mdlm
12-31-2002, 12:42 PM
I’m now halfway through the first six month phase of my poker study program so I though I would give myself a quiz.
I’d appreciate hearing back from you guys how I did on each question and what my overall grade (A-F) should be.
1. When he is behind in a football game, Brett Favre starts slinging the ball around which both increases the number of touchdowns and the number of interceptions that he throws. Is this a good strategy in football? Is it a good strategy in poker?
2. Some people have suggested that playing any two cards of the same suit is profitable. Why doesn’t anyone ever suggest that playing two connected cards is profitable?
3. What is the best non-pair hot-and-cold hand against 2h2c?
4. (Source: Turmel Test). You have AQ. Flop is KJx (no flush). How many bets do there need to be in the pot to justify a call?
5. (Source: Turmel Test). You have A5. Flop + turn is K985. How many bets do there need to be in the pot to justify a call?
6. (Source: Turmel Test). You have Ah5h. Flop is Kh5c2c. How many bets do there need to be in the pot to justify a call?
7. At the flop, you are first to act, heads up and need 3-1 pot odds to bet. There are two bets in the pot. Do you bet?
8. Hand A beats Hand B hot-and-cold. What is the probability that Hand A will be ahead of Hand B at the end of the flop?
My answers:
1. Slinging the ball around increases variance which increases the probability that Favre will win the game. Since this is the only probability that matters in football this is a good strategy. However, in poker this probability is irrelevant. Variance is bad unless it also increases EV. So this is a poor strategy for poker.
2. The probability that a flush will make it is greater than the probability that a straight will make it and a flush is better than a straight. These are two reasons why a flush draw is better than a straight draw and since the “any two suited cards have positive EV” claim is doubtful, the “any two connected cards have positive EV” claim is almost certainly incorrect.
3. JdTd. JdTd is better than JsTs because red ink weighs less than black ink so red flushes hit more often than black ones.
4. Seven outs (3 for A plus 4 for T). Probability of hitting on turn is approximately 2*7 = 14% so need 6-1 pot odds so need 6 bets in pot.
5. Five outs (2 for 5 plus 3 for A). 46 cards not known. So need 41-5 or 8 bets in pot.
6. Effectively 6 outs (3 for A, 2 for 5, 1 for backdoor flush or straight). 47 unseen. So need 41-6 or 7 bets in pot.
7. Yes. If you bet and opponent calls then you have the 3-1 pot odds. If you bet and opponent folds, you win the hand.
8. I don’t know, but this is a very interesting question.
==>
Comments on Comments
Ulysses disagrees with my assessment of Jones’ statement that when calculating the EV of a river bet you only need to calculate the EV when the opponent calls. Ulysses believes that this statement is correct because Jones is talking about the situation in which “you’re likely to have the best hand.” While it is true that Jones is talking about the situation in which you’re likely to have the best hand, this does not change the EV calculation. The EV calculation is always the same. You always look at all possible scenarios, assign probabilities and EVs to all of the scenarios, and then multiply and add. Even though you are likely to have the best hand, Jones concedes that you are sometimes wrong (that is why it may be incorrect to bet in the first place and you need to calculate the EV). Just as you might be wrong your opponent might be wrong and fold a better hand to a bet. Since the opponent is losing a pot in these situations while you are just losing a bet when you incorrectly bet, even a small probability of error on the opponent’s part will cause large changes in the EV of the river bet. Hence, this scenario needs to be included in the EV analysis and Jones is incorrect in excluding it.
Pudley4 makes a similar point. To summarize, the fact that you think you have the better hand is not what is key. Even if you are not bluffing and you think that you have the better hand and you think that you want the opponent to call, the case that I give above is critical to computing the EV. Obviously, if you were always right about having the better hand then the whole discussion would be moot. Figuring out the EV is only interesting because you might be wrong sometimes and just as you are wrong your opponent will be wrong sometimes. Jones dismisses this situation and says that your opponent will rarely fold a better hand. Jones simply makes no sense on this. If your opponent never folds a better hand, then your river bet is good if you win slightly more than 50% of the time (it’s slightly more than 50% instead of just 50% because your opponent might raise). Since he is assuming that you usually do have the better hand there is nothing to discuss. I’ve done some pencil and paper calculations and in some very reasonable cases if your opponent folds a better hand only once out of every 50 hands then you will do well to bet even if you have negative EV when the opponent calls, so considering this case is critical to getting the EV right. The fact that Jones ignores this case is an error.
Ulysses says that he doesn’t understand how I am calculating the probabilities. I am saying something very simple. The probability of being dealt QJ is the same as the probability of being dealt J7. When someone limps, the probability that they have QJ goes way up and the probability that they have J7 goes way down (you get information). When the BB checks the probabilities are roughly unchanged (you get no information).
