A Conversation with Bob Wilson,
the Dean of Poker Software

Published in 2 issues of Poker Digest

Wilson Software has been the acknowledged leader in the poker software industry since its inception. Comments from top players about the company's products include:

"I highly recommend this software." - Dr. Max Stem, World Series of Poker title holder

'There's no better software on the market at any price." - Lou Krieger, poker author and columnist

"These poker programs are the most powerful ever created." - Mike Caro, poker author and columnist

In this exclusive interview, writer Larry W. Phillips talks with Bob Wilson, the dean of poker software, about a variety of topics, including computer simulations, general playing strategy, the ups and downs of Omaha high-low split, and the future of poker software.

Larry Phillips: How did you get started in this line of work?

Bob Wilson: Oh, I've been working with computers since the mid-'60s as a programmer.

LP - Were you ever a player?

BW - I've played poker since the 1960s ... hold'em and Omaha mostly, both high and high-low, but I was never a high-limit player. Then I started to tie the two together - computers and poker.

LP - What year did you start your company?

BW - I started Wilson Software in 1989. Turbo Texas Hold'em was our first product and is still our most popular program. There were five DOS versions; soon we will release the fourth Windows version. That's eight major upgrades in 11 years. We also offer a tournament hold'em program, plus high and high-low split Turbo programs for both stud and Omaha.

LP - How many simulations have you done?

BW - Ten billion.

LP - Billion? With a "B"?

BW - Yes. There are various decision points that occur in a hand different bet actions that occur on the different streets - so we do simulations for each one. It often takes 10 million simulations to isolate the correct results.

LP - You go deeper into the variables that occur at these decision points?

BW - Yes. I've got a specialized version of the program that isolates them, so I can check the effectiveness in every situation, test all the different combinations.

LP - Certain pivotal spots in a hand pop out that you focus on?

BW - Well, actually we try to do them all. There are separate decision points for the flop, the turn, and the river - and for first position, middle position, and late position. Every combination is tested, and the specific dollar results are isolated in the test version.
An example might be the willingness to call two bets but not cold-call two bets with a hand at a certain point in the game. That sort of thing. We test until we find the most profitable outcomes, then we lay judgment on top of that. We ask: "is this realistic? Would this work in a real game? Or would opponents pick up on that?" So it has to pass the reasonability test, too, We have people who give us input - who look at these situations in a real-world sense. One is a friend and player, Mike Gilbert, who has been advising me.

LP - He's the guy who says, "Here's a point in the game we should look at?"

BW - Well, he looks at the whole thing, as a player, then gives us his advice on how the data applies to real-world situations.

LP - Your latest software seems quite "user friendly" - the ability to fold quickly, zip to the end, start new hands immediately; player stats, colored 3-D charts, read-outs on play; a nice drag-and-drop feature to obtain different player profiles in the games; and an icon that pops up to warn you when making the wrong play.

BW - That's an example of the advancing software.

LP - There are some people who are critics of poker software.

BW - Sure. It's not the "magic bullet" that will solve every poker problem. It needs to be used in combination with other types of experience - such as live play, discussing hands with other players, reading books, and so on.
Poker books and poker software are not competing products; each complements the other. Our programs expand on the strategy information available in poker books. In addition to strategy, the better books cover theory and include expert analysis. We regularly recommend them to our customers. However, books are limited in the detail they can cover. Our Turbo programs give players the opportunity to specify unlimited situations and test for the best play in each case. They also give advice and evaluate your play. Even if you don't always agree with the advice, it can help you think about situations. Last, but not least, the programs let you practice and improve your skills without risking real money.

LP - Is there such a thing as a best way to use poker software?

BW - Well, I think that isolating certain situations in your game is a good way to go: using the "stack the deck" feature, for instance, to work on specific situations in your game where you're weak, or setting the program to play short-handed for practice at that. Play around with the program, read the advice, and think about what's going on in different situations.

