Theory of Poker: Chapters 5-7 Discussion

[MEbenhoe]

Well I'm trying something different for this section because I think in the first section some people might not have been able to ask questions they needed to ask because too much of the focus might have been on the discussion starter questions I put out there. I could be way off too, but for this section I'm just gonna post chapters to start posting under and then hopefully you'll post your thoughts and questions on these chapters. Whether this sucks and you think the other way was much better or if you like this better just let me know.

[MEbenhoe]

This section is for discussing Pot Odds

[MEbenhoe]

This section is for discussing Effective Odds

[MEbenhoe]

This section is for discussing Implied Odds and Reverse Implied Odds

[Smokey98]

I think that is best as I didn't get too much out of the first go around, however I think that I'm getting pot odds and all fine.

[sethypooh21]

I think the perhaps the most overlooked, and one of the more important concepts in this chapter is the relationship between position ond pot odds. P.41. We've all had it happen, we call a flop bet with a marginal draw thinking "pot odds" and suddenly it gets raised behind us and reraised by the orignal bettor. It's two back to us. It could get capped behind us. We might have correct odds to call each individual time, but as a whole, calling the 3-4 bets is clearly a losing proposition, all because we ignored the possibility of more action behind us.

[chaos]

I agree. This is an important concept. Caro also mentions that you can loosen up if the raiser or bettor was immediately to your left, then when it gets back to you you get to "close the action." If you do not get to close the action you must take into account the possibility of a raise behind you reducing your pot odds.

[Gamblor]

Agreed - I'll call with a ton of hands - bottom pair, overcards, etc. etc. when closing the action.

A lot of times I'll see a lot of turns if I'm the limper closest to the preflop raiser because I have the "relative button" - as in, assuming the preflop raiser bets the flop, I'll be last to act.

[Gamblor]

I remember reading somewhere once that you could actually make a profit if you bought in for 1 BB every hand and waited for high cards and pairs before throwing it all in. You wouldn't have to pay off rivers and you'd get to see all 7 cards.

Anyone have a clue on how this relates to (Reverse) Implied Odds?

[MEbenhoe]

[ QUOTE ]
I think that is best as I didn't get too much out of the first go around, however I think that I'm getting pot odds and all fine.

[/ QUOTE ]

How would you like a combination of the two? Like for each section have one question to get the discussion started and then it should easily spread out from there?

[RobGW]

Regarding no limit with small blinds. As in a cash game or early tourneys. Chapter four says we should play tighter with small blinds. Chapter seven says we are getting great implied odds in no limit. What strategy do you guys think is better and why?

[MEbenhoe]

[ QUOTE ]
Regarding no limit with small blinds. As in a cash game or early tourneys. Chapter four says we should play tighter with small blinds. Chapter seven says we are getting great implied odds in no limit. What strategy do you guys think is better and why?

[/ QUOTE ]

The way you make this decision is by looking at the size of the blinds compared to the size of future bets. In NL where you can often limp in cheap, but still break someone if you hit your hand the implied odds are through the roof. This is also true in limit hold em games with mini blinds or in 7 card stud games where the bring in bet is abnormally low.

[Gamblor]

Remember also that implied odds require a degree of "hiddenness" if you hit your hand.

For example, if you hold A /images/graemlins/heart.gif J /images/graemlins/diamond.gif and the board is T /images/graemlins/diamond.gif 8 /images/graemlins/spade.gif 9 /images/graemlins/club.gif, you're not going to get much action if a Q hits so you might be better off folding to a solid bet.

[Boylermaker]

[ QUOTE ]
The way you make this decision is by looking at the size of the blinds compared to the size of future bets. In NL where you can often limp in cheap, but still break someone if you hit your hand the implied odds are through the roof. This is also true in limit hold em games with mini blinds or in 7 card stud games where the bring in bet is abnormally low.

[/ QUOTE ]

Regarding no-limit, another way to say it would be to look at the size of the blinds relative to the size of the chip stacks on the table. With small blinds in no-limit, you obiously would want to come in with more speculative hands, as, as stated, you can break someone. But this is only worthwhile if the size of the stack you are breaking is significant.

[Boylermaker]

I think a combo of the two is good. Start one thread with a question to get the discussion going, and start a second for posters' miscellaneous questions from that chapter.

