Professional No-Limit Hold 'em Study Group Day 1

[Matt Flynn]

Hi everyone welcome to the study group. We're starting with any and all questions on Odds and Outs or Bet Sizing, the first two chapters in PNL1.

Fire away!


Some thoughts:

1. Implied odds are what count. Always consider what you might win or lose if you hit your hand, or if your opponent hits his.

2. Poker players tend to be optimistic about their implied odds. So be cautious. Consider whether you're discounting your outs enough, and also whether you really are going to get half of your opponent's stack on average when you hit your flush.

3. Bet sizing is an art form and not something you can learn overnight.

4. A thought experiment: suppose you had to play online poker exclusively and the only bet you could make was a fixed percentage of the pot. They offer you a "bet pot" button with no slider to adjust, but you can change "bet pot" to any size. What would you choose?

I would choose 2/3 the pot or perhaps slightly more, like 70% of the pot. 2/3 is an excellent all-purpose bet size. It's also rarely THE best bet size. Rather, it's a good compromise.

5. Continuation bets can be small, but if you use a small cbet like 1/3 or 1/2 the pot you had better be making that same bet size with at least some of your bigger hands. Otherwise an astute opponent will eat you alive. This is a major and common weakness seen in live $1-$2 games. A tight player makes a big raise preflop then makes a small bet postflop, say $10 into a $30 pot. When that happens, first look around to see if anyone else in the pot looks interested. If not, seriously consider making a 2/3 pot raise or similar. If your opponent plays back at you after such a weak cbet, he's less of a mark and should get more respect later on.

By the same token, when you make small cbets and face a couple decent opponents mixed in with the weak ones, you often want to play all hands that way when in pots against the better opponents so you can sucker them in. However, when you end up in pots against the weak ones, just make bigger bets with your big hands until they fight back against your smaller cbets when you miss.

The trick is to exploit them as much as they'll let you. Against very bad opponents that are commonly found in low-limit live games and the lowest online limits, you can get away with this stuff and should take advantage.

[ciro bonano]

Maybe I've missed the PSA, but the odds table in appendix B is incorrect. I don't understand how such a thing can happen, checking e.g. wikipedia is the _least_ one can do, right? </rant>

Some minor note which I find useful because I'm a nit: the 'multiply by four' rule can be improved a bit to
p = outs * 4 - (number of outs above 8), e.g. for 15 outs
p = 15*4 - 7 = 53% instead of 60% (real number 54%).

Then a real question:

On page 12 you discuss the K[img]/images/graemlins/heart.gif[/img]J[img]/images/graemlins/heart.gif[/img] hand on a T[img]/images/graemlins/heart.gif[/img]2[img]/images/graemlins/heart.gif[/img]2[img]/images/graemlins/spade.gif[/img] board, and discount the 9 possible flush outs to 5 outs, which seems quite drastic to me. Now I'm reading it for the second time, you don't mention the number of opponents (or pre-flop action); what did you intended? If it was heads-up and raised preflop I'd be quite happy with my number of outs; if it was limped 7-way multiway I wouldn't. What are your thoughts on this?

[Matt Flynn]

[ QUOTE ]
Maybe I've missed the PSA, but the odds table in appendix B is incorrect. I don't understand how such a thing can happen, checking e.g. wikipedia is the _least_ one can do, right? </rant>

Some minor note which I find useful because I'm a nit: the 'multiply by four' rule can be improved a bit to
p = outs * 4 - (number of outs above 8), e.g. for 15 outs
p = 15*4 - 7 = 53% instead of 60% (real number 54%).

[/ QUOTE ]


You are of course right. Divisors should be 47 and 46, not 49 and 48. This was an Excel eff up and was right when we first did it and checked it. It was my fault, not Sunny or Ed's.

Nice modification of the Four Times Rule!


[ QUOTE ]
Then a real question:

On page 12 you discuss the K[img]/images/graemlins/heart.gif[/img]J[img]/images/graemlins/heart.gif[/img] hand on a T[img]/images/graemlins/heart.gif[/img]2[img]/images/graemlins/heart.gif[/img]2[img]/images/graemlins/spade.gif[/img] board, and discount the 9 possible flush outs to 5 outs, which seems quite drastic to me. Now I'm reading it for the second time, you don't mention the number of opponents (or pre-flop action); what did you intended? If it was heads-up and raised preflop I'd be quite happy with my number of outs; if it was limped 7-way multiway I wouldn't. What are your thoughts on this?

[/ QUOTE ]

Good point. We didn't want anything complicated to confuse the discussion, because discounting outs is plenty if you don't already know to do it. Here we were indeed thinking of a multiway pot, so we also didn't count the kings and jacks.


Deeper discounting should be used when the penalty for hitting a bad "out" is worse than the gain from hitting a good out. For example, if you opponent holds a full house and you hit the flush here, you will lose more money, and maybe a lot more. But if you hit the flush, you won't necessarily make anything more from him. By discounting your outs more, you help adjust for the nastiness of hitting yet still being behind.

[bozzer]

Matt,

in the book you recommend 2/3P as a good default bet size if you're unsure about how much to bet. Thinking about just on the flop, do you like 2/3 pot as an average bet size, or should flop bet sizes tend to be a bit larger than this in typical low stakes online/live conditions?

