Glossary of Poker Terms

Back to main poker terminology page

Terms starting with V

Value
There are many potential reasons to bet or raise (e.g., to get people to fold, to manipulate the size of the pot, to express anger, to impress someone watching from the rail, etc.). Betting for value is one of the better ones. Value means the return you get on your investment; the expected increase in your equity in the pot (your return), as compared to the size of your bet or raise (your investment). Typically this means either that you believe you will receive action from inferior hands, or that the the chance you will win the hand makes the bet worthwhile.

Variance
If you have a sufficient advantage at the game you're playing, you expect to make money over the long haul. This is true whether the game is poker, blackjack, or craps, and whether your advantage is due to skill, cheating, or psychic powers. However, over a small period of time, you may do better or worse than what your average should be. For example, you may expect to make one big bet per hour at the poker table, but in a given hour it may not be uncommon for you to win or lose twenty big bets. Variance is the statistical measure of dispersion, or just how widely your results will be distributed. When variance is high enough, a small advantage may be of no use during your lifetime. When variance is low enough, a small sample will be much more likely to reflect your real advantage (or disadvantage). In other words, variance describes just how long the long haul is. In poker terms, high variance means that a small number of hands will not be very representative of your long-term expectation.
Here's a simple non-poker example. A slot machine that pays you $1 every time you put two quarters into it (or vice versa) has no variance whatsoever. Your expected win (or loss) is $.50 per spin, and you get exactly that every spin. On the other hand, a slot machine that takes the same two quarters (or $1) and usually just eats them but one time in ten thousand spits back $10,000 (or 5,000) will have identical expectation. If you play enough games, you will tend to average the same $.50 per spin profit (or loss). But because you need so many more spins to get a representative sample of the possible outcomes, your variance is very large.

Variance is such a strong contributor to poker results that it often obscures the importance of good play. The best player at the table may start with the best cards and still have far less than a 50 percent chance of winning the hand. A skilled professional can lose money over days or weeks, without necessarily doing anything wrong. And while bad play may have negative expectation, it is often rewarded in the short term - players who draw for incredible longshots do sometimes get lucky, despite their poor judgment. Variance is what makes losing players think they have a chance in the long run, and what gives them a real chance in the short run.

Situations in poker may be higher or lower in variance. For example, in a situation where you know it will cost you a few bets to draw for a real longshot, but the pot is large enough to justify the calls, your expectation may be positive while your variance will be much higher than you'd like. This sort of situation is typical of high-variance bets - high potential payoff with a small probability of winning. As well, different qualities of the other players at the table can contribute to your overall variance at a given table. If many of the players are maniacs, willing to cap the betting and see the flop with any two cards, your variance may be high at that table. On the other hand, exceptionally weak and passive players, who may fold a high proportion of the time when they are raised, will reduce your variance. Obviously the variance you experience in your play will be affected not just by the nature of the game, but also by your style of play and by the styles of those you play with.

Although in the context of poker it's often used loosely, "variance" is a statistical term with a precise definition. Given accurate estimates of your variance and expectation (along with some assumptions about the distribution of your outcomes), it is easy to calculate confidence intervals, or ranges, within which your results are most likely to fall over different periods of time. If all this is news to you, pick up a book. Learning a little about statistics wouldn't kill you, especially if you want to play poker seriously.

Back to main poker terminology page