What can I do to calculate my odds in a hand?

Question

How do I calculate the odds of making my hand for any given game of poker?

Answer 1

The simple case, for making your hand on the next turn, is to add up the number of cards that can give you the winning hand on the next turn, and divide by the total number of unseen cards left in the deck. This means that, in games like Stud, you need to remove any potential outs that have been folded by others from the cards that can give you a winning hand.

In general, if you have seen 5 cards, and you have 4 to a flush, there are 47 unseen cards, and 9 that will give you a flush. If you think a flush is the winning hand, you have a 19.14% chance. If you are playing a flop game and you are up against a set, you have to discount the one flush card that would pair the board, which means you have 8 outs out of 47, or 17%.

So, for the "next card" solution, you have:

percent = (#outs) / (#unseen cards)

When there are multiple cards left, more advanced statistics become involved. You have to consider cards that may give your opponents the best hand, as well as conditional probabilities. In flop games with only the turn and river to go, and shared board cards, an approximation of the formula would be:

percent = ((#flopouts)/47) + 
           (1-(#flopouts)/47)*((#turnouts)/46) * 
              (1 - ((#opponentflopouts)/47) + 
                    (1-(#opponentflopouts)/47)*((#opponentturnouts)/46))

In that case, opponent outs are outs that will give your opponent a better hand than you. Depending on the hand you are drawing to, and your opponent's hand, the formula can be simpler - in the case of a flush draw against a set, you have 8 outs twice, but your opponent will have 6 cards on the turn that give him a full house, and 9 cards on the river that will do the same.

Answer 2

There is a nice tool at pokertablestats.com and another at propokertools.com to help you calculate your odds.

A good rule of thumb I always use for calculating odds is to multiply the number of outs(13) by 4 on the flop and 2 on the turn. This way it will be much easier to remember and you are never that far away from the correct percentage.

If you do not understand the term "outs", this is basically just the number of cards you need to hit to make your desired hand. If you are holding for instance 2 spades in your hole and there are 2 more spades on the flop, you are holding a so called flush draw. There are 4 spades showing now and with a total of 13 spades in the deck, this leaves 9 more spades possible to hit on the turn or river. 9 outs for hitting your flush.

If you want to use my easy to remember rule of thumb this makes it:

On the flop = number of flush outs x 4 = (13-4)x 4 = 36%

On the turn = number of flush outs x 2 = (13-4)x 2 = 18%

These numbers are somewhat higher than the absolute correct percentage though. Just remember to subtract a couple of percentages after calculating and you are never far away from the correct one :)

*pokertablestats.com tool no longer exists

Answer 3

You can use the rule of 2/4 as described above, or by thinking in reverse, do some simple math to get the odds of completing your hand by the river.

Instead of directly calculating the odds of building your hand, think of the odds of NOT getting your card -- it makes it much simpler. As an example, after the flop you have 4 to a flush. What are the odds of NOT getting your flush with 2 cards to go? There are 9 cards that will fill the flush. 47 cards remain. So on the turn, your odds of NOT hitting are 38/47, and if you are unsuccessful on the turn, then your odds on the river of not hitting the suit decrease (slightly) to 37/46. Then the probability of not hitting on either is simply the product of the two odds: 38/47 * 37/46 (and conversely, your odds of hitting are 1 - [38/47 * 37/46].

If you can do long division in your head, you would know that 38/47=~.81, and 37/46=~.80. .8 x .81 =~.65. So you have a 65% chance of NOT hitting the flush, and a 1-.65 = 35% chance of hitting it.

Since long division in your head may not be a skill you wish to develop, just use approximations -- you will be close enough to determine your proper betting behavior. Reduce the fractions to see that 38/47 and 37/46 are both around 4/5 or .8, with a shade left over (9*4 = 36, 9*5 = 45, and 36/45 is very close to the division you want, but slightly less than either 38/47 or 37/46). multiple them together you get .64. You know the real answer is a shade over 64%... which gets you very close to the actual answer without doing complex mental gymnastics.

The harder part is calculating your opponents odds. The same technique applies, but you must make informed guesses as to what he is holding. Building a profile of your opponents' hands is the essence of Texas Hold 'em decision analysis, but not germane to the calculation question.

Suffice to say that you should be very leery of playing with idiot-savants who can do Rainman-like calculations rapidly in their heads... they are going to be a step ahead in knowing the odds of beating you. You can get better and faster with practice, though!

Answer 4

It's hard to do all the exact math to get the exact percentages while you're sitting at the table and everyone is staring at you while you noodle your chances. You have to do it all in your head in less than a minute. The other methods described here are actually more accurate, but I don't think that getting the exact percentage is that important, but you need to be close.

The way I do it on the fly is:

1. figure out if you are ahead or behind in the hand - you have to be able to put your opponent on a hand or a range of hands. If can't do that, then you're always gambling and simply "playing your own hand." Then none of the percentages really mean anything.
2. If you are behind, then count the number of outs you have (the cards that will make you "ahead" in the hand) then multiply by 2. That is your percentage to improve on the turn or river.