[+EV]

[ QUOTE ]
Maybe I've missed the PSA, but the odds table in appendix B is incorrect. I don't understand how such a thing can happen, checking e.g. wikipedia is the _least_ one can do, right? </rant>

Some minor note which I find useful because I'm a nit: the 'multiply by four' rule can be improved a bit to
p = outs * 4 - (number of outs above 8), e.g. for 15 outs
p = 15*4 - 7 = 53% instead of 60% (real number 54%).

Then a real question:

On page 12 you discuss the K[img]/images/graemlins/heart.gif[/img]J[img]/images/graemlins/heart.gif[/img] hand on a T[img]/images/graemlins/heart.gif[/img]2[img]/images/graemlins/heart.gif[/img]2[img]/images/graemlins/spade.gif[/img] board, and discount the 9 possible flush outs to 5 outs, which seems quite drastic to me. Now I'm reading it for the second time, you don't mention the number of opponents (or pre-flop action); what did you intended? If it was heads-up and raised preflop I'd be quite happy with my number of outs; if it was limped 7-way multiway I wouldn't. What are your thoughts on this?

[/ QUOTE ]

If I remember correctly there is actually a modification to the 2X rule as well. You add a few percent per out based on +8 outs. I don't remember this is exactly right but I am sure it could be verified by someone else who remembers more precisely.

+EV

[boardertj]

On page 18 in the hand where you hold 8c6c on a Qh7c5d flop you say that you are 5-to-1 against hitting your straight on the next card.

When determining outs on the flop why did you only calculate your chance with only one card to come? Since there are two cards to come wouldn't you use the 4 times rule?

8X4= 32% chance of winning
We would then be sitting nearly 2-to-1 against, making it profitable to call getting 3.2-to-1 pot odds before even considering implied odds.

Obviosly it is still a call (or arguably a raise) but I was just wondering if my logic and understanding of the concept is correct.

[Matt Flynn]

[ QUOTE ]
Matt,

in the book you recommend 2/3P as a good default bet size if you're unsure about how much to bet. Thinking about just on the flop, do you like 2/3 pot as an average bet size, or should flop bet sizes tend to be a bit larger than this in typical low stakes online/live conditions?

[/ QUOTE ]


this primarily depends on how often you bluff vs. bet with a made hand on the flop.

if you bluff a lot, it also depends on whether smaller bets get respect: if they will fold to smaller bets and you bluff a lot (relative to having a made hand) then you'd go for half pot or even less.

if you don't bluff often, you mainly want to get paid off and force others to pay to draw, so you would bet more, 2/3 or higher, sometimes even > pot.

if your opponents are oblivious to your bet sizes, you just do what works in the moment.

[bozzer]

[ QUOTE ]
On page 18 in the hand where you hold 8c6c on a Qh7c5d flop you say that you are 5-to-1 against hitting your straight on the next card.

When determining outs on the flop why did you only calculate your chance with only one card to come? Since there are two cards to come wouldn't you use the 4 times rule?



[/ QUOTE ]

future betting means it is unlikely you will see the turn and river for the price you are getting given now.

actually i was going to bring this up - on one page in the book it talks about using the 4x rule in this sort of situation and I couldn't work out why - presumably the PNL advice is to not assume you'll get a free river card?

[Matt Flynn]

[ QUOTE ]
On page 18 in the hand where you hold 8c6c on a Qh7c5d flop you say that you are 5-to-1 against hitting your straight on the next card.

When determining outs on the flop why did you only calculate your chance with only one card to come? Since there are two cards to come wouldn't you use the 4 times rule?

We would then be sitting nearly 3-to-1 against, making it profitable to call getting 3.2-to-1 pot odds before even considering implied odds.

Obviosly it is still a call (or arguably a raise) but I was just wondering if my logic and understanding of the concept is correct.

[/ QUOTE ]


this is because your opponent can drop a big bet on you on the turn.

if you're calling all-in you won't face another bet and so count it as two cards to come. when there's a lot of money behind, you may not get to see that second card and so should not count it. you might allow yourself a little wiggle room for those times your opponent checks the turn, picking something between the odds of hitting in one card and the odds of hitting in two.

on the other hand, you may well want to semibluff the turn if he checks, so again you wouldn't figure it out based on having two cards coming.

[Disconnected]

[ QUOTE ]
On page 18 in the hand where you hold 8c6c on a Qh7c5d flop you say that you are 5-to-1 against hitting your straight on the next card.

When determining outs on the flop why did you only calculate your chance with only one card to come? Since there are two cards to come wouldn't you use the 4 times rule?

We would then be sitting nearly 3-to-1 against, making it profitable to call getting 3.2-to-1 pot odds before even considering implied odds.

Obviosly it is still a call (or arguably a raise) but I was just wondering if my logic and understanding of the concept is correct.

[/ QUOTE ]

The key thing is what the odds are for the next card, which is 5:1 against. If the flop bet is all-in, you would evaluate your chances of hitting with two cards to come.

When you evaluate pot equity on the flop, you use the 4x rule, but with pot odds, you're only concerned about the next card.