Ace
11 points
If Zemel9 ever folded KK here I would eat my monitor.
Feb. 14, 2013 | 5:48 p.m.
Just to make it clear, playing GTO is only technically the BEST option, if villain is playing GTO too. Otherwise he is going to be unbalanced and exploitable in certain spots. But regardless of how he plays, if you play GTO your overall gameplan is going to be +EV.
Unfortunately, noone will play true GTO for a long time if ever.
Unfortunately, noone will play true GTO for a long time if ever.
Feb. 14, 2013 | 5:45 p.m.
okay guys I asked this question on 2+2 and the answers I got there made more sense; http://forumserver.twoplustwo.com/showpost.php?p=37160034&postcount=11 explains it.
I think a lot of people in this thread seem to be under the misconception that you can exploit a true GTO gameplan, and by it's very definition, you can't.
I think a lot of people in this thread seem to be under the misconception that you can exploit a true GTO gameplan, and by it's very definition, you can't.
Feb. 14, 2013 | 5:43 p.m.
Yeah I hear you on point 2).
I think your original equation on point 1) seems correct, but rearranging it quickly I get z=(-)62.
I'm still super confused because I can't see what was wrong in my 4bet calcs, can you spot it?
I think your original equation on point 1) seems correct, but rearranging it quickly I get z=(-)62.
I'm still super confused because I can't see what was wrong in my 4bet calcs, can you spot it?
Feb. 14, 2013 | 3:44 p.m.
Yeah legit point, you are definitely right on point 2).
I don't really get what you are saying on point 1) though,? could you explain it abit clearer?
Also, point 3). I am not sure I follow. I agree it will be +EV to get them in, but I don't think it's the most +EV action.
Surely there is no point taking a path with QQ/ AK that leads to us getting them in vs an underdog when we can 3bet enough vs him without doing so?
I don't really get what you are saying on point 1) though,? could you explain it abit clearer?
Also, point 3). I am not sure I follow. I agree it will be +EV to get them in, but I don't think it's the most +EV action.
Surely there is no point taking a path with QQ/ AK that leads to us getting them in vs an underdog when we can 3bet enough vs him without doing so?
Feb. 14, 2013 | 1:06 p.m.
Which is obviously the problem I have when I look at the scenario I set out above.
Feb. 14, 2013 | 2:11 a.m.
Yeah I get what you are saying about the strategy pairs etc, but my overall understanding of GTO is different to yours. My understanding is that it is not always the optimal strategy, but that regardless of your opponents strategy, it cannot lose.
EG, In a basic game of Rock paper scissiors, the GTO strat is clearly choose 33% of each at random. No matter what my opponent does, this strategy cannot result in me losing EV wise. At worst I can break even.
I don't think this is a fluke of the way rock paper scissors is set up. I am pretty certain when i say that GTO strategy cannot lose no matter what counter strategy our opponent tries to enact.
EG, In a basic game of Rock paper scissiors, the GTO strat is clearly choose 33% of each at random. No matter what my opponent does, this strategy cannot result in me losing EV wise. At worst I can break even.
I don't think this is a fluke of the way rock paper scissors is set up. I am pretty certain when i say that GTO strategy cannot lose no matter what counter strategy our opponent tries to enact.
Feb. 14, 2013 | 2:11 a.m.
I also just worked out the optimal 3bet strategy vs UTG, pretty tired now so may well be wrong, but can anyone comment on this too?
Also when we are the 3bettor in late position, we need to 5bet enough that the 4bettor in early position cannot profit by 4betting any two cards. If they 4bet, they are risking 17 (20- their 3bb raise) to win 20 (their 4bet) + 10.5 (the blinds plus the 3bet) so we need to defend 1-20/30.5, 34% of our 3bet hands to stop them exploiting us.
This leaves us with a 34% value to 66% bluff range.
We can only 5bet hands for value that have >50% equity vs their calling range. The only two such hands are KK and AA.
We also have to 5bet bluff, enough to make villains weakest 4bet hands breakeven when they call off our 5bet. They need to call 80 into 121.5, so they need 80/201.5, 39.7% equity to call.
Thus we need to give JJ and AQs 39.7% equity, which can be done by making our 5bet range KK+, A5s 1/2*A2ss. That's right, I want you to use a randomiser every time you 3bet A2ss vs utg and get 4bet, and 5bet shove it 1/2 the time.