So far so good, though, as I think the discussion group was an excellent idea.

[sethypooh21]

[ QUOTE ]
In NL where you can often limp in cheap, but still break someone if you hit your hand the implied odds are through the roof. This is also true in limit hold em games with mini blinds or in 7 card stud games where the bring in bet is abnormally low.

[/ QUOTE ]

This is interesting about the mini-blinds games. I've heard it either way, as in you should play looser cause you get in cheap or tighter because your draws are likely to be expensive relative to the pot. I think a mini blind structure probably increases the relative value of mid/low pocket pairs, while decreasing the value of connectors and suited cards somewhat linearly with the postflop aggression of the rest of the table.

Loose limping can also be severely punished. Some one can effectively 3-bet you at any time. (Call the $2, raise to $6.)

[Boylermaker]

[ QUOTE ]
I think a mini blind structure probably increases the relative value of mid/low pocket pairs, while decreasing the value of connectors and suited cards somewhat linearly with the postflop aggression of the rest of the table.

[/ QUOTE ]

I think it would increase the value of suited connectors as well. The idea being that you sometimes flop a big hand with suited connectors, not necessarily just a draw, and you are risking a relatively small number of chips to see that flop.

[RobGW]

Qwiz: Does everyone know the quick way to figure your pot odds besides memorizing? Hint: Use the namesake of this forum.

[As Zehn]

I have a question on calculating odds. If you have four to a flush on the flop what are the odds of getting the flush by the river. Is it 9/38 + 9/37 = .48 and then 1/.48 = 2.083 to 1?

[As Zehn]

Reading the section on the impact of position on pot odds was worth the cost of the book.

[BugsBunny]

1 - (38/47 * 37/46) = 0.3496762257

In plain English subtract the probability of not hitting the flush on either the turn or river from 1. You're left with the flush probability.

Odds are (1 - 0.3496762257)/0.3496762257 = 1.8597883598 to 1 against

The other way to do this is:
9/47 + (1 - 9/47) * 9/46 =
0.1914893617 + 0.8085106383 * 0.1956521739 =
0.1914893617 + 0.158186864 = 0.3496762257

In plain english, the probability of hitting on the turn (9/47) is added to the probability of hitting on the river (9/46) when you didn't hit the turn (1 - 9/47).

[Smokey98]

I think the best way to do this is by asking if people are struggling with a certain aspect of the chapters and then elaborating on those areas.

[theghost]

[ QUOTE ]
Qwiz: Does everyone know the quick way to figure your pot odds besides memorizing? Hint: Use the namesake of this forum.

[/ QUOTE ]

I use 13-(# of outs)=odds, so for 5 outs you need 8 bets in the pot to call (13-5=8).
It works to give rough odds when you have 5 through 9 outs (and you almost always have odds to call with 10 or more outs).
You need to remember the odds for 4 or less outs:
4- 10.5
3- 15
2- 22
(roughly)

Another trick I use to give me a rough figure on % chance to make your hand:

on the flop, with two cards to come, 4(# of outs)=%
on the turn, with one card to come, 2(# of outs)=%

Basically, 2% per out per card coming (so a 4 outer on the flop has a 16% chance to come in by the river: 4x4=16).

Admittedly, this is *very* rough - but sometimes usefull. The #'s come in a little low (conservative) because there are 47(46) cards left (where 50 would make 2% per out per card precise).

[sammy_g]

[ QUOTE ]
Qwiz: Does everyone know the quick way to figure your pot odds besides memorizing? Hint: Use the namesake of this forum.

[/ QUOTE ]
One trick I've heard is to multiply your outs by two then add two to get a percentage. For instance, if you have 8 outs, that's

8(2) + 2 = 18

So you have an ~18% chance of catching one of your outs on the next card.

Is this what you were referring to?

[New York Jet]

First time poster, long time reader.

I use two different methods.

Method 1. 2% x Cards to come x Outs (i.e. flush draw on the flop would be 2% * 2 cards to come * 9 outs = 36%)

Method 2. On the flop, use 4% x number of outs and on the turn, use 2.2 x number of outs.

