Rare strategy quirk in a Deuces Wild practice session

Practice may not make perfect in video poker, but it can spring some surprises on you.

I got a surprise of my own recently as I was practicing my strategy on Not So Ugly Deuces Wild. It's a game

I hadn't played in some time, but I was planning a day at a casino that offered it. With expert play, NSUD pays 99.7 percent with expert play. That figured to be the best game I'd find on that trip, so I figured I'd better put in a little practice time on the WinPoker software I use.

Here's the hand I was dealt: Queen of clubs, 9 of diamonds, 8 of diamonds, 4 of spades, 3 of hearts.

High pairs don't pay off in Deuces Wild --- the pay table starts at three of a kind --- so I wasn't going to hold the low Queen. Straights in Deuces Wild pay only 2-for-1, so you need four cards before you start thinking about possible straight draws. The best straight possibilities here were only two-card sequences.

Flushes pay 3-for-1 in NSUD, better than the 2-for-1 in full-pay Deuces Wild, so we do look for flushes often. Still, there were only two cards of the same suit in this hand.

Straight flushes pay 10-for-1, another step up from full-pay Deuces, which pays 9-for-1. But two cards to a straight flush? Not likely.

My conclusion: Toss the entire hand. Take a chance on five fresh cards.

The software's conclusion: A pop-up box, warning me I was making a mistake.
I changed my play to holding the 8-9 of diamonds, the only feasible play I could see here. At least it would give me starts on possible flushes, straights, and a long shot at a straight flush.

That, the computer accepted. It played out the draw, and then I clicked on the "analyze any hand" option to check out the numbers.

Sure enough, the calculations told me that holding the 8-9 of diamonds would bring an average return of 1.6075 coins per five wagered, while tossing the entire hand would bring only 1.6074 coins. I was wrong by one ten-thousandths of a percent.

Not a make-or-break hand obviously, and if you're playing in a casino and decide to toss the entire hand, well, I won't quibble. In order for holding the suited 8-9 to be the correct play, all the circumstances had to be in place.

To start with, the NSUD pay table had to be in place. In full-pay Deuces Wild, which pays less on flushes and straight flushes than the Not So Ugly variety, the best play is to discard all five cards. Even in Illinois Deuces, which matches NSUD in paying 3-for-1 on flushes but retains the full-pay return of 9-for-1 on straight flushes, the expert play is to toss the lot. The same hand in Illinois Deuces returns an average of 1.6012 coins with a five-coin discard, but only 1.5828 when holding the suited 8-9.

Beyond that, the situation regarding other possible straights and flushes had to be the same. Remember the hand: Queen of clubs, 8-9 of diamonds, 4 of spades, 3 of hearts
.
If holding the two consecutive diamonds with a straight flush possibility meant tossing a third diamond, the percentages would shift. If the 3 of hearts was a diamond instead, the best play would be to throw away the entire hand. Three diamonds with no straight flush possibility wouldn't yield enough to hold them, and throwing away a third diamond would diminish flush possibilities enough that holding the 8-9 would no longer be worthwhile.

Same deal with straight possibilities. The only possible straight involving three cards in the original hand is 8-9-Queen. There are two gaps on the inside, so the only combinations that can complete the straight are 10-Jack, 10-2, Jack-2 or two wild deuces.

What if the Queen of clubs was a Jack instead? Then there would be only one gap, and the combinations that would result in a straight would increase to 7-10, 10-Queen, 7-2, Queen-2, or 2-2. Throwing away the Jack would decrease the chances of building 8-9 into a straight that here too, the best play would change to discarding all five cards.

It seems by random chance in practicing with the WinPoker software, I ran into the right hand on the right pay table to learn a little something. If I'm playing with the Not So Ugly Deuces Wild pay table, and if I'm dealt a hand with 8-9 suited, no other cards of the same suit, and no straight possibilities with less than two gaps, I'll be holding the 8-9.

That's a rare situation, and who knows when I'll run into it again. But it's a play I'll never forget, and one I'd never have noticed had I not taken the time for a little practice.

Do casinos make their profit from winning players?

Q. Sitting in the buffet, I caught a snippet of conversation that I was wondering if you could explain. (I wasn't listening in, they were so loud I couldn't help hearing.) One guy was saying that the casino really makes its money off the winners, and the other guy said something like, "Oh really? You mean we pay for all this when we win?" That doesn't really make sense to me.

