor and Cannabis (AGLC) operates. AGLC’s 2024-25 annual report lists the site’s results as net sales, and its financial statements recognize gaming revenue from the net win or loss on play.

For the year to March 31, 2025, PlayAlberta.ca posted casino net sales of C$197.3 million and live dealer net sales of C$21.4 million, out of C$269.9 million in all. Those are win figures. The report prints no wager totals beside them, so neither converts into a hold percentage.

The split still helps table players, since live dealer games had their own line, and whether the new market’s reports keep it will shape how much a blackjack reader learns from them.

A Sportsbook’s Hold Starts From The Handle

Sports betting uses the online base too. A sportsbook’s hold is its win as a share of handle, the total amount bet, so it reads like an online table figure, not a drop-box number.

When the new market’s first sportsbook figures arrive, divide revenue by handle before judging them. LineUps covers online sports betting in Alberta on a dedicated provincial page, a useful companion to those numbers, and every player in that market must be 18 or older and inside the province.

Alberta’s sportsbook record is thin so far. PlayAlberta.ca’s sports line produced C$8.0 million in net sales in 2024-25, and again the annual report left handle out.

Pricing sets the margin any handle figure starts from. When a book lists both sides of a game at -110, a bettor risks $110 to win $100. If $110 arrives on each side, the book collects $220, pays the winner $210 and keeps $10, or 4.55% of handle.

Actual hold wanders from there. Parlays stack the margin of every leg they combine, and heavy action on a side that wins can leave a book behind for a week.

Side Bets And Misplays Sit Inside The Online Figure

A wager-based table figure captures blackjack as people actually play it, which is rarely the game the chart assumes.

Side bets push the number up first. A Blackjack Review guest post on September 13 put a common 25/12/6 Perfect Pairs paytable at a house edge of about 4.1% in an eight-deck example, more than seven times the main game’s 0.56%. Every dollar placed in the side circle raises the blended percentage for the table.

Decisions push it up next. The chart’s 0.56% belongs to a player who follows basic strategy on every hand, which another Blackjack Review post described, “applied perfectly,” as “the single largest controllable factor in any blackjack session.” Standing on 15 against a dealer’s 10, or skipping a split the chart calls for, hands the house a bigger share of that hand.

A published online blackjack figure above 1%, then, does not prove the tables pay 6:5, because side bets and misplays can produce it at tables with the best rules in the lobby. A session budget should count the side bets as well as the main wager.

The Alberta iGaming Corporation (AiGC) directs anyone worried about their gambling to 211 Alberta, a free and confidential service that answers by phone, text or online chat 24 hours a day.

Wynn Publishes An Expected Range For Its Tables

Wynn Resorts treats a quarter of table hold as one reading inside a band, printing an expected range beside its Las Vegas and Boston table results. Its second quarter of 2026 showed how wide those bands run.

As a Yahoo Finance recap of the quarter noted, Wynn’s Las Vegas tables kept 23.9%, inside a 22% to 26% range. Encore Boston Harbor kept 18.1%, at the bottom of its 18% to 22% range and down from 21.3% a year earlier.

Long records show the same spread. The UNLV Center for Gaming Research tracked Nevada’s monthly blackjack hold from January 2004 through February 2026 and found a 12.77% average, with single months as low as 8.23% and as high as 18.12%.

A market in its first months has less play behind each figure, so its percentages can swing further. A few large player wins at live dealer tables, or a run of weekends when favorites cover, could push an early Alberta figure well away from where it later settles.

AiGC, a Crown corporation, has said it will publish quarterly market performance reports, although the Reports section of its website still holds only quarterly expense reports and annual compensation disclosures. Its first market report will open that band, not close it.

Readers following the market between releases can find betting coverage on the LineUps YouTube channel, and table players can keep the rules card close, since it still describes every hand they play.

IMAGE CREDIT: Marisol Quintero

BLOG MENU