New York Jet

[As Zehn]

When should you use effective odds v implied/reverse implied odds? This section is beginning to really confuse me. Apparently I know even less than I thought. /images/graemlins/confused.gif

[SeppDeitrich]

another point is that the skill of your opponent's postflop play would have a larger effect on the value of your starting hands in this game. Compared to a game with a normal blind structure more money is going into the pot after the flop compared to the amount in the pot before the flop, so the advantage you gain from superior postflop play is magnified. So with regards to suited connectors, they may go down in value for games with tight/ good opponents, but go way up against loose/ bad opponents.

[Boylermaker]

[ QUOTE ]
When should you use effective odds v implied/reverse implied odds? This section is beginning to really confuse me. Apparently I know even less than I thought.

[/ QUOTE ]

The easy answer is that you should always consider them. But you probably already do, maybe without knowing it. When looking at stuff like effective odds, implied odds, reverse implied odds, etc., what you are in essence really doing is asking yourself questions like the folowing:

"If I call this bet, will this pot get raised and reraised behind me?"

"What does this paired and coordinated board do to my straight draw?"

"Will I get sufficient action of my hand if I make it?"

I try to not get too bogged down in the math when considering these factors, but by being aware of what is going on in the hand, and in the game in general, and by considering questions like this when thinking about what action to take, you can further increase your chances of making the correct decision. Like if the odds appear okay to call a bet, but you suspect there is a better than average chance that the pot will get raised/reraised behind you, you know that the odds might in fact not be sufficient to call, and you can correctly fold.

[Smokey98]

Can someone draw this out? Put it on a 3rd grade level for me. I'm just not getting that. So if you've 4 to the flush of spades then you need one of the remaining 9 spades to hit the flush. So that gives you a 9/47 or 9 that will hit and 38 that don't. I get that part. So would the odds of hitting it on the turn be 1/5.2? 47 divided by 9? So then how do you take that one step further to the river?

[Smokey98]

Can someone like Ed or David or Mason confirm if the methods that Sammy and New York Jet are using are good methods? No offense guys. /images/graemlins/wink.gif

[theghost]

[ QUOTE ]
So would the odds of hitting it on the turn be 1/5.2? 47 divided by 9?

[/ QUOTE ]
47/9=5.22 so 9 goes into 47 5.22 times. 9 of the 47 are good for you, so 38 are bad.
Your odds are 1 to 4.22 (for a total of 5.22), follow?

[ QUOTE ]
So then how do you take that one step further to the river?

[/ QUOTE ]
You only consider pot odds for the bet at hand. If it's $1 to you and there are $5 in the pot, you can call (or raise for value, but that's another story). You can call because 5>4.22.

If the bet size doubles for the next street (turn), you might not have odds to call again ($2 to you, $7 in the pot would be a bad call because 3.5<4.11).

You might, however want to call this anyway if you are drawing to the 1st, 2nd, or 3rd nuts (or even on any unpaired board, for a flush draw with two in your hand). This is because you anticipate to get paid off when you make your hand (implied odds). Make sense?

[Smokey98]

[ QUOTE ]
[ QUOTE ]
So would the odds of hitting it on the turn be 1/5.2? 47 divided by 9?

[/ QUOTE ]
47/9=5.22 so 9 goes into 47 5.22 times. 9 of the 47 are good for you, so 38 are bad.
Your odds are 1 to 4.22 (for a total of 5.22), follow?

[/ QUOTE ]
So you’re dividing the 38 by 9 and not the 47?


[ QUOTE ]
So then how do you take that one step further to the river?

[/ QUOTE ]
You only consider pot odds for the bet at hand. If it's $1 to you and there are $5 in the pot, you can call (or raise for value, but that's another story). You can call because 5>4.22.
[ QUOTE ]
If the bet size doubles for the next street (turn), you might not have odds to call again ($2 to you, $7 in the pot would be a bad call because 3.5<4.11).

[/ QUOTE ]
If you bet your $2 your opponent has to bet $2 as well making it $9 and pot odds of 4.50.

[MEbenhoe]

The way that Sammy uses is correct for determining a percent chance of hitting your hand, but be warned the less outs you have the more your the result of that equation will be off by. New York Jets way of determining them is a quick way to approximate your percentage chance, but I would never use this method for two reasons; the percentages are only close approximations and will be off by more and more the more outs you have, and his calculations calculate your chance of hitting on the next two cards rather than only hitting on the next card which are the odds you should be concerned with.

