A. When I first read Frank’s "Beat the Craps Out of the Casinos," way back in the 1990s, my thoughts weren’t a million miles from yours. What and when you wager can’t change the numbers that come up on the dice. Given a random game, the house edge is there, unchanging.

My feeling at the time was that the positives of playing the Five Count lay in reducing exposure to the house edge. By not putting money at risk until the Five Count was reached, the player was extending his session, staying in the action in the hopes of eventually catching a hot roll. The player would still lose 1.41 percent of his money on pass and come bets, but would have less money per hour at risk than a player who didn’t Five Count.

I floated that past a fellow gambling writer, whose response was, "Maybe, but then why play at all? You could reduce your exposure to the house edge to zero by steering clear of craps."

But my view has shifted over the years, with a turning point coming when I started to see dice control in action and Frank shifted terminology in his books and articles from "rhythmic rollers" to "controlled shooters." What if, unbeknownst to you, you find yourself at a table with a controlled shooter? Will the Five Count keep us off a disproportionate share of rolls by random rollers, and put us on a disproportionate share of rolls by the controlled roller?

To go a step further, there aren’t all that many trained controlled rollers. But what if a player who is a random roller most of the time just has a stretch where he’s doing everything right, gets in a rhythm that gives him a controlled roll that depresses the frequency of 7s? Will the Five Count get you onto that controlled roll while keeping you off random rolls?

That, I think gives the Five Count value beyond just reducing exposure to the house edge. It’s not going to pick up a controlled shooter in every session, because there’s not going to be a controlled shooter in every session. But as one who has seen some strong dice controllers in action, I want to be in on their rolls, while limiting wagers on the majority of random shooters.

A. Those hold percentages are not the same thing as the house edge. When gaming board statistics list a hold percentage of 16 percent, it doesn’t mean blackjack players have lost 16 percent of wagers. It means they’ve lost 16 percent of buy-ins.

That includes the effect of rebetting winnings as you play. Let’s take an average player --- not a basic strategy player or an advantage player, but the kind of player who fills most of the seats and bucks a house edge of about 2 percent. And let’s say he buys in for $500, wagers $25 a hand, wins a little, loses a little, and sticks around long enough to play 200 hands. He’s made $5,000 worth of wagers. His expected loss at a 2 percent house edge is $100.

But if he actually loses $100, casino statistics won’t see that as 2 percent of his $5,000 in wagers. They’ll see it as 20 percent of his $500 buy-in. That’s the hold percentage.

Electronic games are reported differently. There, actual wagers are tracked, and what you see on gaming board reports as a hold percentage or win percentage is the percentage of money wagered that’s kept by the house. A casino marketing exec once argued the point with me, claiming that slots were a better bet than tables, because his casinos slots were holding 8 percent while the tables were 20 percent. I’m not sure I ever convinced him that they were completely different statistics that couldn’t be used as a basis for comparison.

A. He could just be on a streak, but the game has changed, too. When game inventor Derek Webb sold Three Card Poker to Shuffle Master in 1999, Shuffle Master started developing alternative pay tables for the Pair Plus portion of the game. That’s normal business for Shuffle Master, which offers casino operators choices in pay tables on games such as Let It Ride and Caribbean Stud Poker.

Pair Plus is a bet that your three-card hand will include a pair or better. In Webb’s original game, payoffs were 40-1 on straight flush, 30-1 on three of a kind, 6-1 on a straight, 4-1 on a flush, and even money on any pair. The house edge was 2.32 percent, relatively low for a newer table game.

That pay table remained in play at most casinos for quite a while after the sale, but in recent years it’s been replaced by a change to 3-1 payoffs on flushes. With that one change, the house edge soars to 7.28 percent. Other pay tables are available, with none as player-friendly as the original, but the 40-30-6-3-1 schedule is the most common. Your brother-in-law’s troubles could result from that change.