Anyway, we now have our 5bet for value range sorted as 1/2 *A2ss, KK+ A5ss. That equates to 16.5 combos, and should be 34% of our overall 3bet range. Thus we can 3bet bluff 32 combos ish, which obviously you can decide yourself but is something like; K8s-K7s A2s-A7s.
Also when we are the 3bettor in late position, we need to 5bet enough that the 4bettor in early position cannot profit by 4betting any two cards. If they 4bet, they are risking 17 (20- their 3bb raise) to win 20 (their 4bet) + 10.5 (the blinds plus the 3bet) so we need to defend 1-20/30.5, 34% of our 3bet hands to stop them exploiting us.
This leaves us with a 34% value to 66% bluff range.
We can only 5bet hands for value that have >50% equity vs their calling range. The only two such hands are KK and AA.
We also have to 5bet bluff, enough to make villains weakest 4bet hands breakeven when they call off our 5bet. They need to call 80 into 121.5, so they need 80/201.5, 39.7% equity to call.
Thus we need to give JJ and AQs 39.7% equity, which can be done by making our 5bet range KK+, A5s 1/2*A2ss. That's right, I want you to use a randomiser every time you 3bet A2ss vs utg and get 4bet, and 5bet shove it 1/2 the time.
Anyway, we now have our 5bet for value range sorted as 1/2 *A2ss, KK+ A5ss. That equates to 16.5 combos, and should be 34% of our overall 3bet range. Thus we can 3bet bluff 32 combos ish, which obviously you can decide yourself but is something like; K8s-K7s A2s-A7s.
Feb. 14, 2013 | 1:49 a.m.
'Okay so I'll briefly explain it from UTG.
We open a reasonable range of; 66+,ATs+,A5s-A2s,KJs+,QJs,J9s+,T9s,98s,87s,AJo+,KQo
which is 12.5%.
Someone 3bets us to 9x (we open 3x). They are risking 9 to win 3 (our bet) + 1.50 (the blinds).
This means they have that if we fold 9/(9+4.5) or 63%, they automatically profit with any two cards. Which is bad for us, it means they are exploiting us.
So we set about defending roughly 37% of our 12.5% opening range. This is 0.37*12.5, 4.625%.
Right, now we'll assume that when the 3bettor 5bets all in he has roughly 30% equity with his bluffs. He invested 91bb extra to 5bet bluff, so his net loss when he 5bet bluffs and we call is (approx) 91-60=-31
We 4bet to 20, then he 5bets he is risking 31(the amount he loses when we call) to win 20 (our bet)+9 (his 3bet) +1.50 (the blinds). This means if we fold more than 31/100+20+9+1.50, 24%, he auto profits. So we need to 4bet with a 76% value, 24% bluff ratio.
0.76*4.625=3.515. This actually equates to AQs, JJ+, AKo roughly.
We then balance this with 24% bluffs 1.11%, roughly A2s, AJs, ATs, KQs.
and yada yada your range is protected perfectly.'
I recently wrote this for someone, and having never actually done the maths myself (lazy I know) was fairly suprised by the results. Basically all the articles I read in the past were fairly dated and based their ranges on bigger 4bets and 3bets. The implications of a smaller 3bet are fairly dramatic, especially when people 3bet to 8 or 7.5 even IP nowdays.
Could someone check my working, and also would anyone reccomend a different strategy, maybe involving calling? Basically would appreciate some opinions because I was pretty shocked we have to value shove AQs to protect a 12.5% opening range really.
Thanks and gl.
We open a reasonable range of; 66+,ATs+,A5s-A2s,KJs+,QJs,J9s+,T9s,98s,87s,AJo+,KQo
which is 12.5%.
Someone 3bets us to 9x (we open 3x). They are risking 9 to win 3 (our bet) + 1.50 (the blinds).
This means they have that if we fold 9/(9+4.5) or 63%, they automatically profit with any two cards. Which is bad for us, it means they are exploiting us.
So we set about defending roughly 37% of our 12.5% opening range. This is 0.37*12.5, 4.625%.
Right, now we'll assume that when the 3bettor 5bets all in he has roughly 30% equity with his bluffs. He invested 91bb extra to 5bet bluff, so his net loss when he 5bet bluffs and we call is (approx) 91-60=-31
We 4bet to 20, then he 5bets he is risking 31(the amount he loses when we call) to win 20 (our bet)+9 (his 3bet) +1.50 (the blinds). This means if we fold more than 31/100+20+9+1.50, 24%, he auto profits. So we need to 4bet with a 76% value, 24% bluff ratio.