Overall though all these methods are somewhat pointless, because the percentage chance you have of hitting a hand means nothing, except for putting up numbers on poker shows on tv. You should only be concerned with odds. Heres an example to show you what I mean:

Lets say you have an inside straight draw on the flop, meaning you have 4 outs to make your hand. You're trying to determine if you can call a $2 bet with $16 in the pot. Using Sammy's method will tell you that you have a 10% chance of hitting your straight on the next card (you actually only have an 8.5% chance), and using New York Jets method will tell you that you 16% chance of hitting your inside straight by the river. Do these numbers actually do you any good? Not really. Plus its just as easy to figure out that you have 4 outs of hitting with 47 unseen cards so your odds of hitting are 43:4 or 10.75:1, so you don't have odds to call on this hand. Although Sammy's and New York Jets methods could cause you believe that you do have the odds. In all honesty I say use the way I just showed you, its not that hard of math and you're assured an exact answer that you can actually use.

I think people are just obsessed with percentages because they're not used to working with odds, but using percentages will do nothing but cause you to make mistakes in the long run.

[theghost]

[ QUOTE ]
So you’re dividing the 38 by 9 and not the 47?

[/ QUOTE ]
Yes, basically. If we simplify the math a little, and round off the decimals we can look at it this way:
There are 5 groups of nine cards, one of these groups helps you and the other four don't. (9 cards help, 38 don't)
So 4 times you lose, 1 time you win.
You want that one win to at least make back enough to cover the 4 times you will lose, so you need at least 4 bets in the pot to your 1 every time.


[ QUOTE ]
If you bet your $2 your opponent has to bet $2 as well making it $9 and pot odds of 4.50.

[/ QUOTE ]
You're getting the right idea, but remember in the example that you are calling his bet, so his $ is already in the pot (5+2=7, 2 to you). If there are others left to act behind you, you have to anticipate what you think they will do.
If you are pretty sure they will call, you might call because you expect that money to be in the pot.
If you think they will raise, you might not want to call, because it will end up costing you double to see you card. (note that if they raise, you have to call because the bet size is the same and the pot is even bigger.)

[New York Jet]

No need for Ed, David, or Mason to verify the method. I'll show you the math and you can decide for yourself. Notice that Method 2 is much more accurate for the Turn percentages.

Method 1
FLOP
Outs Actual % 4% * Outs Difference
15 54.1% 60% -5.9%
14 51.2% 56% -4.8%
13 48.1% 52% -3.9%
12 45.0% 48% -3.0%
11 41.7% 44% -2.3%
10 38.4% 40% -1.6%
9 35.0% 36% -1.0%
8 31.5% 32% -0.5%
7 27.8% 28% -0.2%
6 24.1% 24% 0.1%
5 20.3% 20% 0.3%
4 16.5% 16% 0.5%
3 12.5% 12% 0.5%
2 8.4% 8% 0.4%
1 4.3% 4% 0.3%
TURN
Outs Actual % 2% * Outs Difference
15 32.6% 30.0% 2.6%
14 30.4% 28.0% 2.4%
13 28.3% 26.0% 2.3%
12 26.1% 24.0% 2.1%
11 23.9% 22.0% 1.9%
10 21.7% 20.0% 1.7%
9 19.6% 18.0% 1.6%
8 17.4% 16.0% 1.4%
7 15.2% 14.0% 1.2%
6 13.0% 12.0% 1.0%
5 10.9% 10.0% 0.9%
4 8.7% 8.0% 0.7%
3 6.5% 6.0% 0.5%
2 4.3% 4.0% 0.3%
1 2.2% 2.0% 0.2%

Method 2
FLOP
Same as Method 1.
TURN
Outs Actual % 2.2% * Outs Difference
15 32.6% 33.0% -0.4%
14 30.4% 30.8% -0.4%
13 28.3% 28.6% -0.3%
12 26.1% 26.4% -0.3%
11 23.9% 24.2% -0.3%
10 21.7% 22.0% -0.3%
9 19.6% 19.8% -0.2%
8 17.4% 17.6% -0.2%
7 15.2% 15.4% -0.2%
6 13.0% 13.2% -0.2%
5 10.9% 11.0% -0.1%
4 8.7% 8.8% -0.1%
3 6.5% 6.6% -0.1%
2 4.3% 4.4% -0.1%
1 2.2% 2.2% 0.0%


New York Jet

[MEbenhoe]

Thats all and good, but you're giving them useless numbers. If you can do math like that you can just as easily do the simple math required to figure out your pot odds rather than your percent of hitting, and with pot odds they're going to be accurate everytime, there's no being off by 2%.