So when the flop comes I have to be as worried about J7 as I am about QJ.
Dynasty correctly points out that a BB holding J7 will sometimes call a raise. In this case, the raise serves to punish the BB. I will still have to worry about J7 on the flop but at least I will be getting paid a bet to worry. /forums/images/icons/wink.gif
Pudley4 notes that there are some 100% suck out hands, such as 44 and 56s, that should be called instead of raised in LP. This may be true, but the analysis is quite subtle for loose-passive hold’em games. Note that it is almost always correct to call a raise preflop if you have limped so if there are enough limpers this is actually a value bet. /forums/images/icons/wink.gif I suppose it just has to be worked out. For example, if the preflop raise often leads to a free card on the flop then this would change the EV of the preflop raise markedly. Also, the preflop raise might “psychologically commit” your opponents to the pot and get them to call flop bets incorrectly.
Pudley4 also says that raising reduces the ROI. Let’s say this is true. It still doesn’t immediately follow that the raise is bad. Maximizing ROI is not exactly what you want to do. If the raise makes you more money than a call then it is worth making even if the ROI declines. Here’s a quick example: A horse is going off at odds of 5-1, but you think the correct odds are 2-1. The more you bet the lower your ROI will be because your bet lowers the odds. Nevertheless, you should continue to bet until the actual odds are slightly above 2-1 because every dollar that you bet that gets odds of better than 2-1 returns more than a dollar.
On the flush draw question, Pudley4 says that there is a difference between value bets and pot/effective/implied odds and that Jones is talking about value bets on the flop while I’m talking about pot/effective/implied odds. I’m saying something very simple: Jones says that you need slightly better than 2:1 to bet a flush draw and I think this is incorrect in some cases. Pudley4 says that a bet on the flop that is called by three opponents is a value bet. I think that this is a good bet in virtually any hold ‘em game but it is definitely not a value bet. Consider the following thought experiment: Suppose that every time you had a flush draw and missed on the turn your opponent bet an extremely large amount of money on the turn. You would have to fold every time you missed the draw on the turn. In this case, using odds of slightly better than 2:1 (which is what Jones recommends) would be a huge error (definitely not a value bet because you are not making money on the flop bet) since those are the odds that you need to make the draw by the river. Instead, you would use the odds you need to hit the flush on the turn. As the thought experiment shows it is not the case that “if you can get 3 people to call your flop bet, you should bet because you will make money off those bets.” If you changed this statement to 5 people, I would agree because that is enough to hit the flush on the turn. Here is another thought experiment: Suppose there is no money in the pot and you will called by three opponents. Would you bet a flush draw? Maybe or maybe not. It depends what happens on the turn. The point is that 2:1 is the minimum possible cutoff and in some games it will be significantly higher. As I mentioned before, I think that the clearest, simplest, and most accurate way to think about draws is to calculate the odds that you need to hit on the next card and then make an implied odds type adjustment if you are on the flop.
Angelo doesn’t understand my logic. My logic is simple: If I let the BB check into the pot he will win some pots that I otherwise would’ve won. If I can get rid of the BB for free then I will make more money. I don’t know if raising before the flop actually does this, but that is the idea. /forums/images/icons/wink.gif
==>
Goal Update
This past week, I spent approximately 21 hours on poker: 8.5 hours playing PokerPages tournaments, 7.5 hours playing ring games, and 5 hours on 2+2.
I did not spend any extra money this week. I have spent a total of $438.46 out of my $1000 budget.
An update on each of the four goals (which are to be accomplished by 3/30/03):
1. Read and study Jones’ “Winning Low Limit Hold ’Em”
I have confirmed one out of the three points I need to achieve this goal. The second and third points are pending based on the discussions with Ulysses and Pudley4.
2. Beat Acespade
Goal Completed on 11/5/02.
Over a period of 100 hours (3600 hands) I beat Acespade’s best lineup at the rate of over 4 BB/hr.
3. Beat Masque World Series of Poker
Goal Completed on 11/17/02
After playing Masque WSOP dozens of time I finally became the Masque WSOP Champion.
4. PokerPages 85% rating in one calendar month playing 20 tournaments
My rating got destroyed last week. In my last ten tournaments or so I only picked up one final table finish. Absolutely horrible. I now have only three shots left to his this goal. If I fail, that will be it for The Newbie. I’m very glad that I made this one of my goals. If I don’t have the psychological strength and technical poker ability to do this, I’m unlikely to be a decent poker player. I played a total of six tournaments and finished #52 out of 91, #22 out of 98, #28 out of 43, #7 out of 59, #72 out of 97, and #75 out of 102. I finished with a rating of 78.06%.
