I'm thinking of starting to implement a formal randomizing tool, whether it be a coin, a dice or a clock.

I'm personally not worried about randomizing to obscure my ranges, but rather as a discipline tool to help my mental game be more consistent and reduce computational load.

For example, instead of thinking about board texture and my opponent's tendencies when considering whether to cbet or not, just roll the dice.

Another reason that I'm going back to this external randomization is that, as with many players, I'm being heavily influenced by unexploitable GTO, this added with the fact that I started playing in Ignition, which has anonymity, makes exploiting and adapting not as relevant.

Finally, I saw a professional poker player that pretty much just played GTO as if it were a 9 to 5 instead of big braining every decision.

It's all pointing towards the RNG.

I doubt I would ever play against any single opponent that:

a) I play frequently enough
b) That pays attention
c) That is competent enough to draw inferences

That I would need to ever do this

Same, that's why I took the trouble of explicitly mentioning this in the OP

I can see why you would skip a wall of text (or speed read) and just go straight into posting what you think are original contributions to a conversation. But try not to do that.

I'm all for speed reading, and I might miss something in a long conversation, like a thread that is 5 pages long, or a university book. But failing to read like 3 paragraphs is just not the way to go about life.

Might really help with those leaks that you just know you should fix, but you keep making anyways. Mine is not folding Top pair to pot sized jams.

Maybe if I just do a 25% or 50% call, I would have an easier time letting go of them, it protects me against the bluffs.

There's a spiritual aspect to this leak, it's like the spot where I thought I would be getting advantage over the fish at the micros, but when people flip overpair or sets, I realize I am the fish. Pulling out the randomizer for this spot, (as opposed to just folding them and being exploitable to bluffs or "missing value"), leaves the "value" I was seeking at the table, it sacrifices it, as I now can neither make or lose money from it, it just ceases to be an exploitative or exploitable spot.

In life in general I'm starting to feel that I was biting more than I could handle, and being beaten (or breaking even) at ****ing NL2 is a bit humbling.

I am even a bit thankful of being beaten, but because I feel not that I'm playing bad, but others are playing amazing. We've come very far with poker strategy over 20 something years, I knew this, I've been venturing into new formats like PLO6 and Short Deck knowing this full well, but I still kept a bit of hope of just being a winner on traditional Hold Em.

Nope, I would need to devote a lot of study to the game to keep up, and keep pushing the limits of the game, to the point where the rake becomes hard to beat, but that is the game.

Another important factor is the time rake, that we all pay, and to be honest, I don't know if I can justify spending enough time on poker to be actually good. My hourly at my actual job is just better no matter how I look at it, and poker is probably costing me as much in time as it is in money.

I like poker better when it's a non-repeated game anyways, who says that we need to run enough games so that we get the same scenario enough times that the variance dissapears. In the end this is a gambling game. Randomness and non repeatability is a core aspect of the game, even if you could play 1000 hours, you would adapt so it wouldn't even be the same game.

This thread is about using a tool for randomizing , but an actual is only the cherry on top of randomizing, the most important randomizing is the one we do with our gut (I'll use my 25% call on this guy, but not that one), with our cards (If I'm gonna call w AQ, I'll do it when the queens are unsuited, and the ace is a club), with an internal counter (Ok I called this one, so I can't call the next 3).


So anyways what I'm saying is that I'm a fish, a rec at best, that can't beat the microstakes and I will resort to literal randomness and severely limiting my playtime.

Thanks for reading my blag

If you want to save yourself some "computational overload" in this spot and actually take a higher EV action in this spot, just overfold by like a LOT. Who cares if they are bluffing you? They aren't doing it nearly enough at 2NL. Maybe call with the TPTKs and some hands with backdoors. Fold the rest. Every time. Try this for a few thousand hands and see how the graph goes.

They aren't going to exploit your overfolds at 2NL. If they could do that, they wouldn't be playing 2NL.