0.76*4.625=3.515. This actually equates to AQs, JJ+, AKo roughly.
We then balance this with 24% bluffs 1.11%, roughly A2s, AJs, ATs, KQs.
and yada yada your range is protected perfectly.'
I recently wrote this for someone, and having never actually done the maths myself (lazy I know) was fairly suprised by the results. Basically all the articles I read in the past were fairly dated and based their ranges on bigger 4bets and 3bets. The implications of a smaller 3bet are fairly dramatic, especially when people 3bet to 8 or 7.5 even IP nowdays.
Could someone check my working, and also would anyone reccomend a different strategy, maybe involving calling? Basically would appreciate some opinions because I was pretty shocked we have to value shove AQs to protect a 12.5% opening range really.
Thanks and gl.
Feb. 14, 2013 | 1:21 a.m.
Sean, I have a question about GTO and a situation that seems paradoxical to me.
GTO is defined as the optimal strategy, or an overall gameplan that cannot be exploited, and can at worst breakeven. I understand we are nowhere close to perfect GTO in NLHE at the moment, but imagine a situation in the future where we are.
We play perfect GTO, and face an opponent who plays completely GTO bar the fact that he never 3bet bluffs the CO.
Thus, everytime we 4bet bluff him with our GTO frequencies after he 3bets the CO, he wins money off us. At all other times, we play the optimal strategy pairs and breakeven. So he profits overall. But that's impossible if we are playing perfect GTO.
GTO is defined as the optimal strategy, or an overall gameplan that cannot be exploited, and can at worst breakeven. I understand we are nowhere close to perfect GTO in NLHE at the moment, but imagine a situation in the future where we are.
We play perfect GTO, and face an opponent who plays completely GTO bar the fact that he never 3bet bluffs the CO.
Thus, everytime we 4bet bluff him with our GTO frequencies after he 3bets the CO, he wins money off us. At all other times, we play the optimal strategy pairs and breakeven. So he profits overall. But that's impossible if we are playing perfect GTO.
Feb. 14, 2013 | 12:26 a.m.
I expect that is something to do with the fact that opening frequencies and 4bets are part of an overall GTO gameplan and if he is not 3bet bluffing he will probably be leaking money in some other area that makes up for the times that we leak some money to him preflop.
But I am still confused, because lets simplify it, to us playing GTO, and not adjusting to anything, and him playing GTO in all spots, apart from never 3bet bluffing the CO. We lose money to him everytime we 4bet bluff him after he 3bets us from the CO, but in all other spots we breakeven (I don't think him never 3bet bluffing the CO affects his other frequencies). So we are losing money. But with GTO we can't lose money. Sorry I am confused and would appreciate someone clearing this up, thanks. also semi drunk so sorry for typos
But I am still confused, because lets simplify it, to us playing GTO, and not adjusting to anything, and him playing GTO in all spots, apart from never 3bet bluffing the CO. We lose money to him everytime we 4bet bluff him after he 3bets us from the CO, but in all other spots we breakeven (I don't think him never 3bet bluffing the CO affects his other frequencies). So we are losing money. But with GTO we can't lose money. Sorry I am confused and would appreciate someone clearing this up, thanks. also semi drunk so sorry for typos
Feb. 14, 2013 | 12:13 a.m.
Nice post Stu. However, I think you misunderstood a few things. Basically, GTO means game theory optimal (obv). Which means it is the optimal approach based on an opponent playing game theory optimal too, which you seem to have said tbf. But basically, if I play perfect GTO (which obv I can't nor can anyone), no matter how you play you cannot beat me. It will no longer be the optimal strategy, because I will be leaking money in certain spots, but it will make me money overall, and probably significantly more than I would make by playing my own (suboptimal) exploitative strategy.
So basically I can play GTO and print of someone playing non GTO, eg if I protect my opening range with proper 4bet frequencies then everytime they try and exploit me by 3bet bluffing a hand that wouldn't fall in their GTO range, I am printing money, as they are trying to exploit the unexploitable.
My question was that if I am defending my opening range with GTO 4bet frequencies then it seems to me that I can be exploited, by someone who just doesn't 3bet bluff. But that seems counterintuitve, because the whole point of playing GTO frequencies is that they can't be exploited. So would appreciate someone clearing it up.