I’d appreciate hearing back from you guys how I did on each question and what my overall grade (A-F) should be.
1. When he is behind in a football game, Brett Favre starts slinging the ball around which both increases the number of touchdowns and the number of interceptions that he throws. Is this a good strategy in football? Is it a good strategy in poker?
2. Some people have suggested that playing any two cards of the same suit is profitable. Why doesn’t anyone ever suggest that playing two connected cards is profitable?
3. What is the best non-pair hot-and-cold hand against 2h2c?
4. (Source: Turmel Test). You have AQ. Flop is KJx (no flush). How many bets do there need to be in the pot to justify a call?
5. (Source: Turmel Test). You have A5. Flop + turn is K985. How many bets do there need to be in the pot to justify a call?
6. (Source: Turmel Test). You have Ah5h. Flop is Kh5c2c. How many bets do there need to be in the pot to justify a call?
7. At the flop, you are first to act, heads up and need 3-1 pot odds to bet. There are two bets in the pot. Do you bet?
8. Hand A beats Hand B hot-and-cold. What is the probability that Hand A will be ahead of Hand B at the end of the flop?
My answers:
1. Slinging the ball around increases variance which increases the probability that Favre will win the game. Since this is the only probability that matters in football this is a good strategy. However, in poker this probability is irrelevant. Variance is bad unless it also increases EV. So this is a poor strategy for poker.
2. The probability that a flush will make it is greater than the probability that a straight will make it and a flush is better than a straight. These are two reasons why a flush draw is better than a straight draw and since the “any two suited cards have positive EV” claim is doubtful, the “any two connected cards have positive EV” claim is almost certainly incorrect.
3. JdTd. JdTd is better than JsTs because red ink weighs less than black ink so red flushes hit more often than black ones.
4. Seven outs (3 for A plus 4 for T). Probability of hitting on turn is approximately 2*7 = 14% so need 6-1 pot odds so need 6 bets in pot.
5. Five outs (2 for 5 plus 3 for A). 46 cards not known. So need 41-5 or 8 bets in pot.
6. Effectively 6 outs (3 for A, 2 for 5, 1 for backdoor flush or straight). 47 unseen. So need 41-6 or 7 bets in pot.
7. Yes. If you bet and opponent calls then you have the 3-1 pot odds. If you bet and opponent folds, you win the hand.
8. I don’t know, but this is a very interesting question.
==>
Comments on Comments
Ulysses disagrees with my assessment of Jones’ statement that when calculating the EV of a river bet you only need to calculate the EV when the opponent calls. Ulysses believes that this statement is correct because Jones is talking about the situation in which “you’re likely to have the best hand.” While it is true that Jones is talking about the situation in which you’re likely to have the best hand, this does not change the EV calculation. The EV calculation is always the same. You always look at all possible scenarios, assign probabilities and EVs to all of the scenarios, and then multiply and add. Even though you are likely to have the best hand, Jones concedes that you are sometimes wrong (that is why it may be incorrect to bet in the first place and you need to calculate the EV). Just as you might be wrong your opponent might be wrong and fold a better hand to a bet. Since the opponent is losing a pot in these situations while you are just losing a bet when you incorrectly bet, even a small probability of error on the opponent’s part will cause large changes in the EV of the river bet. Hence, this scenario needs to be included in the EV analysis and Jones is incorrect in excluding it.
Pudley4 makes a similar point. To summarize, the fact that you think you have the better hand is not what is key. Even if you are not bluffing and you think that you have the better hand and you think that you want the opponent to call, the case that I give above is critical to computing the EV. Obviously, if you were always right about having the better hand then the whole discussion would be moot. Figuring out the EV is only interesting because you might be wrong sometimes and just as you are wrong your opponent will be wrong sometimes. Jones dismisses this situation and says that your opponent will rarely fold a better hand. Jones simply makes no sense on this. If your opponent never folds a better hand, then your river bet is good if you win slightly more than 50% of the time (it’s slightly more than 50% instead of just 50% because your opponent might raise). Since he is assuming that you usually do have the better hand there is nothing to discuss. I’ve done some pencil and paper calculations and in some very reasonable cases if your opponent folds a better hand only once out of every 50 hands then you will do well to bet even if you have negative EV when the opponent calls, so considering this case is critical to getting the EV right. The fact that Jones ignores this case is an error.
Ulysses says that he doesn’t understand how I am calculating the probabilities. I am saying something very simple. The probability of being dealt QJ is the same as the probability of being dealt J7. When someone limps, the probability that they have QJ goes way up and the probability that they have J7 goes way down (you get information). When the BB checks the probabilities are roughly unchanged (you get no information).
