I was watching one of Matt Hawrilenko's lectures at MIT and he said this twice at different points in the talk to really emphasize it. How true is this and if you followed it how far can it take you?

the video is here: https://youtu.be/kn92WXcKr0M

The full quote is: "Let me be very clear on how important this concept of knowing your own hand and knowing where you are in your distribution is. I think you shouldn't think about anything else in poker until you have bought and paid for a house by knowing where you are in your own distribution and shaping it to be balanced." and is at https://youtu.be/kn92WXcKr0M?t=2177

Your paraphrasing of what he said really cuts out the meat of the concept imho.

^^^

Thank you for doing that, I forgot where it was in the video but figured people would see the name Matt Hawrilenko and would understand what I'm talking about.

If you look through my post history you'll see me referring to this video or using the concepts from it several times. Which is why I knew exactly where that quote was in the video. So far only one person has responded to my posts where I use the concepts from that video and his response only showed that he completely missed the point.

I can tell you that I've validated the basic formula by posing situations to pokersnowie and comparing the results. So far the results are very similar. I would not say it is a perfect theory since it was derived from solving a toy game but it handles post flop nl very well as well as helps build balanced ranges. Contrary to what some people may say it is not just a river/all in theory. It applies to all streets post flop.

Learning and applying these concepts well will help almost any poker player a great deal.

I think that ignoring the potential to win or lose money on future streets is a recipe for disaster, particularly vs good players.

I don't know how that really makes sense.

I agree it's a very important concept but you should be exploiting weaker players and you have to think about their range to do that. I think it's a shallow interpretation of the quote any way.

My guess is he simply meant you should always be thinking about and actively working on range construction until you have enough experience.

The "until you can buy a house" was probably just an attention grabbing way to say "for a really long time in your career". Plus it's a tangible goal for people.

Even on the river, your opponent's range matters. Let's look at some math:

https://forumserver.twoplustwo.com/1...22/?highlight=

Notice that our ev changes depending on the opposing range.

He does mention exploitive play but only briefly. Where that fits here is by adjusting the value betting and bluffing ranges based on their play. It is well known that exploitive play is more profitable than a game theory based play but it is also fairly commonly accepted that one should understand the game theory based approach before attempting to exploit players.

Notice I did not say gto and that is because I doubt that this is optimal since the formulas were derived from a toy game(ace, king, queen game).

I think that a rather large amount of people suffer from the confirmation bias which makes it really hard, if not impossible, to understand this theory/take it seriously (notice I said theory and not hypothesis). So I decided to take some time and setup an example situation where we are the button which has open raised, got 3bet by the big blind and have called. For this example I’m going to use the preflop range suggested by snowie for the button pfr and the 3bet call. This theory assumes a solid preflop range and we could argue about the validity of snowie’s preflop ranges but for this example I’m just going to go with it. In this example we do know the opponent’s range but I’ll show a concrete example showing that we do not need to know their range in order to make solid post flop decisions. We simply need to know our range and where we’re currently at within that range. We should be adjusting our preflop range based on the perceived style of play of the big blind. But for this example I’ll assume no prior information on the opponent.


Snowie suggests the following preflop raising range for the button:
44+, A2s+, K2s+, Q6s+, J7s+, T7s+, 97s+, 86s+, 75s+, 64s+, 54s, A4o+, K9o+, Q9o+, J9o+, T9o


Like I said, in this example I am assuming we get 3bet by the big blind. Snowie suggests the following 3bet calling range:
QQ-44, AQs-A2s, K7s+, Q9s+, J8s+, T8s+, 97s+, 86s+, 75s+, 64s+, 54s, AQo-ATo, KJo+, QJo


For the flop I used the random.org generator to generate 3 random cards

The flop I got was 6dQcAd

Snowie suggests that the big blind should be checking just over 70%(72.15%) of the time and betting just under 30%(27.85%). Lets assume for this example that the big blind bets.

Snowie suggests that we should be calling 44.33% or roughly 101x combos. Our 3bet calling range contains 263x combos which when you account for card removal that 101x combos seems right.

