Are The Critics Right?

One of the criticisms of our book, "Middle Limit Holdem Poker", is that we are too loose in calling raises from the big blind. We advocate calling an early position raiser with any pair, any suited ace, or any two unsuited cards higher than a nine like jack-ten offsuit. We have been told by more than one expert that with only one or two players in the hand, this should prove expensive.


I used an all-in poker simulation tool similar to poker probe. I ran 3 cases. Case one was when you have a small pocket pair like deuces. Case two was when you had a suited ace like ace-deuce suited. Case three was when you had two decent unsuited cards like jack-ten offsuit. I assumed heads-up play between the big blind and an early raiser. I assumed the early raiser had a pocket pair higher than nines, ace-king, ace-queen, ace-jack suited, or king-queen suited. For each case I ran the simulation and used a weighted average based on the number of hands for the raiser to have. For example, consider the case with the pocket deuces. There are 16 ways for the early raise to have AK and only 6 ways for the early raiser to have AA. Obviously the AK situation got weighted proportionally more than AA. I came up with a weighted average for each. Here are the results:


For Case One (pocket deuces), you will win 38% of the time so your odds are 1.6-to-1 against.


For Case Two (ace-deuce suited), you will win 31% of the time so your odds are 2.2-to-1 against.


For Case Three (jack-ten offsuit), you will win 30% of the time so your odds are 2.3-to-1 against.


Now your current pot odds are 3.5-to-1. Of course, a skeptic would scream: "But your simulation assumes you get to see all 5 boardcards for free, so it is invalid!"


"BUT IF AN ALL-IN SIMULATION HEAVILY FAVORS CALLING OVER FOLDING, IT IS UNLIKELY THAT THE IMPACT OF ADDITIONAL BETTING ON THE LATER STREETS WOULD CHANGE THE INITIAL DECISION FROM A CALL TO A FOLD, IT WOULD SIMPLY MAKE IT CLOSER" according to one expert.


I believe there is a fairly large overlay when you compare your current pot odds with your "all-in" drawing odds. What does everyone else think?

"There are 16 ways for the early raise to have AK and only 6 ways for the early raiser to have AA. Obviously the AK situation got weighted proportionally more than AA"


no weight to all raising possibilities? how about KK to TT also. along with AQ..


i think you also should factor in the type of player. that narrows down a decision even more precisely.


is using an allin format really going to make the results accurate? i think this would differ quite a bit if you went street by street. again, i think you have to take the other player into consideration. against certain players, you can get trapped with a piece of the flop on a lesser hand. then you may wish you never called in the first place.


not sure if this was what you were looking for...


oh well... i tried /images/smile.gif


b

I think the all-in sim is going to be totally useless here.


Just consider the 2-2 on the river only. I guess you advocate calling all the way to the river to find out if raiser has overcards. On the river AK is going to check and a big pair is going to bet. So you lose an extra big bet on the river when you lose compared to what you get when you win. That is quite a big difference when you calculated your pot odds based on only 1.75 big bets at the start.


When you win with the 22, it is going to be hard to get to the river, let alone win the same amount he will win with a solid pair.


Another big factor with these calls, is that a suited hand will frequently give you a big enough hand to semi-bluff back with, JTo won't connect enough for these plays.


I like the calls with the suited A generally, unless it is some kind of great player that you just don't want to mess with.


D.

>>There are 16 ways for the early raise to have AK and only 6 ways for the early raiser >>to have AA


I hope you made adjustments to this after looking at your own hand.

(e.g. Axs)


And after looking at the Flop

I haven't read your book yet so I'm basing my response soley on your post.


Case 1: Pocket 2's (or other small/medium pocket pairs).


I'm going to assume that the UTG raiser will bet when checked to regardless of whether he has a big pocket pair or a Big Ace/KQs. I think this is reasonable for most middle limit players in this situation.


When you don't make a set, you are going to have a difficult time playing this hand heads-up. Your current pot odds may be 3.5:1 but your effective odds aren't nearly as good. If you call on all streets, you are going to have to put in 3 big bets in order to win 4.25 big bets (1.75 big bets in pot pre-flop plus additional 2.5 from opponent post-flop). So, your effective odds are 1.42:1. You calculated that your odds of winning were 1.6:1 so the odds are somewhat close. Maybe it becomes close to even or profitable when you consider the times when you flop a set and don't have to call a river bet to someone who doesn't bet an unimproved big Ace.


Still, you clearly aren't getting the pot odds to flop a set. Your implied for flopping a set aren't great since you have only one opponent to pay you off. And, the effective odds make calling down with the unimproved pocket pair a losing play. Those all sound like reasons not to call with these hands.


Case 2: A2s or other Axs


I think this is a bigger mistake than calling with a pocket pair. What are you hoping to flop? If you flop top pair, you can be dominated by AK, AQ, or AJ and may feel compelled to pay off all the way to the river (or at least your readers will). If your opponent has KK or a smaller pocket pair, he may not pay you off with an Ace on the flop. So, it seems you will pay off with the worst hand but will not get paid when you have the best hand. If you are willing to fold top pair/weak kicker, then why play the hand in the first place? For the flush potential? The flush won't come often enough to make this call profitable.