A. Kind of leaves you wondering what kind of gambling palaces they could build if everybody won, doesn't it? If the casino makes money off the winners, then more winners must mean more profits, right?

But seriously, there is a way of looking at how the casino makes its money that looks at casino profits as a tax on the winners. Casino games make money because they pay the winners at less than true odds. If 38 people are sitting at a double-zero roulette table and each bet $1 on a different number on a single spin, the 37 losers each will lose their buck, and the one winner will be paid at 35-1 odds and walk away with $36. If the casino was paying true odds, the one winner should be paid at 37-1 odds and walk away with $38. The casino profit is the $2 not paid to the winner that he'd get if true odds were paid.

Same deal with sports betting. In most sports books, you have to put down 10 percent vigorish on top of your bet. Let's say you and I are betting on the same football game, with me betting on Team A and you on Team B. We each intend to bet $100, but we have to pay the vig, so we actually each bet $110. When my team wins --- hey, it's my example; I get to win --- you lose your $110 bet, but I'm paid only $100, along with the return of my $110 wager. The casino profit is the $10 it didn't pay me on my winning bet.

Sometimes the paying of winners at less than true odds is disguised a bit. In baccarat, for instance, bets on banker win more often than they lose, and bets seem to be paid at even money. However, bettors have to pay a 5 percent commission on winning bets, so winners aren't really paid at 1-1; they're paid at (1 minus .05)-1, and that's less than the true odds of winning the wager.

So it goes with every casino game. There are going to be winners, and there are going to be losers, but the house will make money because it pays winners less than the true odds of winning the bet.

My roulette system? Play for fun

Over the years, I've debunked a fair number of roulette systems.

There's the Martingale, where you double your bet after each loss. Hazardous to your bankroll. Bets get very large, very fast. Even if you can afford the action, you'll will bump up against table maximums often enough to sink the system.

There's the cancellation, where you start by writing three numbers, the sum equaling your win goal for the sequence. In a three-number cancellation with a win goal of three units, you'd write 1, 1, 1. Your first bet is the sum of the end numbers, so you start with a two-unit bet. A win cancels the end number --- that's the cancellation part --- so your next wager ins the remaining bet. Win that, cancel out the number, and you've won your goal.

What if you lose? You then write your wager in units at the end of your number sequence. Start with 1, 1, 1, wager two units and lose, and you're left with 1, 1, 1, 2. Your next wager is three units --- the sum of the two end numbers.

The problem? You're still increasing wagers as you lose. It's not as dangerous a system as the Martingale, but a few losses in a row stretches out the sequence, and it's not unusual for your row of numbers to get so long it'd take an unusual hot streak to get you back to even.

One close friend likes to extend play at roulette by betting on black plus an equal amount on the third column. Eight of the numbers in the third column are red, so my friend covers 26 of the 38 numbers --- all 18 black numbers plus eight red. If a black in the first two columns shows up, he breaks even, winning on black but losing the column bet. If one of his eight reds turns up, he makes a one-unit profit, winning a 2-1 payoff on the column while losing on black. But if the ball lands on one of the four third-column blacks, he wins triple --- even money on black, plus 2-1 column.

Of course, there's the pesky matter of the 10 red numbers he doesn't cover, plus 0 and 00. Any of those dozen numbers bring a double loss.

Undermining all roulette systems is that the house edge never rests. Whether you're betting red or black, dozens or columns, single numbers or three-number streets, that 5.26 percent house edge is always working against you. The exception is if your system includes the five-number bet on 0, 00, 1, 2 and 3. Then a little higher house edge is working against you, since the house has a 7.89 percent advantage on that bet.

Still, few players just throw their chips out willy-nilly when they play roulette. Nearly everyone uses a system of some sort, whether it's in how they choose their numbers, or in how they raise or lower their bets.

My friend who likes the black-plus-third-column system pressed me for my system. "Come on," he said.

"Everybody has a system. What do you do?"

And I do have a system. A really easy one. I play family birthdays. So does my wife. When we play together, we're usually on different numbers. I play her birthday, and she plays mine.

Given a balanced wheel, no number is any more likely than any other to come up, so why not have some fun? It might even produce a memorable moment.

It did for me one day in Las Vegas, playing at a wheel with a $1 minimum bet, and with 25-cent chips available. It was my birthday, so I was betting on myself. I put a chip on No. 29, and surrounded it with a chip on each of the corners and splits surrounding it.