So when the flop comes I have to be as worried about J7 as I am about QJ.
Dynasty correctly points out that a BB holding J7 will sometimes call a raise. In this case, the raise serves to punish the BB. I will still have to worry about J7 on the flop but at least I will be getting paid a bet to worry. /forums/images/icons/wink.gif
Pudley4 notes that there are some 100% suck out hands, such as 44 and 56s, that should be called instead of raised in LP. This may be true, but the analysis is quite subtle for loose-passive hold’em games. Note that it is almost always correct to call a raise preflop if you have limped so if there are enough limpers this is actually a value bet. /forums/images/icons/wink.gif I suppose it just has to be worked out. For example, if the preflop raise often leads to a free card on the flop then this would change the EV of the preflop raise markedly. Also, the preflop raise might “psychologically commit” your opponents to the pot and get them to call flop bets incorrectly.
Pudley4 also says that raising reduces the ROI. Let’s say this is true. It still doesn’t immediately follow that the raise is bad. Maximizing ROI is not exactly what you want to do. If the raise makes you more money than a call then it is worth making even if the ROI declines. Here’s a quick example: A horse is going off at odds of 5-1, but you think the correct odds are 2-1. The more you bet the lower your ROI will be because your bet lowers the odds. Nevertheless, you should continue to bet until the actual odds are slightly above 2-1 because every dollar that you bet that gets odds of better than 2-1 returns more than a dollar.
On the flush draw question, Pudley4 says that there is a difference between value bets and pot/effective/implied odds and that Jones is talking about value bets on the flop while I’m talking about pot/effective/implied odds. I’m saying something very simple: Jones says that you need slightly better than 2:1 to bet a flush draw and I think this is incorrect in some cases. Pudley4 says that a bet on the flop that is called by three opponents is a value bet. I think that this is a good bet in virtually any hold ‘em game but it is definitely not a value bet. Consider the following thought experiment: Suppose that every time you had a flush draw and missed on the turn your opponent bet an extremely large amount of money on the turn. You would have to fold every time you missed the draw on the turn. In this case, using odds of slightly better than 2:1 (which is what Jones recommends) would be a huge error (definitely not a value bet because you are not making money on the flop bet) since those are the odds that you need to make the draw by the river. Instead, you would use the odds you need to hit the flush on the turn. As the thought experiment shows it is not the case that “if you can get 3 people to call your flop bet, you should bet because you will make money off those bets.” If you changed this statement to 5 people, I would agree because that is enough to hit the flush on the turn. Here is another thought experiment: Suppose there is no money in the pot and you will called by three opponents. Would you bet a flush draw? Maybe or maybe not. It depends what happens on the turn. The point is that 2:1 is the minimum possible cutoff and in some games it will be significantly higher. As I mentioned before, I think that the clearest, simplest, and most accurate way to think about draws is to calculate the odds that you need to hit on the next card and then make an implied odds type adjustment if you are on the flop.
Angelo doesn’t understand my logic. My logic is simple: If I let the BB check into the pot he will win some pots that I otherwise would’ve won. If I can get rid of the BB for free then I will make more money. I don’t know if raising before the flop actually does this, but that is the idea. /forums/images/icons/wink.gif
==>
Goal Update
This past week, I spent approximately 21 hours on poker: 8.5 hours playing PokerPages tournaments, 7.5 hours playing ring games, and 5 hours on 2+2.
I did not spend any extra money this week. I have spent a total of $438.46 out of my $1000 budget.
An update on each of the four goals (which are to be accomplished by 3/30/03):
1. Read and study Jones’ “Winning Low Limit Hold ’Em”
I have confirmed one out of the three points I need to achieve this goal. The second and third points are pending based on the discussions with Ulysses and Pudley4.
2. Beat Acespade
Goal Completed on 11/5/02.
Over a period of 100 hours (3600 hands) I beat Acespade’s best lineup at the rate of over 4 BB/hr.
3. Beat Masque World Series of Poker
Goal Completed on 11/17/02
After playing Masque WSOP dozens of time I finally became the Masque WSOP Champion.
4. PokerPages 85% rating in one calendar month playing 20 tournaments
My rating got destroyed last week. In my last ten tournaments or so I only picked up one final table finish. Absolutely horrible. I now have only three shots left to his this goal. If I fail, that will be it for The Newbie. I’m very glad that I made this one of my goals. If I don’t have the psychological strength and technical poker ability to do this, I’m unlikely to be a decent poker player. I played a total of six tournaments and finished #52 out of 91, #22 out of 98, #28 out of 43, #7 out of 59, #72 out of 97, and #75 out of 102. I finished with a rating of 78.06%.