So basically I can play GTO and print of someone playing non GTO, eg if I protect my opening range with proper 4bet frequencies then everytime they try and exploit me by 3bet bluffing a hand that wouldn't fall in their GTO range, I am printing money, as they are trying to exploit the unexploitable.
My question was that if I am defending my opening range with GTO 4bet frequencies then it seems to me that I can be exploited, by someone who just doesn't 3bet bluff. But that seems counterintuitve, because the whole point of playing GTO frequencies is that they can't be exploited. So would appreciate someone clearing it up.
Feb. 14, 2013 | 12:09 a.m.
Right, been doing some work on my preflop ranges. I think I might be being retarded here, but if I'm playing a balanced preflop range as described in these articles, and I play against someone who never 3bet bluffs then am I not -EV? If not can someone explain how?
Feb. 9, 2013 | 11:09 p.m.
I think it's reasonable to xback the turn, some of his gutshots now have a pair and he's not folding any pair he called the flop with.
Dec. 28, 2012 | 12:42 p.m.
Sizing is fine and flop is 100% a call, would range him with JJ+ AsK which is maybe being pessimistic and still makes it a super snap.
Dec. 28, 2012 | 1:05 a.m.
Haven't watched the videos you've made so far yet so apologies if you've already covered this, but would be interested to watch stuff about postflop range construction in heads up pots on later streets.
Dec. 28, 2012 | 12:28 a.m.
BN: $258.43
SB: $100
BB: $100 (Hero)
UTG: $174.39
HJ: $171.05
CO: $100
SB: $100
BB: $100 (Hero)
UTG: $174.39
HJ: $171.05
CO: $100
Preflop
($1.50)
(6 Players)
Hero was dealt
T
A
UTG folds, HJ folds, CO folds, BN raises to $3, SB folds, Hero calls $2
UTG folds, HJ folds, CO folds, BN raises to $3, SB folds, Hero calls $2
Flop
($6.50)
J
8
7
(2 Players)
Hero checks,
BN bets $4,
Hero calls $4
Dec. 28, 2012 | 12:26 a.m.
On the flop, obviously we have to x/f a lot of our range. To counteract this I know we have to x/c some stronger hands, but am not really sure which one we should do this with? Is 99 better to x/c than AA because there are less hands to vbet against with? or is AA better because it is less vulnerable. Would be interested to hear input on this
As I see it I think we need to have a big range that bets flop and x/c's turn and river too, (just looking at this from the standpoint of the flop and ignoring what actually fell on the turn) Basically would be interested to hear how people construct their ranges on the flop and turn as preflop cbettor, sorry for the derail and mutliple posts.
As I see it I think we need to have a big range that bets flop and x/c's turn and river too, (just looking at this from the standpoint of the flop and ignoring what actually fell on the turn) Basically would be interested to hear how people construct their ranges on the flop and turn as preflop cbettor, sorry for the derail and mutliple posts.
Dec. 27, 2012 | 11:58 p.m.
Preflop I think you can call so no need to 3bet.
Postflop your thinking is good if we are at lower stakes or playing a fish, as it will be +EV to just barrel off given that the King is going to be such a scare card for most of their range.
However, you have to consider that people are going to recognise that the king is a good barrel card, and if they see you barrel this hand (which has pretty much no equity vs better on the turn) they will assume your barrels are very unbalanced.
Basically, you can't bluff this run out vs anyone semi competent 100% of the time, so pick hands with more equity to bluff with (Adx Qdx 89s etc, or stuff which blocks his Kx range like QJ.)
Also I wouldn't even cbet this flop a lot of the time, it's not particularly good for your preflop 3betting range (I assume), and he's going to fold close to 0, and there are few turn cards which improve you. Balance the fact you will x/f a lot here by x/c or x/r with some stronger hands.
Postflop your thinking is good if we are at lower stakes or playing a fish, as it will be +EV to just barrel off given that the King is going to be such a scare card for most of their range.
However, you have to consider that people are going to recognise that the king is a good barrel card, and if they see you barrel this hand (which has pretty much no equity vs better on the turn) they will assume your barrels are very unbalanced.
Basically, you can't bluff this run out vs anyone semi competent 100% of the time, so pick hands with more equity to bluff with (Adx Qdx 89s etc, or stuff which blocks his Kx range like QJ.)
Also I wouldn't even cbet this flop a lot of the time, it's not particularly good for your preflop 3betting range (I assume), and he's going to fold close to 0, and there are few turn cards which improve you. Balance the fact you will x/f a lot here by x/c or x/r with some stronger hands.