Now lets use the formula from the video to determine our continuation range(includes raises and calls). The formula is 1/(1+(bet/pot)) which in this case I followed the suggestions from snowie and assumed that the big blind bet pot on the flop. This would mean we have a continuation range of 50% of our 3bet calling range which beat a bluff or have a strong draw. After card removal we have roughly 227x combos from our 3bet calling range left. Out of those combos roughly 196x combos have a good draw or outright beat a bluff. So then this formula says that we should be continuing with 196*0.5=98 combos. Which is 3x combos off of what snowie suggests. I could be a bit off on the combos left after card removal and then could also bit a few combos off on the number of combos that have a strong draw or beat a bluff. Either way this gets us extremely close to the post flop decisions made by snowie.

I ran out of time towards the end of this post and didn't really have the time to correctly sort our range relative to the flop or to count all of the combos with 100% accuracy so I made a few educated guesses. The results are extremely close to snowie even with a few errors on my part. I will come back later and post more accurate results as well as continue for the turn and river.

Thank you for the detailed post @alkimia.

Perhaps you'll clarify in later posts but with what you have provided so far it occurs to me that you still need to consider not just your own range but your opponents.

Phrases like "hands that beat a 3 bet" carry an implicit reference to your opponenrs range.

If I said that it was a typo. I mean that beat a bluff or have a strong draw

Besides the nut low, and nut draws, the bold classifications will be dependent on the opponent's range.

This is exactly what I was trying to get it. Thanks, Bob.

Not once in that situation was the opponent's range ever considered. You can easily tell which hands beat a bluff without needing to know the opponent's range and while the term "strong draw" is fairly subjective it also doesn't depend on the opponent's range. This is the confirmation bias not allowing you guys to see that there is a way to play post flop poker without trying to do magic and put a person on a range. In fact, despite what you may think, it is extremely unlikely that you or many other people are very accurate at doing that. It is a lot like how "body language experts" at various intelligence agencies are extremely confident in their assessments but their accuracy ranks up there with a blind folded dart throwing chimp.

What is a bluff? How do we know our hand can beat that hand? You're making assumptions about an opponents range implicitly when you do these things.

Considering your own range is extremely important but your hand's strength and 'where you are in your range' is all a relative assessment compare to your opponent's range.

No one is saying ranging people is easy, we're just saying evaluating your own range without the context of another range is somewhat meaningless. You can rank your own range based on the equity of the hands vs eachother but in the end you're trying to win money from other people and their range will matter as much if not more for your own decision with individual hands.

While I do think that many toy games such as AKQ apply differently on different streets in NLH (AKQ works much better on river calls in position with low SPR, than it does on dynamic flops with high SPR), this video shows how insightful theoretical toy games are at constructing a broad well balanced strategy:

48:09 if you want to design your strategy around a certain threshold, such as risking your tournament on holding trips or better, then you have to redesign your distribution with enough combos worse than trips that you can then fold. This will mean either playing more combos more aggressively on earlier streets, or betting larger sizings with the same combos, bringing more bluffs to the river and more thin value FROM THE SAME STARTING RANGE. It is all about your distribution, your combos and your bet frequency and sizing.

60:00 exploit opponents' weaknesses on earlier streets, and finish the hand with balance. Overfolders will overfold, overcallers will overcall, overbluffers will bloat the pot needlessly. All of this happens most frequently on earlier streets, BECAUSE THOSE STREETS HAPPEN THE MOST OFTEN. If you have expoited a player on the early streets, and finish the hand with balance, you win, end of story. The added bonus is that most of them can not perceive your strategy since the actual showdown is much closer to balanced.

The actual quote in the title of the thread:

"Let me be very clear on how important this concept of knowing your own hand and knowing where you are in your distribution is. I think you shouldn't think about anything else in poker until you have bought and paid for a house by knowing where you are in your own distribution and shaping it to be balanced."

The above concept absolutely crushes the micros. If you are not crushing the nameless rabble of amateur poker players with nothing more than this concept, applied to starting ranges that are well constructed (have board coverage) then you are doing it wrong. Very wrong.

