Awful Roy Cooke article - what am I missing?

[Clarkmeister]

In the Dec. 20 Card Player, the normally dependable, thoughtful and otherwise excellent Roy Cooke wrote what appears to be a flat out incorrect article. Not "it depends", but simply wrong.

The setup is this: In a $30-60 game he limps with a suited ace, and it comes back to him for 2 more bets in a 7 way field. The correct play is irrelevant, what is relevant is how he justifies his thought process. Here are his two general points, and I am paraphrasing. Please pick up the article and read for yourself if you want quotes.

1. When it comes back to him for 2 more bets (after he has already limped in), his EV on a fold is -$30.

This is clearly wrong. His EV on a fold is zero.

2. He can make a -EV call of $60 more dollars because the effect of that negative call is offset by the fact that he increased his EV of his initial $30 from zero to something more than zero by staying involved in the hand.

This is wrong again. If the call is negative for both this hand and his overall range of hands, then he should fold. There is no such thing as using sunk costs to offset the negative cost of a current play. His initial $30 limp is totally irrelevant when deciding whether or not to continue.

Anyways, I was surprised to see something that appears so blatantly incorrect come from the esteemed Mr. Cooke, particularly considering the high regard that Sklansky holds him in. Am I missing something here? Did I misread what he was trying to say?

Thoughts?

[Rick Nebiolo]

Clark,

I also found this article convoluted and confusing but perhaps wrong in a different way than you.

You wrote paraphrasing Roy: The setup is this: In a $30-60 game he limps with a suited ace, and it comes back to him for 2 more bets in a 7 way field. The correct play is irrelevant, what is relevant is how he justifies his thought process. Here are his two general points, and I am paraphrasing. Please pick up the article and read for yourself if you want quotes.

1. When it comes back to him for 2 more bets (after he has already limped in), his EV on a fold is -$30.

You: This is clearly wrong. His EV on a fold is zero.

Me: Except for the softest games, the initial early position limp with a suited ace probably costs most players about a quarter of a small bet. This is huge if you consistently make this mistake. But let’s say Roy judged the game sufficiently soft so the initial limp was slightly positive in expectation.

Now the pot is uncharacteristically raised and reraised; however, he has six opponents. With six opponents in for two more bets (he should be getting about 10 to 1 on his additional money) calling is probably the right play, but not by much. If I’m right, folding would have an EV of about -$3 or so. The only way his EV would be zero as you claim would be if his decision were exactly break even (call or fold). But Roy is wrong to say that his negative EV for folding is the full bet already committed (-$30).

You paraphrasing Roy again: ”2. He can make a -EV call of $60 more dollars because the effect of that negative call is offset by the fact that he increased his EV of his initial $30 from zero to something more than zero by staying involved in the hand.

You: This is wrong again. If the call is negative for both this hand and his overall range of hands, then he should fold. There is no such thing as using sunk costs to offset the negative cost of a current play. His initial $30 limp is totally irrelevant when deciding whether or not to continue.

Me: I strongly agree with you regarding the irrelevance of factoring in his initial limp. Whether to call two additional bets with six opponents is probably a close decision, which means it isn’t that important (if I’m right that it is a close decision). The EV of close decisions is approximately zero.

”Anyways, I was surprised to see something that appears so blatantly incorrect come from the esteemed Mr. Cooke, particularly considering the high regard that Sklansky holds him in. Am I missing something here? Did I misread what he was trying to say?

Roy might have gotten complacent after the praise his previous article received. I think you have it nailed except for the point I made above.

Regards,

Rick

[BruceZ]

If I’m right, folding would have an EV of about -$3 or so. The only way his EV would be zero as you claim would be if his decision were exactly break even (call or fold).

That's wrong. The $30 in the pot has no impact on the ev of a fold. The ev of a fold is always 0 times the probability of winning 0 which is 0*1 = 0. You are computing the ev of putting $30 in the pot and then folding to a double raise, which has no relevance to the current decision since the $30 is already committed. All that matters is whether calling has a positive ev.

[Rick Nebiolo]

Bruce,

I’m not an expert as I should be on the terminology but I’ll try to clarify my point.

You wrote: ”All that matters is whether calling has a positive ev.”

I agree completely with this statement. When I said folding would have an EV of -$3 or so I mean calling is worth about $3 of positive expectation (beyond the cost of the call). If one folds when a call is worth $3 one would forgo that $3 in expectation.