Yes I am speaking about TPTKs and losing to sets and overpairs.

If NL2 is so bad, I should still be winning while playing Nash/GTO.

Exploiting just isn't worth the risk of being beat at NL2.

How can you speak with any certainty that players at NL2 overfold, maybe they overbluff? Maybe a specific player does. And to a big degree in poker you are not playing vs a field, but vs individual players.

Winning at NL2 only because you beat 6 players but lose to 4 players feels like losing to me. The objective is to win or stay neutral vs all. Even losing to rake is a victory, if I lose to rake at NL2, let it be.

No, not making myself exploitable on NL2 on base theory, only would consider playing exploitively against specific players, not field.

Exploiting is the best way to beat microstakes, not GTO. GTO is defensive against other balanced players. It's not designed to make money, it's designed to make yourself unexploitable. But you would never play GTO against players whose leaks and imbalances are apparent. If you input unbalanced ranges into a solver, it would not play the same as it would against a balanced player. That's what you are doing. You're trying to use GTO strategies against the wrong ranges, which is a big mistake.

If most players in the pool underbluff, then play accordingly. Don't worry about the few individuals who are balanced or overbluffing (unless of course you can clearly identify them). My experience is that micro/low stakes pools tend to underbluff, especially for stacks and especially on rivers. But if you're seeing something different in your pool, then play accordingly. But I would not try to adjust to every player unless I had a strong read (or stats, although I don't use a HUD).

I personally think a this is a mistake.

You are right, in a perfect world you would exploit against specific players that you have a good "read" on their tendencies based on prior hands you've played with them.

But what about players you don't have enough hands with or don't know what their leaks are in a particular spot?

In that case, playing against population exploits will be more profitable than playing a balanced GTO range.

Your logic of "beating 6 players but losing to 4 players is still losing" doesn't make much sense imo.

Imagine you have a read on a specific player when they jam 5x pot on the river. If they were balanced, they should be bluffing roughly 45.45% of the time.

However, let's imagine that you know this player is overbluffing, and they are actually bluffing 60% of the time.

Well, now you should call with all your bluff catchers, 100% of the time to exploit them! But if you do this 10 times, you will see that you are winning 6 hands and losing 4 hands.

According to your logic, this is "still losing", remember? The only difference is that we are talking about individual hands instead of individual players, but it works out the same anyways.

Suppose in your pool you have 60% of players bluffing 80% of the time (overbluffing), and you have 40% if players bluffing 30% of the time (underbluffing).

Then when you encounter a random player that you have zero reads on, you can essentially treat them as bluffing 60% of the time! Because they are! Even though there's a 40% chance that this specific player is actually underbluffing, it doesn't matter.

Do you see what I'm saying? A random player from the pool is essentially an average of all the player pool tendencies and you can treat it as such even though it seems counter-intuitive.

In 100bb NL HE, there is first more than one Nash Equilibrium, second it is not EV0 against all strategies, it is only guaranteed to be at least EV0. There's a lot of mistakes villains can make that lose ev against GTO.

This is unlike, say, GTO strategy in Rock Paper Scissor, where your EV is always 0. Imagine it more like playing rock paper scissors ****. Where **** always loses, we alternate 1/3 each between RPS, and we win whenever villain chooses ****.

Hard no on exploiting and becoming exploitable to a "field", with minimal exceptions such as limping pocket pairs, and that kind of stuff.

I've already said that I'm playing GTO and barely exploiting individual players, so no way I'm playing against a field.

A lot of these responses are quite off topic.

The question is, do you use randomizing tools. We assume we are playing a mixed strategy.

If you don't play mixed strategy, gtfo. Even Harrington in the 90s was looking at his clock or suits to figure out whether to cbet or not.

Btw, I'm using a coin. Flip heads twice to call all ins with TPTK, that kind of stuff.