I had a nice streak. No 29s, but the two-number splits, paying 17-1, and four-number corners, paying 8-1, were hitting regularly. My stack was growing, and I increased to two chips on each bet.

I heard the ball drop, and the dealer said, "Uh-oh. Here we go."

It was No. 29.

The four corner bets each brought me 16 chips in winnings. The four splits each brought 34. The single number bet brought 70. Suddenly I had 270 extra chips in front of me, and my nine two-chip bets were still on the table.

The next number: 29. I raked in another 270 chips.

That was the end of the streak. The second 29 was followed by an 11. I exchanged my roulette chips for casino chips, tipped the dealer and left with a profit of more than $125.

When I told my systems-playing friend that story, he said, "Don't you wish you were playing with dollar chips? Too bad it was only quarters."

I told him I had no such wishes. I play roulette for relaxation, and I play only at low-limit tables. I was thrilled to walk away with such a profit.

Besides, I told him, I know what the house edge is on that system, and I don't care to bet too much money against that kind of an edge.

In roulette, the magic number for the house is almost always 5.26.

Should you quit whenever you're ahead?

Q. I had a typical day at the casino. I won a little money at the blackjack table, then I piddled it away a little at a time -- 20 bucks on a slot machine, another 20 at video poker, then I went back to blackjack and lost a little more. The bottom line is that instead of winning a little, I wound up losing about $50 for the day.

My friend and I got to talking about this, and he said that every time you play, there's a time that you're ahead of the game, and if you could just discipline yourself to quit then, you'd win. That seems right to me, but I never seem to be able to walk out ahead.

A. First off, I don't buy the notion that there's a time we're ahead every time we play. Haven't you ever had a day when you've lost your first five or six hands at blackjack, and never recovered? Or blown through a couple of $20 bills at video poker before ever seeing a hand better than two pair? I sure have.

The house edge is working us from the very beginning of our day at the casino, and there are times that we're on the downside of the ledger from the outset and never climb back to even.

On days we do get ahead, how quick do we want to be to end our day? If we win our first hand at blackjack to go ahead $5 on the day, are we going to take our profit and go home? Not likely. Even stopping after big wins will rob us of some of our biggest, most memorable days. On one overnight stay at a casino, I hit a royal flush on a quarter video poker machine for $1,000 in the morning, won a few hundred dollars at blackjack in the afternoon and wrapped it up with an $8,000 royal on a $2 machine in the evening. Should I have stopped after the morning royal and gone sight-seeing or to the movies?

Now then, on days when we get ahead by a fair amount, most of us should take steps to make sure we bring at least some of it home. This doesn't necessarily apply to pros such as blackjack card counters or video poker experts who have a mathematical edge on their games. But average, recreational players need to protect their winnings.

Putting away half of any large win is a good place to start. When I won my $1,000 royal at the start of a big day, I put away $500 to bring home right off the bat. That gave me some extra money to play with, but made sure I'd have some profit, too. How much you put away, and at what level you start your put-asides, depends on your own bankroll and goals.

Blackjack freedom of choice helps if you know how to choose

Freedom of choice is a good thing for blackjack players who know their basic strategy, but for many players, it largely means the freedom to choose poorly.

Players who take the time to learn well are best off with a game that gives them plenty of options. Being allowed to double down on any first two cards and to split and resplit pairs is to the advantage of players who know how to use the options wisely. On the other hand, I see players every week who would be better off if double downs were restricted to hard totals of 10 or 11.

Of course, such players would be better off still if they just took a short time to learn how to use the options presented. Let's take a look at rules that can be beneficial to the player, but need to be handled with care.

Double down on any first two cards: When it comes to doubling down, I've seen some truly odd plays. I once watched a fellow double down every time he started with hard 12. Doubling when it's possible to bust in one card is such an unusual--and bad--play that the dealer was required to call "Double on hard 12!" to the pit supervisor every time the play was made.

I also once played at the MGM Grand in Las Vegas at the same table as a man who doubled on every soft total--hands in which an Ace is at least temporarily being counted as 11 and which can't be busted with a one-card hit. He peeled off $100 bill after $100 bill, getting his comeuppance for doubling Ace-2 against a dealer's 9, and Ace-4 against a 7. The play that had the whole table trying to show him the error of his ways was when he doubled on Ace-Ace against a Jack, instead of the far superior play of splitting the pair. Such players would be better off if opportunities to double down were limited.