The more you apply this concept, the more you start to bet more frequently and larger sizing on earlier streets, since THERE IS VERY LITTLE PROFIT IN BALANCED CALLING except against extreme overbluffers.

53:45 The weakness of toy games is range asymmetry, known to poker players as "range advantage". The math involved in more advanced extensions of (0,1) uniform distributions, is to take the range of the caller and set that as the new uniform and the bettor then has the advantage/disadvantage and bets according to equilibrium. This is the method of Ferguson & Ferguson and allows for complete frequency calculation and descritization to real poker hands with card removal. Thus, in theory, every bet comes from a balanced range and every call comes from a specific threshold combo in the range of the caller.

The opponents' bluffing range construction prior to the river is instrumental in defining "hands that can beat a bluff."

Pair outs matter for draws. The construction of the opposing range affects the profitability of every single draw in our range. The more unpaired hands the opponent could hold, the stronger a draw is due to pair outs.

where in our distribution do we bluff from? does this change on later streets? - it's often said you should bet the very bottom of your range if you get there on the river.

I finally watched the video and I see where some of my confusion was coming from.

There's still a lot the video glosses over that I think still implies your making some sort of decision about a theoretical opponent's range even when only evaluating your own range.

Overall really good vidoe that provides some clarity on toy game results and how those translate to actual poker.

Alkimia's way should be fine with regards to bluff catching in spots where you will never be raising. Using blockers to your opponent's range will do even better though and will sometimes be necessary forcing you to think about your opponent's range.

If you are going to consider value betting or bluffing then you need to know your opponent's range otherwise you're just in the dark.

As for us not being accurate, that isn't true. Versus a lot of people I can narrow their range down to their exact holdings with a strong probability on numerous occasions.

I think the thread got a little sidetracked with discussion regarding our opponent's range, or its insignificance when making decisions. I do think it's significant but we've been over that.

-----

Now, I'm thinking about the substance of what the original post is getting at:

The Practical Application of Minimum Defense Frequency

Let's contrast two scenarios:

Now, I believe that we should be more likely to call in the latter scenario, because the threat of losing lots of money is much greater in the former scenario. I don't think about mdf at all when making such flop decisions. I think that if I did, I would make lots seemingly 0ev calls on the flop and turn that are actually quite -ev if there are more betting rounds.

So basically I think making lots of (0ev) turn calls with hands that can beat a bluff, but that also have very little chance of improving, will supersaturate(for lack of better term) the river range with 0ev bluffcatchers, which just seems bad to me, unless the opponent is the type to give up on the river very often.

Just would like to add, that my opinion on the whole "who's range matters most when betting" has resolved to this:

GTO calls on the river are purely the caller locating her exact combo within her own range.

GTO bets/raises are a little different. The decision to bet (as value or bluff) comes from the location of the bettors holding within the bettors range. However, the SIZE of the bet is determined by the range of the potential caller, and the bet size should push the decision point into the geometric mean of the asymmetric range of the caller, or the point mass of the callers range.

So:

I look at my range and locate my actual holding. I have a hand worth value betting. I decide to bet. How much? Well, what does the caller have lots of that will be a close decision to pay me off? Do they have every combo of the second, third, fourth nuts (I have the actual nuts)? If so, lets bet large enough to make them indifferent with those very strong hands (usually this is all in).
Do they usually only have an overpair, or top pair? Lets size down a little, so that they have to be indifferent with those this time.

In tough games, one can more easily think about one's own range only.

The ranges for both are pretty similar, just on average weaker for the caller, especially in NLH where it also makes the biggest difference and in cases the caller might never be able to call down w/o thinking in GTO.

Thinking about the opponent's range is more important the lower one plays, especially when one is calling, but also after the opponent has called (and before).

It would be weird not to think about the opponent's range as soon as he has called and one has seen the next card or two. As so, one might play the opponent's range rather than (only) one's own range.

Playing like GTO might make the opponent fold often enough even on boards where he is strong, and if he won't fold, one gets more value when one has the goods. One is not exploiting here but plays one's own range only.