LP - It seems to be important to play poker software as if you were playing in a live game - to take it seriously, connect it in your mind to real play.

BW - Yes, that's true. Some just try to beat the game or beat the software, rather than using it as a learning tool for a live game,

LP - Suppose I could consistently beat the software but not a live game. Isn't there a gap of some kind there? I mean, if I'm doing everything statistically right, playing an optimum game?

BW - I'd answer that by saying this software can really help you. But it's not everything. There are more aspects to a live game. We've never said this was the end-all. You can, if you work at it, find weaknesses in the other players in a live game that you can exploit. So if what you really want to do is learn to beat a computer game, that's not the purpose. But you can use it to improve your hand selection, your timing for betting and raising.

LP - And then carry it over.

BW - And then carry it over. You really ought to be asking (in a live game), "Well, now who am I facing here?" And if there are some rocks, then while this hand might be playable on average, it might not be against these guys. Conversely, if there are some real loose players, you have to adjust the other way.

LP - I got the distinct feeling while I was playing your Omaha high-low software, observing the point values of the hands and the computer advice, that hardly anyone is playing this game correctly. I mean, playing all the right cards from the starting hands on, making all the right decisions on all the streets. Is that a fair statement?

BW - Absolutely. There's very poor play in this game, almost universally. Hardly anyone plays this game correctly. Many hold'em players, in particular, don't have a clue about how to play Omaha.

LP - Ray Zee, writing about Omaha in his book High-Low-Split Poker, states: "If you are simply the tightest player both before the flop and on the flop, you have a significant edge," and "if you just play tight before the flop, you will beat even the toughest games."

BW - I agree. It's not borne out statistically to play a lot of hands in Omaha high-low. The lesser hands simply don't make a profit over the long run.

LP -Do players and friends argue with you about certain computer results and card situations?

BW - (Laughs) Oh, sure.

LP - What do they most seem to disagree with you about? Based on your simulations, what kinds of things do you have the most trouble convincing them are true?

BW - Three of them come to mind, common misconceptions. First, that in Omaha high-low, "any four cards will do" - that you can play any four cards. Second, that it's a river game, with disproportionate events occurring on the river, which isn't true. Our simulations show that in Omaha, reasonably strong hands on the flop stand up well by the river. Players don't seem to realize that.

LP - So they don't bet and raise when they should, because they're afraid of the river monster. Emotions get magnified on the river so that a certain play sticks in memory more than it should and distorts the play.

BW - Yes, that's exactly it. Overall, the good hands do stand up.

LP -How about premium hands before the flop in Omaha high-low? How far ahead of anybody are you with a good hand before the flop?

BW - I would turn that around and say the stronger the point count of the hand, the more money that hand's going to win over time. And it's going to win in two ways. First, it's going to win more pots; second, it's going to win more money. I mean, it's nothing magic. You've simply got a set of cards, and each set is either going to be junk, or you have a high hand, or a low, or you're going to have a pair, or maybe a straight possibility or flush possibility, and sometimes multiple combinations. You add all these things up, and your hand's worth X points.
There's a good, straight-line correlation in the stats report of the software. The numbers are negative, then they go to break-even, and then they get profitable - 30 points, 40, 45, and so on as you go up. It's a pretty good approximation of the profit potential of each of those hands.

LP -Here's a common Omaha high-low situation. You sit down to play in a game starting out with the determination that you're going to play tight. So you wait about an hour for a premium hand. You finally get one and bet it out, but it misses the flop and you fold. Then you wait another hour, get another premium hand, and you lose on that one, too. Then you go on tilt, lower your standards, decide to play any four cards, start hitting the flop over and over, and run your stack up about $500.
Is there any statistical explanation as to why you can play premium cards and miss the flop, then switch to playing garbage, hit the flop, and build yourself a big stack? I mention this only because I've done it myself any number of times. And if it's not you doing it, then it's someone else at the table. Every Omaha player knows what I mean by this.