Case 3: JTo and other big offsuit cards


I was surprised to read this in your post. I think folding AJo, ATo, KJo, KTo, QJo, QTo, and JTo to "legitimate" EP raises is very easy. AKo is a 3-bet. AQo is a call. KQo is a tough decision but it's probably best to fold it.


The possibility of making second best hands here is very high. I'm not sure what you're looking to flop. Top pair seems like a trap hand most of the time.


Overall


The biggest problem with the computer simulation isn't that it assumes you get to see all five board cards. The big problem is that it doesn't account for the extra bets your opponent will win from you when he is ahead since he has position on you. And, it doesn't account for the bets your opponent will save when you out-flop him.


I see players following your advice getting into trouble by paying off better hands but not getting paid themselves when they outdraw their opponent.

<BLOCKQUOTE>We advocate calling an early position raiser with any pair, any suited ace, or any two unsuited cards higher than a nine like jack-ten offsuit. </BLOCKQUOTE>


This is a big range of hands, and I thought that conventional wisdom says to avoid trap hands like KJ and QJ. My hunch is says JT is a safer hand to play.


Second, your UTG player is tighter than many. There are plenty of halfway decent players who will also play hands like AJo, KQo, and ATs for a raise. I'm not sure how including those hands will affect the value or JT (or QJ or KJ) in the blind.


I would be curious to see how much your values change when you adjust your playable hand upward (like for QJ and KJ), and how much they change when UTG gets looser. If they don't change much, then you will have a lot more confidence in the numbers.

There will also be times when your odds improve due to implied odds on future betting rounds.

AK is an underdog to any pair including 22 hot and cold, but you would certainly call with that. Your odds improve since you can get away from it easily when you miss the flop.

The biggest problem with the computer simulation isn't that it assumes you get to see all five board cards. The big problem is that it doesn't account for the extra bets your opponent will win from you when he is ahead since he has position on you. And, it doesn't account for the bets your opponent will save when you out-flop him.


I see players following your advice getting into trouble by paying off better hands but not getting paid themselves when they outdraw their opponent.


I think these are very good points. It takes a lot of skill to play marginal hands out of position.


If the EP player is predicatable, I think you could play the recommended hands. If he's a tricky player, it would be difficult.


"If your opponent has KK or a smaller pocket pair, he may not pay you off with an Ace on the flop"


If he doesn't pay you off on when an Ace flops with these hands, then all the recommended hands will beat these hands whenever an Ace flops (assuming BB bets them).

Yes, I did. To quote from my lead post: "For example, consider the case of pocket deuces. There are 16 ways for A-K and only 6 ways for AA."


Now in the case of ace-deuce suited there would only be 12 ways for AK and only 3 ways for AA.


The flop was not considered because this was run as a preflop "all-in" simulation.

I think a major factor would be how well you play vs. the EP raiser, since skill is such an important factor in heads-up play. If you play better then EP, then I think it would be profitable to play the recommended hands (although hands like 22, 33, A2s, A3s are very difficult to play). If you play worse, you probably want to fold all except premium hands. If evenly matched, maybe somewhere in between (???)


It's a good question!

What if you put a bet in on every round? Now getting only a 30 percent return is a big losing proposition. How does this impact your decision?

Consider case one. If the board comes Ace-high or King-high, I will check-fold when the flop gets bet unless I flop a set so I don't get trapped for additional bets on later streets. If the board flops all small cards, then I will definitely play since my hand will be best most of the time.


With ace-little suited, I may bet the flop or check-raise the flop if it comes ace-high. If I continue to get played with on the turn, I will usually give it up and not pay off all the way to the river.


The key is learning to play well once the flop comes. It is not true that you will simply pay off all the way. Your winning chances are reflected in the simulation run. The concern is how much additional money do you lose when you lose versus how much you win when you win. The critical considerations are how well you play once the flop comes and how much money is already in the pot before you get involved.

"The concern is how much additional money do you lose when you lose versus how much you win when you win."


But when you automatically fold with an ace or king high flop you are essentially allowing your opponent to successfully bluff you. That costs money as well.

I will only put in a bet on every round if I obtain an improvement. If I obtain an improvement, my winning chances will frequently be better than 30%. In case one with the pocket deuces, I will only play if I don't flop a set or no ace or king flops. In case two with ace-little suited, I will only play if I flop top pair or better or I flop a decent draw. Similarly with case three having jack-ten offsuit.

I am assuming that my opponent has AA,KK,QQ,JJ,TT,AK,AQ,AJ suited, or KQ suited. If the flop comes ace-high the only hand he can bluff me is KQ suited. If the flop comes king-high, then the only hands he can bluff me are AQ and AJ suited.

If you're getting 3.5-1 when you're a 1.6-1 underdog, then you're making .9/2.6 bets preflop. If you put in a bet on every round that's another 2.5 bets which would have to lose this same .9/2.6 bets or 13% to break even.

