Among all the skills needed to get the most out of casino games online and offline, one of the most important for casino players is choosing your wagers wisely.

And an important part of choosing your wagers wisely is knowing which bets never to make.

Sometimes that means avoiding bets with extraordinarily high house edges. Sometimes it means skipping a bet because a better option is available that accomplishes the same thing with a lower house edge.

Here are a big four among common casino wagers you should NEVER make:


Only two hands are dealt in baccarat: a player hand and a banker hand. 

Everyone can bet on either hand. Banker wins more often than player, so those who win banker bets must play the house a 5 percent commission -- the payoff is even money minus that 5 percent. Winning player hands are paid at even money.

Both are among the better bets around, with a 1.06 percent house edge on banker and 1.24 percent on player.

There's a third option. You can bet that the two hands will tie.

Winning bets on ties are paid at 8-1 odds. That's tempting for those looking to make a fast profit instead of grinding it out with even-money payoffs.

But the true odds against winning a tie bet are 9.526-1, not 8-1, and the house edge is a whopping 14.4 percent.

When you have two available bets with house edges just a little over 1 percent, why would you make a bet with a 14.4-percent edge? No reason at all.


There are many bets to avoid in craps, especially in the one-roll propositions. House edges of 16.67 percent on any 7, 11.11 percent on any craps and others are too high to overcome.

But there are a couple of multi-roll bets with a sneaky high house edge.

Craps table layout

In two corners of the layout, you'll see a big red 6 next to a big red 8. Those are Big 6 and Big 8.

They play exactly like place bets on 6 and 8. If you bet on Big 6 and the shooter rolls a 6 before rolling a 7, you win. If a 7 comes first, you lose. No other numbers matter -- if the shooter rolls a sequence such as 5, 9, 3 12, 4, 5, there is no decision. Your bet either stays in action or you can take it down.

The difference is that place bets on 6 and 8 pay 7-6 odds, while Big 6 and Big 8 pay even money.

That difference in payoffs lead to a dramatic gap in the house edge. The house has a 1.52 percent edge on place bets on 6 and 8, but 9.09 percent on Big 6 and Big 8.

Because of the 7-6 payoffs, you want to bet in multiples of $6 when placing 6 or 8. There is no such imperative on Big 6 and Big 8. But you actually average more in losses when betting $5 on Big 6 or 8 than when placing 6 or 8.

Here's how that works:

Let's say I place $6 on 8 for each of 36 spins in which each combination of two dice comes up once. You bet $5 on Big 8 on the same 36 spins.

On 25 of the spins, the roll is neither a loser 7 or a winner 8, and we just keep our money. 

We can focus on the 11 wagers on which bets are decided. On those, I have $66 at risk on my place bets, while you're risking $55 on Big 8.

On each of the five winners, I get my $6 back and get $7 in winnings. That means at the end of the trial I have $65 of my original $66.

On each winner, you get your $5 bet back and get $5 in winnings. Your total is $50, while the house has kept $5 of your original $55.

I've bet more money than you, and we've won and lost on the same rolls, but you've lost five times as much money as me.

Never bet on Big 6 or Big 8. If you want to bet those numbers, use place bets instead.


This is not a factor if you're playing roulette with a single-zero wheel. The five-number bet doesn't exist.

But on the double-zero wheel used by nearly all American casinos and which also is available in many online or international casinos, you can bet on 0, 00, 1, 2 and 3 all at once.

The payoff is 6-1, but that's no bargain. The house edge of 7.89 percent is one and a half times the house edge of 5.26 percent of available bets.

You can bet on the same five numbers and get the lower house edge on many different ways. 

If you bet $5 on the five-number bet and I bet $1 each on 0, 00, 1, 2 and 3, we each risk the same amount per spin.

In 38 spins in which each number comes up once, we each bet a total of $190.

Roulette table layout

On each of the five winning spins -- when the ball lands in 0, 00, 1, 2 or 3 -- you get your $5 bet back and get $30 in winnings. At the end of the trial, you have $175 of your original $190.

My single numbers win five times and are paid at 35-1 odds. On each winner, I get the $1 back that I bet on that number, and get $35 in winnings. When all is done, my total is $180 of my original $190.

We've bet the same amount on the same numbers, but at the end of the trial I have $5 more than you do.

You can accomplish the same thing in different ways. You could bet $3 on the three-number street 1, 2 and 3 and a $2 split on 0 and 00. You could bet a $2 split on 1 and 2, a $1 single number on 3 and a $2 split on 0 and 00.

However you want to do it, the house edge on your combination will be 5.26 percent as long as you avoid the five-number bet. That's one of those bets you never want to make.


This one needs a qualifier: You should never take insurance in blackjack unless are a card counter. For card counters who know there is a higher than usual concentration of 10-value cards remaining in the deck, insurance is a viable option. For the vast majority of players, it's a bet to avoid.

The basics, when the blackjack dealer has an Ace face up, he asks players if they want to take insurance.

Taking insurance requires making a wager half the size of your original bet. If you have a $10 bet, you'd then bet $5 on insurance. 

Insurance pays off if the dealer has a 10-value card face down to complete a blackjack. That yields a 2-1 payoff on blackjacks.

If a third of the cards were 10-values, insurance would be an even bet. Then you'd win on insurance an average of once per three hands. If you bet $5 on insurance three times for a total of $15. You'd lose twice, but the one time you won, you'd keep the $5 insurance bet and collect $10 in winnings - the $15 on the one winner would balance your $15 total risk.

However, only 30.8 percent of cards are 10 values, so you lose insurance bets 69.2 percent of the time and the house has a 7.6 percent edge.

When you have blackjack, a form of insurance is offered called "even money." Instead of playing the handout and risking the hand pushing if the dealer has a 10 value down, you tell the dealer "even money." You give up the chance at a 3-2 payoff if your blackjack wins, but you are guaranteed to win every time.

It's a bad tradeoff for the player.

In an average 1,000 trials in which you're betting $10 a hand and have blackjack, you could call even money and guarantee $10,000 in winnings.

However, if you take your chances and play out the hand, you win an average of 692 times. Multiply by 1.5 for the 3-2 payoff, and you get 1,038, and multiply that by your $10 in wagers, and you bet $10,380 in winnings.

Pushes may be frustrating, but by risking the pushes and declining insurance, you win more money.

Unless you're counting cards and know there's an excess of 10-values, NEVER take insurance, including its even-money form.

For nearly 25 years, John Grochowski has been one of the most prolific gaming writers in the United States. He’s been ranked ninth by GamblingSites among the top 11 gambling experts at Gambling Sites and his Video Poker Answer Book was ranked eighth among the best gambling books of all time.