A live dealer blackjack table running a cam-only feed—no studio presenter, just an overhead camera on the shoe—resolved 9.2% fewer hands in the operator's favor over a 90-day sample than the same game with a live host on camera, according to deal data shared by two mid-size studios that asked not to be named. The gap isn't a rigging story. It's a throughput and error story, and it shows up most clearly in the games where a human voice is doing quiet work the camera can't.

Where the 9% Comes From

The number comes from comparing hand outcomes across the same blackjack variant, same RTP configuration, same shuffle procedure. What changed was the dealer.

Cam-only tables in the sample dealt 4.1% more hands per hour but also logged 2.7% more misdeals and voided rounds. Those voids get reshuffled and replayed, and in the replay the house edge lands slightly differently depending on where the cut card sits. Over 90 days and roughly 610,000 hands, that compounds into the 9.2% swing.

Two things worth flagging. First, the sample is small enough that I'd want a year of data before treating 9.2% as a fixed property. Second, "fewer hands in the operator's favor" is not the same as "better for the player." A spike in voided rounds is a worse experience even when the math tilts your way.

The Baccarat Problem

Baccarat showed the widest variance in the sample, and it's the game where cam-only setups make the least sense. The squeeze—the slow reveal of the third card—is the entire product. Strip the presenter and you're left with a card-flip animation that could just as easily be RNG. Players noticed: session length on cam-only baccarat tables ran 31% shorter than on hosted tables in the same lobby.

Why Operators Are Still Shipping Them

Cam-only tables cost less to run. A hosted table needs a presenter on shift, a studio floor, and the labor scheduling that goes with it. A cam-only table needs a rig, a shoe, and a compliance officer signing off on the feed. At scale, that's the difference between a table that pays for itself in a month and one that takes three.

The studios I spoke with are deploying cam-only feeds mostly as overflow—spinning one up when hosted tables hit capacity on a Friday night. That's a defensible use. The trouble starts when operators treat cam-only as a full substitute rather than a pressure valve.

What Players Should Check

If you're sitting at a live table and can't see a presenter, three things are worth confirming before you buy in:

  • Table limits and void policy. Cam-only tables in the sample voided 2.7% of rounds. Ask what happens to your bet when a round voids—refund, push, or replay.
  • The RTP disclosure. Most jurisdictions require it. A cam-only table should publish the same figure as its hosted twin. If it doesn't, that's a signal.
  • Feed latency. Cam-only rigs in the sample ran 180–240ms behind hosted tables, which matters more in baccarat than blackjack.

None of this is a reason to avoid the format outright. It's a reason to know which table you're on.

The Question Nobody's Answering

Here's what the data doesn't settle: is the 9.2% gap a permanent feature of cam-only dealing, or a teething problem that disappears as rigs and procedures mature? The studios I talked to lean toward the second answer, but neither would commit to a date when their cam-only tables would match hosted performance. Until someone publishes a year of audited hand data, players are making a judgment call on a 90-day sample—and operators are pricing that uncertainty into the table limits rather than the marketing copy.