Theory Discussion of a Point From "Poker Tournament Formula"

[SplawnDarts]

There's been a long thread in "publications" about PTF, and the correctness of numerous theory assertions contained therein. I want to look at one of them in depth.

Background: A key concept explained in HoH is that there are cases where variance is good for you. Specifically, when in the "red zone" with an M less than 5 or so, you want to either get healthy or go broke. Festering at a low M accomplishes nothing and you will not likely survive long enough to get a big enough hand to have a big edge. Hence Harrington advocates getting all-in with almost any two when you can be first in while in the red zone.

The question: PTF puts forward a similar concept, namely that in fast MTTs, early variance is good for you (although it's not stated quite that way). You want to get a big stack or get out. PTF says to get aggressive faster in fast tournaments regardless of M, presumably to get that variance. Is this correct?

The Example: To accomplish this, PTF advocates plays such as preflop button calls during early levels with any two - a play that is clearly normally cash-game -EV. However, it is high variance.

The positions: There are several positions one could take on what Snyder has to say. I've named the ones I think are of interest for ease of discussion:

P1: Snyder is full of it. When M is big, chip EV is closly related to tournament EV, so play for +CEV.

P2: Synder is correct for the reasons stated in PTF - you want to go big or go home.

P3: Snyder is correct, but for the wrong reason. The reason to enter preflop with any two is that your opponents are poor postflop, so you actually have an edge despite weak cards. This is the "smallball" argument.

Ground rules: Be clear if you're talking about expected tournament value (TEV) or expected number of chips (CEV) when discussing expectations.

FWIW, I believe either P1 or P3 is correct depending on the situation. I'll expound later.

Discuss away!

[jeffnc]

[ QUOTE ]
The question: PTF puts forward a similar concept, namely that in fast MTTs, early variance is good for you (although it's not stated quite that way). You want to get a big stack or get out. PTF says to get aggressive faster in fast tournaments regardless of M, presumably to get that variance. Is this correct?

[/ QUOTE ]

That was not my understanding from reading the book exactly, but I did get the impression he values a large stack more than Harrington does and is willing to accept more variance to get it. Or by variance do you mean variance in your place of finish? Because then I think it's pretty easy to show that a lot more bustouts is worth it for a few higher finishes. But that's true regardless of the blind speed.

I thought the main point was that in a very fast tournament, NOT playing aggressively is what's risky simply because the quickly increasing blinds pose more of a threat than most players realize.

[smbruin22]

i think snyder has the call any two from the button due to the post-flop position (so post-flop play) not due to the variance necessarily.

but from reading the book, i don't think he recommends you play these systems exactly ... he has a position system, a card system and a stack size system... one thing i was never completely sure about was how you were supposed to integrate these systems. i remember "look at what weapons you have", but i didn't think that was clear enough.

i wish people would stop citing that one strategic tip. he's assuming deep stacks near the start. and this isn't a book like kill phil is perceived to be. it's not a system. i think he assumes (or you should) that you have played some poker already. if you haven't, i would take out some of the position plays.

say what you want, but PTF woke me up to aggressive small-ball play as much as any other book...

[SplawnDarts]

[ QUOTE ]
i think snyder has the call any two from the button due to the post-flop position (so post-flop play) not due to the variance necessarily.

but from reading the book, i don't think he recommends you play these systems exactly ... he has a position system, a card system and a stack size system... one thing i was never completely sure about was how you were supposed to integrate these systems. i remember "look at what weapons you have", but i didn't think that was clear enough.


[/ QUOTE ]
Yeah, that was less than helpful.
[ QUOTE ]

i wish people would stop citing that one strategic tip. he's assuming deep stacks near the start. and this isn't a book like kill phil is perceived to be. it's not a system. i think he assumes (or you should) that you have played some poker already. if you haven't, i would take out some of the position plays.


[/ QUOTE ]
I think it IS intended to be a system, both based on the book title and on Snyder's blackjack background and outlook on the game. I could be wrong, though. Maybe Snyder will answer if he was trying to provide general strategic ideas or a specified system.
[ QUOTE ]

say what you want, but PTF woke me up to aggressive small-ball play as much as any other book...