Percentages all too often cause people to make mistakes as well. Too many people think 20% means you need 5:1 to call when you really only need 4:1 to call.

[New York Jet]

[ QUOTE ]
The way that Sammy uses is correct for determining a percent chance of hitting your hand, but be warned the less outs you have the more your the result of that equation will be off by. New York Jets way of determining them is a quick way to approximate your percentage chance, but I would never use this method for two reasons; the percentages are only close approximations and will be off by more and more the more outs you have, and his calculations calculate your chance of hitting on the next two cards rather than only hitting on the next card which are the odds you should be concerned with.

[/ QUOTE ]

Not so fast my friend. Method 2 will calculate the Flop percentages within 1% up to 9 outs and the Turn percentages within 0.4% up to 15 outs. If you have more than 15 outs on the Turn, but still need to calculate whether calling is correct, you should think about changing tables because your's is way too tight.

[ QUOTE ]
Overall though all these methods are somewhat pointless, because the percentage chance you have of hitting a hand means nothing, except for putting up numbers on poker shows on tv. You should only be concerned with odds. Heres an example to show you what I mean:

Lets say you have an inside straight draw on the flop, meaning you have 4 outs to make your hand. You're trying to determine if you can call a $2 bet with $16 in the pot. Using Sammy's method will tell you that you have a 10% chance of hitting your straight on the next card (you actually only have an 8.5% chance), and using New York Jets method will tell you that you 16% chance of hitting your inside straight by the river. Do these numbers actually do you any good? Not really. Plus its just as easy to figure out that you have 4 outs of hitting with 47 unseen cards so your odds of hitting are 43:4 or 10.75:1, so you don't have odds to call on this hand. Although Sammy's and New York Jets methods could cause you believe that you do have the odds. In all honesty I say use the way I just showed you, its not that hard of math and you're assured an exact answer that you can actually use.

[/ QUOTE ]

Two more points to make here.
1. I'll agree that determining the percentages for turn and river are not accurate when considering calling. They provide you with the incorrect effective odds. However, the method you suggest is also inaccurate. Your method does not take in to account the implied odds. A method I use is to split the difference between the Turn odds and the River odds. So in the case of an inside straight draw, I'll take the odds for both cards (5.1:1) and the odds for the last card (10.5:1), and average them out (7.8:1). So I will call an inside straight draw if I can get 8 to 1 odds. Now you must use a little common sense. If you are heads-up, 8 to 1 will not be enough, but 10 to 1 would be.

2. I could not agree more that percentages are not as useful as odds. What I suggest is memorizing the odds for calling the last card. Normally, you only need to know your odds with 9 or less outs. Remember the odds after the flop are slightly greater than after the turnl. Also remember that the odds for the turn and river are slightly less than half of the final card odds. If you have more than 9 outs, you should almost always call (or move to a looser table).

Here are some odds I use for after the flop to account for the implied odds. These are for a low limit, loose table. They would not work as well for a tight table.
9 Outs 3 to 1
8 OUts 3 to 1
7 Outs 4 to 1
6 Outs 5 to 1
5 Outs 6 to 1
4 Outs 8 to 1
3 Outs 11 to 1
2 Outs 17 to 1
1 Out 34 to 1

New York Jet

[MEbenhoe]

[ QUOTE ]

1. I'll agree that determining the percentages for turn and river are not accurate when considering calling. They provide you with the incorrect effective odds. However, the method you suggest is also inaccurate. Your method does not take in to account the implied odds. A method I use is to split the difference between the Turn odds and the River odds. So in the case of an inside straight draw, I'll take the odds for both cards (5.1:1) and the odds for the last card (10.5:1), and average them out (7.8:1). So I will call an inside straight draw if I can get 8 to 1 odds. Now you must use a little common sense. If you are heads-up, 8 to 1 will not be enough, but 10 to 1 would be.