In multiple-deck games where the dealer stands on all 17s, it's to the basic strategy player's advantage to double on hard 11 unless the dealer's up card is an Ace, on hard 10 against everything except an Ace or 10-value, and on hard 9 against 3, 4, 5 or 6. If the dealer hits soft 17, double on 11 against all dealer up cards, including the Ace. No doubling on hard 12 or above, or on hard 8 or below.

As for soft hands in stand-on-all-17s games, double soft 17 or 18 if the dealer shows a 3, 4, 5 or 6, soft 15 or 16 against 4, 5 or 6, and soft 13 or 14 against 5 or 6. If the dealer hits soft 17, also double on soft 18 vs. 2, and soft 19 vs.6.

Resplitting pairs: I've occasionally been in casinos that allow only one split--if you split 8, 8 and are dealt another 8, you're stuck with 16 as the start to one hand. Most allow you to resplit pairs, so in that situation, you could have three separate hands, each starting with 8. Some allow up to three splits, giving you a total of four hands.

Players who do strange things such as splitting 5s or 10s--awful plays--are better off with a rule that stops them before they split again. But really, being allowed to resplit is to the player's advantage, provided the player knows when to split in the first place. As far as the basic strategy player is concerned, if splitting the pair is the proper play the first time, so is each potential resplit.

When should you split the pairs? Always split Aces and 8s, but never split 5s or 10-values. For everything else, it depends on the dealer's up card.

If allowed to double down after splitting pairs, split 2s or 3s if the dealer shows 2 through 7, 4s against 5s or 6s and 6s against 2 through 6.

If doubling down after splitting is not permitted, split 2s or 3s against 4 through 7, never split 4s and split 6s only against 3 through 6.

Regardless of whether doubling after splits is permitted, split 7s against 2 through 7.

The trickiest play is splitting 9s. Split (and resplit, given the opportunity) against 2 through 6 and against 8 or 9, but stand against 7, 10 or Ace.

Surrender: Surrender has become a rare option,. but when it's permitted, you can give up half your bet instead of playing out the hand. It's been about 15 years since I've seen early surrender, where you can make the play before the dealer checks for blackjack. In late surrender, you can't surrender when the dealer has blackjack -- you lose the whole bet.

I've seen players surrender 14s against 7s and 12s against 10s. I even watched one player surrender every 16, regardless of dealer up card.

If the dealer stands on all 17s, you can limit losses by surrendering hard 16 when the dealer shows a 9, 10 or Ace, and surrender hard 15 against a dealer's 10. If the dealer hits soft 17, surrender hard 15 vs. 10 or Ace, hard 16 vs. 9, 10 or Ace, and hard 17 vs. Ace.

A player who surrenders more than that could do with a little less freedom of choice.

Faulty app leads blackjack player astray

Q. Here is a question about a murky corner of the basic strategy table for blackjack: Do you hit or stand on soft hands with three or more cards?

The only blackjack book I have at hand is Henry Tamburin's "Blackjack: Take the Money and Run." Tamburin says, "As a general rule, a player should never stand on a soft hand that totals 17 or less." I read this to include hands with three or more cards as well as hands with two cards.

I recently purchased an application that allows many ways of practicing blackjack. The application is a solid piece of work that takes into account most rules variations. Plenty of knowledge and care have gone into its creation. If I follow Tamburin's dictum with a soft hand of 17 or less, I get scored with an error. The default strategy table in the application follows the rules for hard hands with all hands of three or more cards.

Example: Dealer's upcard is a 3. You are dealt a 2, a 3 and an Ace. You have a total of either 6 or 16. The way I read Tamburin indicates that the best play is a hit. The application says I should stand.

Am I misinterpreting Tamburin's rule?

A. It seems your blackjack app is not quite as solid as you would like to think. It is leading you astray on soft hands.

You are not reading Tamburin wrong. He is correct and your application is in error. It seems to be defaulting to basic strategy for hard hands after you receive a third card. That's a serious bug, since you can't bust soft hands with a one-card draw.

Look at it this way. If you stand on soft 16, how can you win? The only way is if the dealer busts. If instead you hit, what's the worst that can happen? You can't bust the hand, so the worst you wind up with is another hand that can win only if the dealer busts. You could draw a standing hand, or you could draw one that's no worse than your starting point. There is no downside to taking a hit.

There is no downside to hitting the hand, and the upside is that you could draw a low card that will turn your hand into a winner.