[/ QUOTE ]

That's good, I guess.

[SplawnDarts]

[ QUOTE ]

That was not my understanding from reading the book exactly, but I did get the impression he values a large stack more than Harrington does and is willing to accept more variance to get it.


[/ QUOTE ]
I meant chip count variance.
[ QUOTE ]

I thought the main point was that in a very fast tournament, NOT playing aggressively is what's risky simply because the quickly increasing blinds pose more of a threat than most players realize.

[/ QUOTE ]

This is the part that a lot of people disagree with.

Perhaps the problem should be restated. If, at the start of a tournament, you could flip a coin and either get an additional 1/3 of your chip stack or loose that amount, would it be +TEV to do so? I believe the answer is provably "no".

However, if you had an M of 3 late in the tournament and were offered a flip for 3M worth of chips, you would probably take it.

[alanbrown]

I heard a lot about Snyder before I read his book (currently half way through). Actually, it was the fact that Sklansky seemed so obsessed with him on this site that made me curious about what he had to say.

And I'm liking his book a lot. His concept of measuring the quality of a tournament by its patience factor is the single most valuable things I've taken from a poker book in a long time. It's one of those things that makes you go huh! how come no-one did this before?

However, I digress. The topic under consideration is whether his plan of pushing harder, earlier in 'faster' tournaments is a good idea.

I think it is.

Obviously in a no blinds game everyone should be ultimately tight. And in a game where everyone plays for their whole stack on every hand nobody should ever fold. Snyder has found a methodology for measuring how far along that tight/loose spectrum one should play.

His point, in a nutshell, is that how loose you should play isn't just a function of the current blind level but also how soon those blinds will go up. An extreme example to illustrate would be to talk about how one's strategy should change if the blinds doubled every orbit. You couldn't then just treat each blind level as an independent decision. The fact that the next one is upon you immediately means you must stay ahead of that breaking wave or get crushed in it.

I'm not sure I agree with Snyders exact conclusions. He considers being <30BB to be short stacked and that you're now looking at being an All In player. That seems a little extreme to me. But his idea of linking strategy to the Patience Factor of the tournament is absolutely the right idea.

As for the question of stack utility, I'm unsure about that one. I agree that I wouldn't flip a coin at the start of a tourney for a third of my stack. The reason being that I believe that I have a >50% chance of increasing my stack by 1/3 by playing pokah.

I agree that Snyder's book is definitely a system also.

I'm not sure how I feel about chip utility. I think it's more relevant in live games than online games where you can just fire up another tourney. But in live games the ability to cover your opponents AIs does sound like it would make them steer clear of you a little more.

[SplawnDarts]

[ QUOTE ]

As for the question of stack utility, I'm unsure about that one. I agree that I wouldn't flip a coin at the start of a tourney for a third of my stack. The reason being that I believe that I have a >50% chance of increasing my stack by 1/3 by playing pokah.

[/ QUOTE ]

The concepts are intertwined, though. If you wouldn't want "artificial" variance at the start of a tournament, there's also no reason to want variance brought on by the poker itself unless it also comes with +CEV. I think you got to the heart of it - you believe you can do better than that by playing - ie. you're looking for +CEV because you know it's +TEV as well. That's sound thinking.

Now, many of the plays Snyder recommends are likely -CEV (since they would be cash game losers) but are hugely +CVAR.

As such, I believe that it is provable that the only reason you would make such plays is if you felt +CVAR -> +TEV. And the fact that you wouldn't flip the coin indicates that you don't think that's true. I agree with you about the coin (ie. I believe +CVAR -> -TEV when stack are deep), and I believe that it follows directly that Snyder's advice is wrong anytime his wacky loose plays are -CEV, which is probably most of the time.

As such, I think either P1 or P3 can be correct, but P2 cannot be.

[alanbrown]

[ QUOTE ]
As such, I believe that it is provable that the only reason you would make such plays is if you felt +CVAR -> +TEV

[/ QUOTE ]

Can you clarify what you mean by +CVAR. Does it have the same units as TEV? I ask because you're comparing the 2 of them and I'm confused as to if that can be done.