BW -Well, if you're lucky enough to do that ... that's just pure luck hitting that flop. The truth of the matter is that the better hands, the higher point-count hands, are going to like many more flops. There's just no escaping that fact. With a good hand, what you're saying is, "If I've got all these different things working for me, if I've got a hand that's worth 40 points, it's going to hit so many more flops than a hand that's worth 20 points. It just is."

LP -So you have to stick with playing the quality hands.

BW -Yes.

LP -It sure seems like positive reinforcement rolls in like the tide for playing wrong in that game, while you keep getting your wrists slapped for playing right. it's probably why the game makes people crazy.

BW -We say, for a loose game, including the blinds, that you can see the flop 35 percent of the time and still make money. Thirty-five percent - that's not bad.

LP -At 35 percent, you're getting to play.

BW - Yes, much more than you can play in hold'em.

LP -When I'm playing Omaha high-low, I often see the situation where a player has, say, A-2-3-K double-suited and is against two bad players who are playing any four random cards, and the two bad players run him down. I know that long-term the good hand is the one that will show a profit. But it just seems like ...

BW -Well, I haven't done that test, but let's do a simulation. Let's give a guy that hand.

LP -A-2 of spades and K-3 of diamonds.

BW - We'll make it a loose lineup. That means this guy will be up against -

LP - Flaming idiots.

BW - (Laughs) And we'll get several kinds of numbers for this guy.

LP -Let me ask the question in a different way. In a poker room, you sometimes hear somebody say, "Before the flop, no hand has any great advantage over any other hand."

BW - That's totally inaccurate. OK, now here's the hand we were just talking about. I played 10,000 hands. This is a $20-$40 game. The player holding the A-2-3-K is winning - he won $987,000. In 10,000 hands, he's winning close to $100 a hand.

LP - So, in other words, when you see the two bad players run him down in a real game, they're just lucking out.

BW - What you're dealing with is the phenomenon of short-term luck. Now this is a loose game. We have five or six players seeing the flop; it even got as high as eight players. And how many pots does this guy win? He's taking away all the money, of course, but ... ah, here it is. "Pots Won": 3,687 pots out of that 10,000 hands. So it's very misleading because of the luck swings. Here's another stat - very interesting - with that hand, which is a power hand. Right?

LP - Right.

BW - He lost 11 times in a row with that hand, at one point.

LP - He did?

BW - Eleven in a row. That was his longest losing streak. Now the guy in seat No. 6, that poor guy lost 127 in a row.

LP - (Laughs) Serves him right.

BW - So in terms of bad luck, well, the other guys are running 71 losses, 88 losses, 34 losses ... in just a reasonably short time.

LP -OK, so when this guy loses his 11 hands in a row, everybody in the poker room chats it up, emphasizes it, makes a mental note of it, blows it out of proportion in their minds as to how often it actually statistically happens.

BW -That's right, because 49 percent of the time, the best hand on the flop won the pot. So it's amazing, as I mentioned earlier, that a lot of people say: "It's a river game. ... My gosh, I get rivered all the time. ... it doesn't matter what four cards you play." The folklore is wrong.

LP -What they're really saying is, "it hurts a lot," and they just want to share it with everybody.

BW -(Laughs) That's right. The third thing, another misconception, concerns raising to thin the field, which is often wrong.

LP -You don't want to thin the field?

BW -No.

LP -You want them in? if it does thin the field, then raising is bad?

BW -Sklansky and Mammoth will tell you - and a lot of others as well - that you bet to either increase the pot size or thin the field. It's just my opinion that if raising with a quality hand does not thin the field, then I don't think that's a disadvantage. What you're saying is, "I can play for higher stakes on the hands that I have an advantage in."

LP - There's also the argument that goes: When you thin the field, you often just get out the bad players and leave the good players in.

BW - That's true, You can test the effectiveness by doing it both ways (on the computer). We really recommend, in the early positions, not raising.

LP - In Omaha high-low.

