Between September 2021 and August 2022, Michael Addamo had a phenomenal streak of results and nearly $12 million in prize money. But then something changed, and in the last five months, he has not cashed in a live tournament. What's happened?
Towards the end of Michael's historical run, I posted a video in which I proposed to fight against his super-aggressive style with counter-aggression. This idea was based on the assumption that he was overdoing his big bluffs. It seems that since about the middle of last year, his rivals have begun to put out the fire with gasoline, so to speak, abandoning a more passive approach based on catching bluffs. Passive lines gave Addamo two ways to win: 1) realizing full equity in the hands, 2) pushing opponents out of the pots with bluffs.
Of course, I am far from thinking that my videos have somehow influenced Addamo's current results. His downswing is most likely due to the usual variance. However, I want to take the opportunity to show why counter-aggression is effective against active players like the Aussie.
Michael's opponent in this hand will be Daniel Negreanu. I highly recommend the review of his recent victory in the Super High Roller Bowl tournament, published on his YouTube channel. Daniel gives everyone the opportunity to look into the head of one of the best players of all time, it's very cool.
The hand is played in the early stages of the tournament, and the stacks of all players are very deep. Everyone folds to Addamo and he raises to 4.5bb with .
Negreanu defends with .
We don't know if Addamo is using one or two raise sizes. For this analysis, we will assume that he always raises to 4.5bb. You can take a look at the difference between 3bb raise and two sizing strategies. Spoiler: they are slightly different.
On a suited flop, Addamo continues with a 1bb minimum bet.
Daniel raises 8.5x and Michael gives up after thinking for a while.
And that’s all? Why probe into such a meaningless hand with such a small pot? But, even small giveaways sometimes help to gain wisdom.
Let's first talk about Daniel's decision to raise with middle pair. Solver probably doesn't play like that, right?
Well, yes and no. Indeed, the solver doesn't want to raise such a big sizing with Q8o. The expectation of a call, according to his calculations, is 5.9674, and a raise is only 4.4365. The difference is pretty big! So if you compare Daniel's line with the Nash balance, he didn't play according to the GTO. However, if we listen to how Daniel explains his line, we can see that he came to his decision using the principles of the GTO game.
First, he correctly assessed how widely his opponent uses min-bet in this situation.
The fact that the aggressor has a high probability of having a flush forces the defending player to discount many of his potential outs. Therefore, a fair share of the BB's range should be given up for even such a tiny bet. Small sizing means that all kinds of combinations will be represented at different frequencies. It will have flushes, sets, and any pairs, and complete garbage. Against this range, Q8o is somewhere in the middle: they beat many hands, but they lose hopelessly to many others. Why is this raise necessary? Aren't these the kind of hands we want to play small pots with?
Yes, they do, but not always. Daniel explained that he would be raising with a balanced range that included very strong hands, very weak hands, and average hands. Why does he raise and the solver doesn't?
Daniel gave two main reasons. First, he wants to clarify the situation, to understand where he is. Of course, for a solver this is complete nonsense, but it can be useful for a person. Daniel theorizes that one reason for Addamo's success is that he goes against the tendencies of the field. I talked about the same in my videos. Addamo noticed some lack of aggression in the game of most MTT pros. Perhaps they are not strong enough to defend against aggression, because they played a lot in mass tournaments, in which the field is very under-bluffing.
Aggressive play brings initiative, and initiative brings clarity. For a computer, this is completely unimportant, but for a person, it is a great psychological help. When we have the initiative, we usually have a good idea of where our hand is in relation to our opponent's range. We know if we're hitting or bluffing. The defending side does not have this clarity. The aggressor's opponent, unless he's trapping with the nuts, never knows if he's ahead or behind. To feel confident in his decisions, he needs to defend with a well-balanced set of combinations and with an ideal frequency. Of course, this is not possible for a human being. And uncertainty breeds mistakes. I don't think it's nice to pay $100,000 to enter a tournament and get three barrels from Addamo with a 200bb all-in on the river when you only have one pair.
Second, Daniel mentions that his raise prevents Addamo's hands from realizing equity. This reasoning, of course, is quite in the spirit of the solver. Daniel correctly recognized that his second pair was ahead of most of his opponent's hands, but he could be beaten on the turn and river.