[SplawnDarts]

[ QUOTE ]
[ QUOTE ]
As such, I believe that it is provable that the only reason you would make such plays is if you felt +CVAR -> +TEV

[/ QUOTE ]

Can you clarify what you mean by +CVAR. Does it have the same units as TEV? I ask because you're comparing the 2 of them and I'm confused as to if that can be done.

[/ QUOTE ]

Sorry - VAR is my abbreviation for variance. EV is my abbreviation for expected value. A C prefix means in units of tournament chips. A T prefix means in units of cashout $$$. a '->' should be read as 'implies', not 'equals'.

So the part you quoted would read

[ QUOTE ]
As such, I believe that it is provable that the only reason you would make such plays is if you felt an increase in variance (in units of chips) implies a larger expected cashout

[/ QUOTE ]


Does that clear things up?

[WRX]

[ QUOTE ]
The question: PTF puts forward a similar concept, namely that in fast MTTs, early variance is good for you (although it's not stated quite that way). You want to get a big stack or get out. PTF says to get aggressive faster in fast tournaments regardless of M, presumably to get that variance. Is this correct?

The Example: To accomplish this, PTF advocates plays such as preflop button calls during early levels with any two - a play that is clearly normally cash-game -EV. However, it is high variance.

The positions: There are several positions one could take on what Snyder has to say. I've named the ones I think are of interest for ease of discussion:

P1: Snyder is full of it. When M is big, chip EV is closly related to tournament EV, so play for +CEV.

P2: Synder is correct for the reasons stated in PTF - you want to go big or go home.

P3: Snyder is correct, but for the wrong reason. The reason to enter preflop with any two is that your opponents are poor postflop, so you actually have an edge despite weak cards. This is the "smallball" argument.

[/ QUOTE ]

Splawndarts, as I'll explain, I think that you, as well as others who have responded in this thread, have the premise of Arnold's argument only partially correct. The parts of his argument that have become so controversial are a direct reaction to the advice given, among other places, in David Sklansky's Tournament Poker for Advanced Players (2003), pp. 19-20: "The [tournament prize structure] factor (which would indicate that you avoid close gambles with a slight advantage, to keep from going broke) only takes on significance near the very end of a tournament. But there is another reason to eschew close gambles even early on. This reason has nothing to do with the prize structure, and in fact is something you should be aware of even if the tournament paid only one winner. What I am speaking of involves the presumption that you are one of the best players in the tournament. That being the case, you should avoid close gambles, especially for large portions of your chips." David goes on to explain that this principle is a consequence of the likelihood that you will have better opportunities as the tournament progresses, and the fact that if you lose a close gamble now, and are knocked out of the tournament, you will not be able to take advantage of those better opportunities later. Both David and Mason have given similar advice in other writings.

I believe this advice is sound, as applied to the situations David had in mind when he wrote it. The principle is somewhat challenging to put into practice, as it is difficult to quantify what amounts to a gamble that is so "close" that it should be avoided. Nevertheless, the reasoning is good.

Arnold wrote in The Poker Tournament Formula of the need to play aggressively in fast tournaments, and to work to build your stack early. Now, he wrote PTF as a very practical work, and did not give lengthy explanations of the theories behind his advice. However, it appears from what he did have to say, and what he has since written at Blackjack Forum Online, that his advice to depart from the above-quoted principle, and to play faster in fast tournaments, was for the following reason. David and Mason's advice to eschew close gambles is premised on the likelihood that you, as one of the better players in the tournament, will enjoy a number of better opportunities in future hands, before the blinds increase to the point that they are eating you alive. However, in a fast tournament, the number of hands you will get to play before you face that extreme pressure will be very much less than it would be in a traditional tournament with long blind levels. That means that the probability of you seeing future better opportunities is severely reduced. Therefore, if you encounter an even slightly +CEV situation now, it is likely to be the best you will ever get. In a fast tournament, you need to exploit every advantage, even the small advantages--and the faster the tournament, the truer this is. If the tournament gets to the point where everyone has very low "M", it's nothing but a crapshoot, and you will no longer be able to gain an edge through your superior skills.