BW - Yes. In the first three positions. You might reraise after it comes back to you, of course.

LP - You might drive people out, otherwise, when you have a premium hand, and you'll end up two-handed or something.

BW -Yes, that's correct.

LP - How much worse a hand can you have in the little blind in Omaha? You know, when you're already in for half a bet. Is there any way to generalize about that?

BW -Yes, there is. In the point count for an average game (in the software), you need 32 points to call. But if you're in the small blind, you need only 18. So it's close to counting for half.

LP - So you can have a significantly weaker hand?

BW - Yes. But after that, it would depend on the specific hand.

LP - T.J. Cloutier, in his book Championship Omaha, seems to favor limping in with premium hands in Omaha high-low, rather than raising with them. This has the effect of leaving a lot of callers in, and then he makes his profit on the later, double-betting rounds. This also has the effect of not tipping off the strength of your hand. What does the computer say about this approach?

BW - Well, it depends on the opposition. As we were saying earlier, a lot of players don't fold when you bet (or raise) a good hand. This is especially true at the lower limits. People don't fold up shop; they don't, get out.

LP -So you should raise if this is the case, as opposed to limping in.

BW -You get more family pots, because people like to see some money come in - the gambling people. Now, if it's a tight game, where people are actively dropping-out because of betting, that's a different thing. Our testing shows that if you're not driving the players out, dramatically, it's more profitable to bet (or raise) with these good hands. But if it's having that effect, then you have to adjust.

LP - So it depends on the game.

BW -And that's important. There's this folklore that nobody should be raising in Omaha games. People will say: "What's wrong with this guy? What's he doing raising in this game? Everybody knows you shouldn't be raising." But you're giving away money if you don't. In fact, if you raise and it's not thinning the field, that's good.

LP -Omaha players get used to adverse comments like this - as well as calamitous occurrences. What sets the Omaha player apart is that he actually gets used to abuse.

BW -(Laughs) Oh, I agree.

LP -In Cloutier's book, he doesn't seem to look too favorably on the Omaha high-low hand, 2-3-4-5. What does the computer say about that hand? And about 2-3-4-type starting hands in general? Why does it rate these hands so highly? Don't they have to flop an ace?

BW - Well, let's set up a test, a simulation. I'm going to do it all offsuit. OK? I don't dislike the hand myself. (A minute later.) It shows the 2-3-4-5 winning pretty steadily.

LP -The average player is often afraid of this hand, thinking he has to have an ace flop.

BW. (Still running the test on 2-3-4-5.) It's showing a fair amount of scoops.

LP -Yeah?

BW. I've got it in a loose game - six players seeing the flop.

LP: But it's making money?

BW - (Surprised) Twenty dollars a hand - in a $10-$20 game. It's winning a big bet per hand.

LP -I think Cloutier was talking about this hand in a tournament context.

BW.- it seems to depend on how often players want to see the flop. It's winning a bunch now. Let me tighten up the game. (A minute later.) Oh, now it's losing.

LP - The 2-3-4-5?

BW. Yup. In a tighter game, where only three or four players see the flop, it's losing $6 a hand.

LPSo if you're playing with tight players, it's not as good a hand. The aces are probably going to be in use in a tight game - the ones you need to flop.

BW. Right. Several of our tests have shown dramatic differences for some hands as the number of players who see the flop goes up. And this has a bearing on one of the more common questions we get - on the question of how to use the software. You need to model your play on the type of game you play in. This is very important. How you should play in a game that's got three players seeing the flop is going to be quite a bit different from one where you've got six players seeing the flop, Hands that will work in one game won't work in another. If you're not aware of what kind of game you're playing in, there's a problem right there. You should be practicing in that context and adjusting to it. Also, you want to stay away from really tight games, because you're not going to beat those games.

LP: You need to be aware of what happens to the table as a whole, how they are reacting as a group.

BW.- Sure.