It's another 5 bets, that you would put in, so these would have to lose 6.9%.

"Consider case one. If the board comes Ace-high or King-high, I will check-fold when the flop gets bet unless I flop a set so I don't get trapped for additional bets on later streets. If the board flops all small cards, then I will definitely play since my hand will be best most of the time"


player dependent, and if any paint hits, your subject to a bluff. and you could find yourself in a guessing game...what if the flop come J or T high? could this player raise with ATs or AJ? know your player..


"The critical considerations are how well you play once the flop comes and how much money is already in the pot before you get involved"


you didnt add flop play to your sim did you? i thought it was an allin sim


some of the hands neither will win much, since it predicates on hitting your hand on the flop. say if you flop a set with low cards on the board, and the other guy has AK. say you c/r or bet out the flop, he may not be in unless he improves on the turn. depending on his read and knowledge of you. i believe your the one with more risk here. bad position against a guy who raised the field.


if the table was tight, some of these chips may be tough to recoup. if its a looser, better game, theres better situations to get into.


personally, i like at least 1 more caller before i call with a low pkt pair in BB to an EP raise.


b

See above, I put it in the wrong place. Betting on every round must lose no more than 6.5% to break even. Of course it will lose more than this if you always bet every round, the question is can you achieve this 6.5% loss or better for every 5 bets by choosing when to bet. Then if you don't bet you will also lose some of your initial preflop advantage the times when you would have drawn out, so this must be canceled too.

It doesn't seem like the all-in sim will be very useful given the way you will be playing these hands.


D.

I have always found your posts here to be interesting and helpful and would love to read your book. Is your book available on a website or will I be able to find it at a bookstore?


Thanks

http://www.diamondcs.net/~thecoach/ is Bob Ciaffone's site. He likes you to books there, and will autograph them and reply to email questions about the book. He doesn't take credit cards, though.


I didn't see any site for Jim.


There's many places on the web that sell the book, which you can easily find by using Google or some other search engine.


I like the book very much. It's done a bit different than HEFAP21 (which I also like very much), which discusses fewer situations but in a lot of detail. MLHE has lots and lots of problems and short answers.

The reason that the "all-in" simulation is useful is that it gives you a feel for what your initial overlay is (that is how much greater your pot odds are compared to your drawing odds). If the overlay is high enough, you should be playing despite the fear of domination and so forth. I believe David Sklansky has put up some posts on this (e.g.- the "7-3 suited problem" and so forth).

I'm short of time at the moment, but this is an area where much discussion can happen. In fact, the important ideas here have come up on these forums many times.


The real question should be "Does this simple mathematical model represent hold 'em well enough to draw valid conclusions from?"


In our stud book, we did draw conclusions from simulations like this. We have a whole appendix devoted to it. The reason we were able to do so is that in stud it is frequently correct to chase. So assuming that each player puts a bet in on every street can often lead to very good insights on proper strategy, and these insights can then be reasonably adjusted for some obvious events (like the high card pairing it's doorcard).


However, David and I believe that hold 'em is a very different matter. We have warned over and over that these type of simulations do not give valid advice for playing hold 'em. If we would have thought that they would have, we would have used them.


Your small pair example is a case in point. First, I believe that calling a legitimate raise out of the big blind when no one else is in with a small pair is a very close play. But you say it is 38 percent which would not be close. But upon further examination we discover that you are frequently folding on the flop unless you make a set. So what does this do to the 38 percent.


First, you will be folding some hands that will win unimproved, and second, you will be folding some other hands that will make a set on either fourth or fifth street. This will certainly lower your overall win rate to a figure far smaller than 38 percent. My guess is that your true win rate is between 20 and 25 percent. Now 3.5-to-1 looks approximately break even. (Note: I am not saying that you are playing the small pair incorrectly. In fact, I would probably play it similarly.)


And yes, for all the other readers, I am one of the critics of your calling strategy out of the blind in this situation as being too loose. Anyway, there is much more to discuss on this topic and hopefully others will join in.

(n/t)

In the 7-3 problem there were a few differences:


- suited: you can more frequently go past the flop on those hands that you might win with.


- 3 players: you will more likely have to hit the flop some to win. Actually your plays are probably a lot better with 3 then 2.


D.

Excellent points. Actually, the 7-3 suited problem was a "wake-up call" for me on the value of computer, all-in, simulations. Perhaps it does not apply here with only one opponent. I wonder what the impact would be if you added some dead money to the pot. Suppose an established player had returned to the game and posted both blinds. Further suppose that this player folds when the early player open-raises preflop. Now your initial pot odds have grown from 3.5-to-1 to 5-to-1. Does this make calling the raise from the big blind with pocket deuces, ace-little suited, and jack-ten offsuit correct?

The more likely you are to go to the raiver, and the more likely the action on each street is bet call as opposed to some raising or no betting at all, the better the information you can obtain from "hot and cold" simulations.

I have the book and have found it very helpful. You can order a copy from Conjelco (http://www.conjelco.com/), which is an online gambling bookstore with a very good selection of poker books.