In short, the faster the tournament, the less of an edge the skilled player has over the field. However, to maintain even a small edge, the skilled player needs to exploit every available advantage.

Responding to Mason's challenge in another thread, I would like to suggest some hands that should be played differently in slow and fast tournaments. This will have to be brief, because I'm about to go away for a few days, and I won't be able to follow up during that time. The differences in tactics are matters of degrees, perhaps small degrees. They won't mean a totally different approach to the play of a hand. Instead, they will mean that a close decision should tip one way or the other.

Examples could be taken directly from David's TPFAP . At pp. 20-21, he suggests that if an opponent moves all in with what you know to be A-K suited, you should seriously consider throwing away 4-4, which has only a small advantage. However, in a fast tournament, it might be correct to call with 4-4, given a low probability of seeing a better opportunity any time soon. David also suggests that when pot odds make calling a bet with a drawing hand only marginally +CEV, it is probably correct to throw away the hand. However, in a fast tournament, one may need to grab that small advantage.

Other examples involve close decisions as to whether to call or fold on the end--perhaps holding second pair and a mediocre kicker, when there has been little action in previous betting rounds after the flop. If you estimate that a call for a big part of your stack would have only a small +CEV, given what you believe to be the odds of you having the best hand, it may be correct in a slow tournament to fold. However, in a fast tournament, you may be compelled to take advantage of that +CEV situation, having little hope of finding better opportunities later.

It is true that Arnold does not speak of many of these situations, because PTF doesn't devote a lot of space to post-flop play. However, the reasoning given above tends to justify his advice advocating aggressive pot-stealing before the flop, and aggressive bluffing and semi-bluffing on the flop, more aggressive than might be correct in slower tournaments.

Some of the messages here assume that Arnold's advice to play faster in fast tournaments is based entirely on the chip utility principle he has espoused. Briefly, this holds that by maintaining a comparatively large stack, a skilled player remains a "full featured poker player," able to engage in a variety of tactics and most effectively exploit his skill advantage over opponents. Short-stacked play, in contrast, is relatively simplistic, and offers less opportunity for capitalizing on edges. Therefore, a "coin-flip" gamble can in fact be +$EV, because if successful, it leads to favorable opportunities later.

Note that the chip utility principle, if valid, is an idea distinct from the considerations discussed above, which independently indicate exploiting marginal +CEV situations in fast tournaments. Also, the chip utility effect is not unique to fast tournaments, or to tournaments in general. To the extent it is real, it comes into play in cash games, too. The one reason a chip utility effect might justify especially aggressive "coin-flip" tactics in a fast tournament is that the pressure of the rising blinds will be forcing other players into confrontations with each other, so that other players will be doubling through, and you may want to keep up with their stacks.

Arnold didn't speak much of chip utility in PTF . It was only later, in writings at Blackjack Forum Online, and to a small extent here, that he discussed his chip utility theories at length. I frankly think that he ran into trouble by overestimating the spectrum of situations over which a positive chip utility effect exists, and by failing to recognize how the percentage payback prize structure distorts correct tactics as one's stack becomes very large, and as one approaches reaching a paying finish position. For anyone who may be interested, I've previously commented on this subject at length both here and at Blackjack Forum Online.

The PTF 's suggested play of calling on the button with any two cards, with the intention of bluffing on the flop, has been given more attention by some commenting here than is justified. That play is a very small part of the advice contained in the book. It is not a play that Arnold has advocated making at every opportunity. Obviously it is a pure position play, and should be used with discretion, based on how other players have demonstrated they are likely to respond. Arnold suggests using this play for the purpose of accumulating chips, when it appears to be a +CEV maneuver, not as a reckless gamble.

The arguments between Arnold and Mason (and David) have centered on two points--the effect, if any, of tournament speed on strategy, and the effect of the percentage payback prize structure on strategy. (Unfortunately, they have also descended to personal attacks.) I think that Mason was straightforward in his review as to why he thought PTF was a good book that would help many players. His criticisms of Arnold on certain points of theory did not call into question the validity of the book's practical play advice. Possibly with minor exceptions, I don't understand Mason to have said that he considers any of that advice to be faulty.