LP: I'm always surprised by certain Omaha high-low hands that are rated highly by the computer. An example would be a hand Re As5s6c8h. This hand in particular has always mystified me. It's rated as a betting or raising hand, yet doesn't seem to have that much going for it: a fairly weak high, single-suited, an A-5 draw for low. But it always seems to register high on the point value chart.

BW - I think it's a pretty strong hand. Let me do it quickly on the computer. (A minute or two later.) I just ran a test. The number of players who see the flop is the biggest thing. in a multiway game, with six players seeing the flop, you've got straight possibilities, plus several low draws.

LP - Most of the premium hands in Omaha high-low are well- known, of course: A-2-K-K, that sort of thing. But here the logic of the hand isn't too apparent.

BW - The hand is not overpowering, for sure. The flush draw adds to it considerably. It seems to pick up a lot from that and gets paid off well when it hits. I don't know how often the guy makes the hand and then loses with it. But when he does win ...

LP -He gets paid off.

BW - He gets more money, The flush is making a good part of the money. Or he's tripping up, or hitting the junk match-ups that he sometimes would hit, and he's also got several ways working on the low.

LP -I got the impression that high hands had been downgraded slightly in your newest Omaha high-low software - hands like K-Q-J-J unsuited, K-J-J-10, Q-J-Q-10, A-Q-Q-10.

BW -A better way to put it is that aces have been upgraded. That would be a more accurate characterization - their value has gone up. Testing has shown the dramatic value of the ace;

LP - I always thought hands with four high cards were fairly good.

BW - If they're all unsuited, it wouldn't be one of my favorite hands.

LP - What would be your opinion of two-pair hands in Omaha high-low, such as K-K-8-8?

BW - Other than if both pair are suited, terrible.

LP -I've sometimes wished there was a "how have my cards been running?" feature in the stats section of poker software. A lot of times when you're losing, you're not sure whether it's you or simply, the fall of the cards. is there some variation of this in the software?

BW -The short answer is, there is not. Have you used the "challenge" feature in the soft-ware?

LP -The part with the computer opponent called "Sherlock Holmes"? Where you play 200 hands against a table of opponents, then he plays the same 200 hands against the same table, and you compare scores?

BW - I think that's valuable.

LP - Actually, that touches on what I was asking - about a feature for how my cards have been running. Because you know they were running the same for him.

BW -Right.

LP - So if you lose $300 and he plays the same hands and loses $380, it gives you some idea of where you are. Both of you were getting the same cards.

BW -Right. We're hoping to eventually have a capability where you can go back and it'll say, "Let's look at each hand,' and give you a commentary on what you did and why the computer thought it was wrong.

LP -More extensive explanations.

BW -Yes.

LP -What's the future? What do you see ahead - the next breakthrough or line of development in poker software? Computer opponents who are smart enough to adjust to our play?

BW -Oh, there's a question. Yes, I think they (computer opponents) will get progressively better. But more feedback, mainly, in terms of reports and information on your play, Also, I think it'll go in the direction of multiple choices, where you have a choice to fold or raise, multiple scenarios.

LP - Eventually it gets to be subjective at certain points in the game, where there's no absolute bottom he. Maybe that's the challenge of it.

BW -Sure. But I think it'll also go in the direction of a more detailed review of a player's decisions.

LP -How about this as an idea? Is there any way you could put a small electric shock into the software, a buzz that shoots up through the keyboard Something that when players start chasing a hand or playing bad cards it gives them a jolt? Maybe if we got zapped every time we made a bad play, it would shorten the learning curve dramatically.

BW - (Laughs)

LP -Here's a final question, one that my friend Gilby wanted me to ask. When are you going to invent a poker software that'll keep us away from the craps table? I told him, 'You're probably on your own, Gilby.'

BW - (Laughs) That's right - that's the answer. You're on your own, Gilby.

Editor's Note: Larry W Phillips is the author of Zen and the Art of Poker, recently published by Dutton/Plume Publishers and is available in most bookstores.