what's more important? the right frequencies or the right combos?
Say for example you're on the river considering whether or not to bluff raise all in as I was today. You ask yourself whether what you have is a good combo or not. You end up shoving only later to find that the combo you used was -11bb.
This happens often where I end up being pretty inaccurate on rivers. Am i right in thinking it's better to have an overall aggressive strategy and make quite a few large errors rather than waiting for complete confidence in your bluffing hands? My bluffing frequencies are likely to be very nitty if I choose the latter approach.
12 Replies
Say for example you're on the river considering whether or not to bluff raise all in as I was today. You ask yourself whether what you have is a good combo or not. You end up shoving only later to find that the combo you used was -11bb.
This happens often where I end up being pretty inaccurate on rivers. Am i right in thinking it's better to have an overall aggressive strategy and make quite a few large errors rather than waiting for complete confidence in your bluffing hands? My bluffing freque
Unless you are blocking the nuts or other calling hands, your river bluff combos should probably be about the same as each other in terms of their success rate.
In my opinion the single biggest reason to limit yourself to certain combos of bluff hands IS to keep your bluff frequency naturally under control.
If you are going to pick combos outside the "acceptable" range to bluff because you feel like you have a good spot to bluff, are you also passing on bluffs with "better" combos in other spots where you don't feel it that time? If you aren't doing that, you're going to end up bluffing at too high a frequency and burning money.
Unless you are blocking the nuts or other calling hands, your river bluff combos should probably be about the same as each other in terms of their success rate.
In my opinion the single biggest reason to limit yourself to certain combos of bluff hands IS to keep your bluff frequency naturally under control.
If you are going to pick combos outside the "acceptable" range to bluff because you feel like you have a good spot to bluff, are you also passing on bluffs with "better" combos in other spots w
For myself in the past, I've usually ended up underbluffing quite a lot due to not bluffing because of having some small doubt about whether my hand is a bluff or not. I'm naturally a quite defensive player. The thing with bluffing those rivers is, you ether find the 0ev bluff, or mess it up and find the -10bb one. If you're not capable of finding the right combos, you have two options , ether underbluff and potentially get exploited against sharp players, or make a bunch of hugely negative ev bluffs to maintain the right frequency. It is quite hard to find the ideal combos, especially in mtts when things are so dynamic.
Lets just say, hypothetically, is it better to underbluff on the river in a large pot by 25%, but with the right combos. Or bluff at the full equilibrium frequency, but with some large inaccuracies?
In terms of the exploitability of your strategy, frequency is far more important than combo selection.
i like your question, because it boils down to a problem which i have encountered within my play style / thought process. when i am bluffing i often focus on good properties and neglect frequencies. this is extremely prominent in this scenario: BB vs UTG, hero holds JT on AKTx(x) and considers bluff raising. JT blocks str8/2pair/set and has low sdv, therefore has perfect properties for bluffing. thing is that heros value range is very thin w/ some QJ combos, so there's not much room for many bluff combos. next to JT, there are QT, KJ, KQ which all share the same properties. if he now starts blasting off all of these combos, just bc they have nice properties, his frequency is way out of line.
this scenario is also nice to showcase that your gto strategy can differ from the solvers one. say the solver picks JT but you decide to use KQ. from villains perspective it doesnt matter much, as long as his 0 ev bluffcatchers are indifferent. heros frequency needs to fit and that's all. of course, you can also look at the whole thing from a very detailed theoretical perspective, where every little EV counts. Then it is important which combos you choose, but a human will not be able to achieve this level of accuracy anyway (apart from Linus and friends).
The particular combos chosen with which to bluff almost always block hands that your opponent might call. The bluff frequency should be low enough so that most of your bluffs are combo specific for this reason.
In the more general sense of your post, calls can be made with frequencies spread over chunks of range, while bets are quite specific to combinatorics.
Just my opinion.
Just saw a great Kanye quote:
“If every time someone asks you a question and you try to say the right answer, your entire life is a test. Just say what you feel. It just is. We just are”
Making the highest EV play you can every time is the most important thing, not necessarily choosing the "Right" anything.
within the constraints of the question though, i would say its all contextual. Depends on if the site has HUDs, is it anon, is it live, how good are your opponents, is the spot in question very frequent or not...