Regards,

Rick

[Ed Miller]

I remember, when reading the article, thinking that his logic was convoluted and distinctly different than the standard "calculate the EV of each decision" logic that you outline. I just assumed at the time that he was solving the same problem and coming to the correct conclusion using a different, more confusing, method. I didn't have the patience to wade through it, so I just ignored it...

The reason I mention this is that I went back to read the article on the website to wade through it... and it isn't there. The rest of the CP issue is all there, but Cooke's article is not... strange...

[AceHigh]

"I mean calling is worth about $3 of positive expectation (beyond the cost of the call). "

I'm not sure I understand this, are you saying calling has an EV of -$57?

I'm pretty sure this call is -EV.

[AceHigh]

"Whether to call two additional bets with six opponents is probably a close decision,"

Aren't you afraid of being dominated by both a big Ace and a big pair? I would be worried that say AK and JJ are both out there and now I need to make 2 pair, trip x - as in Axs or my flush to win.

A6s wins about 13.6% of the time vs. AKo and JJ, splits .4% (used TTHv5 to come up with numbers, just running to showdown). With 4 more opponents cutting into your EV, it seems to me folding is the best option.

[Rick Nebiolo]

It seems many articles and writers are archived but not all are all of the time. They probably do this to encourage subsripctions or later book sales. I believe other online versions of magazines might have a similar policy e.g., The Atlantic Monthly has an online version but it isn't always complete nor are all archives available.

~ Rick

[Rick Nebiolo]

AceHigh,

I wrote above: "Whether to call two additional bets with six opponents is probably a close decision,"

You wrote: “Aren't you afraid of being dominated by both a big Ace and a big pair? I would be worried that say AK and JJ are both out there and now I need to make 2 pair, trip x - as in Axs or my flush to win.”

I’m sure Roy would plan to get away from the hand if he flopped a lone ace and faced heat. He surely understands once his initial call is raised and reraised the primary value of his hand comes from making a flush and to a lesser extent the two pair and trip hands you describe. With six opponents, he has to pay two more bets and will end up with a pre flop pot of about 21 bets (I don’t think Roy mentioned whether the blinds folded or called). That is getting about 10 to 1 on his additional money and clearly he flops enough flushes and flush draws to make this call worth it.

”A6s wins about 13.6% of the time vs. AKo and JJ, splits .4% (used TTHv5 to come up with numbers, just running to showdown). With 4 more opponents cutting into your EV, it seems to me folding is the best option.

Run your simulation with the A6 suited, the dominating ace, the pair of jacks, and four random hands. That would more clearly demonstrate the value of Roy’s hand. The extra opponents don’t cut into Roy’s EV, they contribute to it when he makes flushes. But this isn’t what the thread is about. Let’s agree that it is a close decision (relative to the size of the bet).

Regards,

Rick

[Rick Nebiolo]

AceHigh,

My wording “beyond the cost of the call” is confusing. I’ll try again.

It costs Roy $60 to call two more bets. I’ve argued it is a close decision whether he should call or not but it appears to be slightly positive (see my posts above, below, and to the side [img]/forums/images/icons/grin.gif[/img] ). Relative to a $60 bet, $3 of positive (or negative) expectation would seem to indicate a close decision.

If I’m right and calling $60 more has a positive expectation of $3, then Roy figures to win an average of $3 in this pot in the long run FROM THE POINT OF HIS DECISION to call two more bets. If he folds, he forgoes that $3 of positive expectation. To my way of thinking, forgoing a play that has positive expectation is a decision with negative expectation.

Some might ask: “If he folds what about the $30 he already put in the pot? How does that factor into his expectation.?”.

My answer: The money he already put into the pot doesn’t factor into the expectation of his SUBSEQUENT decision (whether to call two more bets). Think of this money as you would a blind bet.

If Ray had a crystal ball (or he knew the game was aggressive) and he could see he would be paying three bets to see the flop and end up with six opponents, then his initial decision to enter was a mistake (I’ll estimate an EV of about -$15 or so). But in another post I assumed this hand was an aberration for this game and Roy was probably correct to enter.

You know, this one time I really wish we had QuadNines/Mark Glover/Mark Glover’s Editor around to straighten this one out [img]/forums/images/icons/grin.gif[/img]

Regards,

Rick