Some basic out counts after the flop are:

• One Pair improving to two pair or trips: 5 outs twice = 10 outs = 20% or 1 in 5
• Inside Straight draw: 3 outs twice = 6 outs = 12% or 1 in 8 (which is worse than flopping a set with a pocket pair which is 1 in 7)
• Open-end or double belly-buster straight draw: 8 outs twice = 16 outs = 32% or 1 in 3
• Flush draw: 9 outs twice = 18 outs = 36% or a bit better than 1 in 3
• Open-end Straight draw and flush draw: 17 outs twice = 34 outs = 68% (even behind, you are the statistical favorite - it's a correct move to raise or value bet here)
• Two-pair improving to a full house: 4 outs twice = 8 outs = 16%
• Trips improving to full house or quads: 7 (turn) + 10 (river) = 17 outs = 34%

Of course, there is always the dreaded redraw which actually reduces these percentages, but you can include this easily enough in your calculations accordingly. And a runner-runner is usually less than 2%. Yup, a runner-runner is a 1 in 50 shot. Damn 2% suckout river-rats.

Your action, once you've calculated your odds, depends on your pot odds and perhaps implied pot odds. If your pot odds are better than your card odds, then you have a positive Expected Value (EV). If not, then you have a negative EV - so just fold.

Answer 5

Suppose you're at the flop with a pair and you want to know your chances of improving to a set by the end of the hand.

You've seen five cards from the deck (the two in your hand and the three on the flop); this means there are 47 cards in the deck that you haven't seen, and two of them will help you. Thus, the turn has a 2/47 chance of improving you to a set. If you don't improve on the turn then you still have two outs, but there are only 46 cards left in the deck, so you have a 2/46 (or 1/23) chance of improving on the river.

The odds of making your hand are equal to one (100%) minus the odds of not making your hand; in this case, that would be a 45/47 chance of not improving on the turn and a 44/46 chance of not improving on the river; this is a combined (45/47)*(44/46)=~91.58% chance of not improving on the turn or the river, which means you have approximately an 8.4% chance of hitting your set.

(Of course, this doesn't always mean an 8.4% chance of improving; you could unexpectedly backdoor a flush or straight)

Answer 6

The brute force method is to take all of the cards in the deck, deal out every possible combination of cards for a single player in the full course of the game and mark the total amount of times each possible hand occurs. Once you have the full stats, you can work out percentages for each hand by adding all of the hand counts together and then dividing the amount for each type of hand by the total.

Assume for a second there is a game with a deck of 7 cards, and each player gets 2 cards. The cards are a 2, 2, 4, 4, 6, 6 and 8. The hand ranks are that you either have a high card or you have a pair.

There are (7 - 5)! / 2 = 21 total combinations of cards. Three of the 21 combinations are pairs, so 18 of the hands are just high card hands. So in this game your chances of making a pair are about 3 / 21 ~ 14.29% and your chances of making a high card hand are 18 / 21 ~ 85.71%.

Answer 7

Just starting with 2 hole cards it gets complex fast
WIKI Poker Probability[]

For one card to come it is simple
% outs / remaining cards * 100
odds (remaining - outs) / outs : 1
odds is actually odds against
your odds (against) need to be less than you pot odds to mathematically call

Say you have a 4 flush on the flop
Opponent bets the pot
You pot odds are 2 : 1
You immediate hand odds are
(47 - 9) / 9 = 4.22
4.22 > 1 You are NOT getting immediate odds to call
If your opponent bet 1/3 the pot giving you 4:1 to call you are still NOT getting immediate odds to call

Let say 2 cards to come - it is more compex

Combination

47 cards left
combin(47,2) = 1081 (number of 2 card combos remaining)
combin(9,2) = 36 (you get both flush)
9 * (47 - 9) = 342 (you get one flush
(36 + 342)/1081 = 0.349676226 (the number you will see in poker odds)

Not

1 - ((not hit first card) * (not hit second card))
1 - ((47-9)/47*(46-9)/46) = 0.349676226

Possibilities

• first card hits and second misses
  (9/47)(46-8)/46 = 0.158186864
• first and second card hit
  (9/47)(8/46) = 0.033302498
• first card miss and second card hits
  ((47-9)/47)(9/46) = 0.158186864

Add those up and you get 0.349676226

Break it down
(9/47)(46-8)/46 + (9/47)(8/46) + ((47-9)/47)(9/46)
= (9/47)*(1 - 8/46) + (9/47)(8/46) + (1 - 9/47)(9/46)
= (9/47) - (9/47)(8/46) + (9/47)(8/46) + (1 - 9/47)(9/46)
the middle two cancel
= (9/47) + (9/46)(1 - 9/47)
= 0.349676226

People often go to (9/47) + (9/46)(1 - 9/47) without proof
A lot of people think that does not include both flush cards yet it does

Generalize

Just sub in number of outs for 9