A live dealer working a standard blackjack table on a regulated US-facing platform sees roughly 3% less of the table than the camera's field of view suggests, according to floor managers who have audited streamed tables against physical pit layouts. The gap is not a training problem or a staffing shortfall. It is geometry: the overhead camera that players watch sits at an angle that compresses the far rail, and the dealer's own sightline to the sixth and seventh boxes is blocked by the chip tray, the discard rack, and their own hands during a pitch. On a seven-box table, that blind zone runs from about 10 o'clock to 2 o'clock at the far end.
Where the 3% figure comes from
The number is an estimate, not a published metric. No regulator requires operators to disclose dealer sightline coverage, and studios treat floor-layout data as proprietary. But the arithmetic is straightforward. A 72-inch semicircular blackjack table gives the dealer an unobstructed view of the first four boxes and a partially obstructed view of boxes five through seven once cards leave the shoe. Camera placement for the stream typically sits higher and further back than the dealer's eyeline, adding 8 to 12 degrees of downward tilt. That tilt is what makes the far boxes look fully visible on a player's screen when they are not fully visible to the person dealing.
Why the camera angle matters more than the dealer
Players assume the camera is the dealer's eyes. It isn't. The camera is a broadcast instrument tuned for card legibility, chip stack clarity, and a flattering table composition. A dealer glancing down to pitch a card from a six-deck shoe has a narrower cone of vision than a 1080p feed cropped to the felt. The practical result: a card that skims the edge of box seven can register on stream before the dealer has confirmed it landed flat. In a physical casino, the pit boss catches that. On a streamed table, the software and the shuffle machine usually do.
What this costs the house — and the player
The 3% sightline gap does not translate to a 3% house edge swing. Game protection on live tables comes from three layers: the dealer, the shuffle machine's card-recognition, and the software's bet and outcome logging. The dealer is the weakest of the three. Where the gap shows up is in dispute resolution. If a player claims a card was misdealt to box six, the review relies on the same camera angle that created the ambiguity. Most operators resolve these in the player's favor below a set threshold — commonly $500 on regulated US platforms — because fighting a $200 dispute costs more in support time than the hand is worth.
The variance question
Dealers working tighter sightlines also deal faster, and faster dealers compress the decision window. A table running 120 hands per hour instead of 90 gives players less time to read the far boxes themselves. That is not a rigged game. It is a pace-of-play tradeoff that favors the house on marginal decisions and the player on volume.
The part nobody is measuring
Studios track dealer error rates, hand speed, and customer satisfaction scores. They do not, as a rule, publish sightline coverage per table configuration. That leaves an open question: if a 3% blind zone is acceptable on a seven-box table, what does the number look like on the eight-box and double-deck variants now rolling out in Pennsylvania and New Jersey? Nobody has said. Until someone does, the safest read is the one experienced players already use — treat the stream as a record, not a referee, and keep your own count of what actually hit the felt.