I use my watch on like a third of all faced riverbets. Wish I could remember to use it more, I find using RNG to be invaluable.

If you want to randomize, then I find the easiest way is to just use a poker chip.

You can pick a spot on the edge if the chip and treat it like a clock and shuffle your chips and see where the mark on the chip ends up relative to your thumb.

The nice part of using this for randomizing is that you can easily do it in live games as we'll just by shuffling chips.

But as an aside, you wrote several paragraphs in your OP describing why you want to randomize that just don't really make sense. You said you want to "reduce computational overload" so you don't have to consider board texture, opponents range, etc.

Essentially you are saying that you just want to randomize all of your actions and never consider the flop texture or runout or ranges? This doesn't make sense.

Even when randomizing, you still need to decide which hands will randomize and which won't. Then with the hands that will randomize, you have to decide which actions will be done with what frequency, etc.

All of those will be decided by board texture, ranges, runouts, etc.

What you are proposing just doesn't really make sense. You seem to misunderstand what the point of randomizing is and how it can be helpful.

But you seem set on ignoring anything anyone says that doesn't confirm what you wanted to hear, so go ahead I guess.

Here's a cool river flip I got in today

PokerStars, Hold'em No Limit - $0.02/$0.05 - 9 players
Hand delivered by CardsChat


LP: $3.21 (64 bb)

BB (Hero): $5.90 (118 bb)

Pre-Flop: ($0.07) Hero is BB with A♣ 4♥
, LP calls $0.05, Hero raises to $0.25, LP calls $0.20.

Flop: ($0.55) K♥ 5♣ A♠ (2 players)
Hero bets $0.15, LP calls $0.15

Turn: ($0.85) Q♦ (2 players)
Hero bets $0.20, LP calls $0.20

River: ($1.25) J♣ (2 players)
Hero checks, LP bets $1.19, Hero calls $1.19

Total pot: $3.63 (Rake: $0.18)

Showdown:
LP shows K♦ 4♠ (a pair of Kings)
BB (Hero) shows A♣ 4♥ (a pair of Aces)


BB (Hero) wins $3.45


I flipped 50/50 to call. Villain had almost 100VPIP and almost no PFR.

There's something beautiful about the coinflip, it's symmetric. Since we are playing GTO, we can't tell or depend on villain's playing too tight or too spewy, it is the same to us, it is the yin and it is the yang.

Sounds interesting for live as a discreet method, like the suits method. But how do you mark the chip?
I can see how you might interpret that from what I've written, here's the subtle distinction between what you interpreted and what I meant:

I of course consider many parameters like board texture, vill range given previous streets... when deciding what action to take, whether a pure strategy or a mixed strategy.

However once I choose a mixed strategy, (say, betting half of my missed flops with AK/AQ) I used to consider these parameters when deciding whether this was going to be the time I bet or not, that is, keeping the frequency of the mixed strategy, but not truly being random. For example: this opponent is very sticky, therefore this is one of those times I'm not betting.

So I'm only talking about randomizing to reduce the cognitive load once a mixed strategy has been chosen, figuring out whether we choose a random mixed strategy still costs the same cognitive resources whether we use an external randomizing tool or not.

Does that make sense?

So far I've used the coin like 2 or 3 times per 6 hour session. It's not something I use frequently. I am in fact still using the cognitive mixed strategy method for common spots like cbets and light 3bets.

Arent you forgoing the advantage that you will get from your circumstantial reads and opting to play worse by randomising in those cases? Like, you determined that you have the same ev of betting and checking in this example with your air, isnt checking gonna print/be less -ev if you know your opponent is sticky? Why randomise for a worse option if you know which is better? Except of course if you are afraid of being counter exploited, but youre saying its not about that. Do you think its not worth the computational load?

Im not qualified to give advice, but i see the use of that (again, not taking "being unexploitable" into account) only in spots when you can't lean one way or the other exploitably, so vs either very good regs or unknowns.