The particular combos chosen with which to bluff almost always block hands that your opponent might call. The bluff frequency should be low enough so that most of your bluffs are combo specific for this reason.
In the more general sense of your post, calls can be made with frequencies spread over chunks of range, while bets are quite specific to combinatorics.
Just my opinion.
No. On the river gto will suggest bluffing not only based on blockers but on reverse value, i.e. polarizing.
Example, heads up in a minraise, cbet, check check board 56 will see much higher bluff rate than q7o in a high card board because q7 has showdown value.
So no, bluffing combos are not usually selected based on blockers.
In terms of the exploitability of your strategy, frequency is far more important than combo selection.
Solver strategy uses card strength to randomize because it's all the info it has.
I.e. 67 will bluff 10%, 65 will bluff 90% , even if the value difference is marginal ( 67 only beats 65). Because that was the cutoff of the rankings at the needed frequency (say we needed 20% lo frequency to match our 20% hi frequency)
I think you can simplify by bucketing the combos into a 20% lo bluff and use other factors like reads or player stats.
Yes you are exploitably dropping value by deviating but it's sooo small. The value lost from bluffing with 19th percentile instead of 21th percentile hand is just that you have to check with 21th percentile and you get beat by the matching 19th and 20th percentile in villain's range. It's almost no value.
Better sometimes to ignore the actual gto combos and focus on the frequencies. Use other factors to base your mixed strategy in.
One example would be you play against a player that folds three hands, which suggests he is rather tight, but maybe he just got bad hands and is playing gto.
Conversely in another situation a player just limps his first three hands, but again he might have been playing gto and got pocket pairs and sc.
I'd rather make my bluffing decision by bluffing one and not the other (but never both, as that would be being very exploitable to a different class of plays) rather than on if my kicker is a 7 or a 9.
This is distinct from "lol **** gto" in that we keep gto frequencies, but we mix up the combos based on exploitability factors.
And so far the examples used 3 hand histories, imagine if we have 1000 hands played with a nit rival, we are pretty confident what his 3bet range is, the risk of being exploited by villain having bad cards over 1000 hands is very low.
For myself in the past, I've usually ended up underbluffing quite a lot due to not bluffing because of having some small doubt about whether my hand is a bluff or not. I'm naturally a quite defensive player. The thing with bluffing those rivers is, you ether find the 0ev bluff, or mess it up and find the -10bb one. If you're not capable of finding the right combos, you have two options , ether underbluff and potentially get exploited against sharp players, or make a bunch of hugely negative ev b
This thing about getting exploited by sharp players is the biggest poker myth of all time. As long as you find the mandatory/obvious bluffs in every spot, no one is going to be able to tell if you're underbluffing or not.
The fact is, some spots are naturally overfolded, and others are naturally overcalled. If you overbluff the overfolded spots and underbluff the overcalled spots, you're actually quite balanced from an outside perspective. Not to mention the EV of your strategy would be much higher.
This thing about getting exploited by sharp players is the biggest poker myth of all time. As long as you find the mandatory/obvious bluffs in every spot, no one is going to be able to tell if you're underbluffing or not.
The only reason why it’s a myth is because the sharp players aren’t very sharp. If you took any of the regulars in this CoinPoker challenge and put them in any game then they would exploit any regular by many bb/100.
Also what is mandatory is not what is obvious.
You know theory better than most.
If I gave a person A♠5♠ OTB and they 3bet a CO RFI do you think they would know to 100% double barrel on a Q♦8♦4♠ Flop with a 4♦ OTT? Or would they just X back OTT with their no equity hand?
They would probably X back so now they have no bluffs on a 4 diamond runout. So now we can easily overfold on these runouts because players aren’t finding these non intuitive double barrels.
There are countless spots like this in MDA where we can exploit population because of average reg tendencies. River nodes have very high discrepancies versus theory. Almost no spots are played theoretically correct in defending frequencies or bluffing frequencies.