In this particular case, the solver does not have Q8o do the big raise, preferring to use fives and some eights with weaker kickers for this. If Daniel is raising with all these hands and with Q8o, it means that in this situation he is overdoing it with aggression. Well, perhaps. But is it really that important in practice?
Here's what Daniel himself has to say about balance.
When I play high roller tournaments, most of my opponents play a very similar strategy. Same sizing, same lines... But Addamo is different! He is his own law. And I respect and appreciate it. His play made me wonder if people could even get anywhere close to the perfect solver strategy. And how important is it to be balanced in such conditions, if no one, in a good way, can exploit your shortcomings?
Let's say you play suboptimally for a long time, and people slowly begin to notice exactly where. Once you get caught, you're in trouble. For example, Daniel never, that is, never bluffs on the river! OK? If I'm really not bluffing and everyone finds out about it, I'll be destroyed by folding all my bets on the river. This is exactly what happened in 2016-2017.
Two negative years. Partly due to bad luck, partly due to the fact that everyone began to fold, damn them!
And so they discovered me – and now, that's it, quit poker? Not really. I just had to become more balanced, and bluff more often. But until I was discovered, I made money! And it works both ways: if you're rebluffing wildly and you're not being shown more often than usual, you'll be fine... until your opponents figure it out.
Perhaps the only reason to strive for balance is if you have to play a huge number of hands against the same opponent. But this is not the case in tournaments.
Daniel's comment reminded me of a recent twitch stream. The streamer parsed the hand he played using PioSolver. It was a 6-max hand where Hero had a pair of sevens and went all-in on the river as a bluff. The solver showed that this was a mistake and that it was necessary to bluff with pockets from deuces to sixes.
However, the streamer defended himself by saying that there are no pocket deuces to sixes in his range, so instead he balances value with bluffs with higher holdings.
It sounded absurd to me. If you follow my channel long enough, you probably know that we do not consider the conclusions of the solver to be the ultimate truth, but are trying to penetrate the logic of the game with their help. Meticulously following the solver's instructions only works when the ranges of the real hand exactly match the ranges of the solver – that is, never.
Why is the solver looking for balance? So that he could not be exploited, even knowing his strategy. The goal is not to balance for the sake of balance, but balance to protect against exploitation by the opponent. But if the weakness of your strategy is not exploited in real life, it is not necessary to change towards a more balanced game.
What your range really is doesn't matter. The important thing is how the opponent sees him because it is on the basis of this that he will choose response actions. If he thinks you have 22-66 bluffs and you don't, you don't need to add bluffs with other hands to balance.
In reality, balance is needed only in long heads-up matches of hundreds of thousands of hands against the same opponent. Under such conditions, you can really get exploited, because you can gain distance and accurately determine your range in certain situations. But as we add more positions, more players, and more varied effective stack values, your vulnerability to being out of balance doesn't matter.
You should not, of course, bring this idea to the point of absurdity and always put three barrels of any board – even beginners will realize such a strategy.
Some measure of balance is still needed, but it should not be more than necessary to hide your strategy from your opponent in the hand.
Therefore, the risk that Daniel will be exploited for his optimistic raise is minimized.
If we are not interested in balance, what should we strive for? Well, we already talked about maximizing EV against a particular opponent. Sometimes this requires a strategy that is close to the Nash equilibrium. Sometimes you need to play differently. When making a decision, you should not try to copy the actions of the solver – focus on what you think will bring you the most chips. If you reduce everything to memorizing solver ranges, forget to notice other important factors that affect the course of the hand and often lie on the surface.
You can deviate from this rule, perhaps, only when you start learning the GTO strategy. In this case, yes, memorize and reproduce until you understand the main points. At this stage, adding exploits will make it harder for you to learn. But once you've mastered the basics, don't be afraid to go beyond the strict limits of Nash equilibrium and just play.
Daniel also went down this path. At the very beginning, when he was intensively training with coaches, his behavior at the table was not much different from the stereotypical European bio-robot in a hoodie. But it would be strange to forever discard decades of invaluable accumulated experience of one of the leading players in the world. Now we see the old Negreanu, who is very comfortable at the table, ready to trust his instincts and back away from solver lines when it seems more profitable. It's just that he has more tricks in his arsenal.