[/ QUOTE ]

I merely ignored implied odds in order to prove the point in the difference in calculating odds vs percentages (that arent even accurate). The point still stands as being true; Percentages are completely useless in deciding how to play a hand.

[BigBluffer]

On pages 49-50, Sklansky is writing about the final hand of the 1980 WSOP Championship, where Brunson admits he made a mistake by not betting enough on the flop to make Ungar get out of the hand, allowing Ungar to pick up his gutshot straight on the turn and win the Championship.

Sklansky writes, "Brunson acknowledged he played incorrectly in betting $17,000 on the flop...he should have bet more than Ungar would have been able to call...even in terms of implied odds." $30,000 was in the pot pre-flop.

How much should Doyle have bet on the flop? How do you calculate that amount?

[DonkeyKong]

It's actually pg 57-58 in Theory of Poker...

Apparently, the equation is [(Doyles Stack) + (What Doyle has put In Already)] / (Doyles Bet) > Odds of a Gut Shot

So, (232,500 + 15,000) / X = 10.75x
= ~23,000

Or he needs to bet >23,000 to offer Stu an amount that makes him overpay the right price in terms of IMPLIED odds... call it $25k, or 80% of the pot...

But Doyle had top 2 pair and wanted a call from Ungar. In retrospect, he says he didn't bet enough but if $23k is the right math number, and Doyle wanted a call -- a little less than $23k seems logical. Maybe $20k, or 2/3 the pot.

Re. the $17,000; Stu said "I wouldn't have called too much more than that for a gut shot" so it seems virtually impossible to know what the right bet amount was... Something more than $17k but less than $25k as Doyle wanted a call... I guess you could argue anything from $20-25k... Doyle was behind in chips and flopped a big heads-up hand... betting $17k was tempting Stu but also risking his entire stack...

please, someone correct me if I am wrong but this is how I read it...

[DonkeyKong]

I guess the moral of the story is; err on the side of too much bet and risk a non-call rather than allowing an opponent to hit a miracle card... Stu might have called more anyway, who knows, so Doyle might have been going down the drain on that hand had he overbet the mathematically correct amount. Making an inside straight when your opponent has 2 pair in a heads-up match is going to end up ugly... the only question is how much more Stu would have called.

[BigBluffer]

DonkeyKong, thanks for the math. That really helps. But what about the logic behind the equation? Any insight into how the equation was arrived at?

By the way, it IS discussed on pages 49-50 of my edition. /images/graemlins/wink.gif

[DonkeyKong]

<<Any insight into how the equation was arrived at?>>

well it is clear how it was arrived at... It was Doyles stack size divided by the odds of making a gutshot. but in reality, it was impossible for Doyle to put Stu on exactly a gut shot draw. he can't possibly know that.

What you can do is read the flop texture and bet an amount related to likely draws that your opponent might have. Meaning if there is a coordinated board, your opponent might very well have a typical 8 or 9 out draw. You would then make a corresponding bet to ensure that your opponent isn't able to draw for a good price.

Implied odds refers to the amount of money that you are likely to win that is in addition to what is already in the pot. Since they were playing NL, Doyles entire stack is used for this calculation since Stu figured that Doyle would not be able to read the fact that Stu made a straight given the cards on the board. Stu correctly called $17,000 because the implied odds were good for Stu. He was 10.75 to 1 to make his draw but would get paid 14x+ (all Doyles chips) if he made it.

If you are interested, buy Bob Ciaffone's book 'Pot Limit & No Limit Poker'... I read it and its pretty good. I am not much of a guy to talk about NL though so I will stop there. I am a still a hack when it comes to NL.

[BigBluffer]

[ QUOTE ]
Stu correctly called $17,000 because the implied odds were good for Stu. He was 10.75 to 1 to make his draw but would get paid 14x+ (all Doyles chips) if he made it.


[/ QUOTE ]

For some reason, I couldn't put it together until you put it that way. Sometimes it takes a little longer than usual for things to sink into my head. /images/graemlins/crazy.gif Thanks.

And I agree that the calculation of how much Doyle should bet is made more difficult because of the number of different draws/made hands Stu could have. He could have two pair. He might even be sucking Doyle in with a set. It's those unknowns that make the decision not so cut and dry. Of course hindsight is 20-20. /images/graemlins/wink.gif

BB