Dec. 27, 2012 | 11:58 p.m.
Totally depends what uni your at and course you're doing. If you are doing media at Oxford Brookes then quit (no disrespect meant) but if you are doing econ at LSE than don't.
Dec. 13, 2012 | 11:17 p.m.
I'll give no.2 a stab because bored 1 tabling a tournament.
Basically poker is a solvable game, it's just incredibly complex. But there is a solution or strategy such that if you play it, you cannot lose money. This is known as GTO, or being perfectly balanced. The best way to understand it is look at a simple game like rock paper scissors. The GTO solution to this game is pick rock 33.33%, paper 33.33%, and scissors 33.33%. If you do this it is impossible for you to lose (at least EV wise) no matter what your opponent does.
However, most of the people we play in poker are very unbalanced. To continue the analogy, a typical opponent will pick rock 45%, paper 30%, and scissors 25%. Therefore we can exploit them by playing unbalanced ourselves and picking rock paper 100% of the time to maximise our EV (assuming they don't catch on.)
The GTO solution to poker has not been found, but good players who play other good players on a regular basis will try and play as close to GTO as they can.
Just reread your question slightly and realised what I wrote above probably over simplified it a lot, but i'll leave it here anyway, hopefully someone will find it useful, and someone else can go into frequencies more.
Basically poker is a solvable game, it's just incredibly complex. But there is a solution or strategy such that if you play it, you cannot lose money. This is known as GTO, or being perfectly balanced. The best way to understand it is look at a simple game like rock paper scissors. The GTO solution to this game is pick rock 33.33%, paper 33.33%, and scissors 33.33%. If you do this it is impossible for you to lose (at least EV wise) no matter what your opponent does.
However, most of the people we play in poker are very unbalanced. To continue the analogy, a typical opponent will pick rock 45%, paper 30%, and scissors 25%. Therefore we can exploit them by playing unbalanced ourselves and picking rock paper 100% of the time to maximise our EV (assuming they don't catch on.)
The GTO solution to poker has not been found, but good players who play other good players on a regular basis will try and play as close to GTO as they can.
Just reread your question slightly and realised what I wrote above probably over simplified it a lot, but i'll leave it here anyway, hopefully someone will find it useful, and someone else can go into frequencies more.
Dec. 10, 2012 | 1:10 a.m.
My computer says the link is high risk, if anyone can confirm it's safe or post it on weak tight would be cool
Dec. 8, 2012 | 7:45 p.m.
Not entirely sure if my line of thinking is good in above post though so don't take it as gospel, I'm kind of bad too in these spots, think it arises from having not played enough HU.
Dec. 8, 2012 | 7:43 p.m.
BTN can 3bet a super wide range here, and without some history I would assume he's never 3bet/folding as default.
I also think generally if you are calling this hand preflop then you have to x/c this flop really as you have flopped a fairly strong hand and villains generally cbet all their worse 7x here (I think anyway.)
You can x/r 87s, 67s kind of hands which are going to play much better and have less value vs his cbetting range.
I also think generally if you are calling this hand preflop then you have to x/c this flop really as you have flopped a fairly strong hand and villains generally cbet all their worse 7x here (I think anyway.)
You can x/r 87s, 67s kind of hands which are going to play much better and have less value vs his cbetting range.
Dec. 8, 2012 | 7:42 p.m.
I can definitely see your point but it's worth noting that in zoom fish go bust at lower limits so quickly and you obviously change tables in a massive player pool so the value of that information is massively decreased when compared to normal tables.
Dec. 8, 2012 | 4:42 p.m.
Yeah with flop action I think utg has a set more often than a nut flush draw. Generally at 50nl people play nutflush draws way too passive IP deep i've found.
With that in mind, I think call flop fold turn is much better. Once he slightyl overships turn your pair of 9's is rarely relevant and without stoving it I'd guess he needs a nfd or AA quite a lot to make it a call (assuming he has all set combos which I think is a reasonable assumption deep)
Just stoved it and you have 41% vs (2 combos of AA, two combos of KK, all sets, 87s 65ss AKss AQss) which i don't think is a majorly optimistic range so calling is okay I guess.
With that in mind, I think call flop fold turn is much better. Once he slightyl overships turn your pair of 9's is rarely relevant and without stoving it I'd guess he needs a nfd or AA quite a lot to make it a call (assuming he has all set combos which I think is a reasonable assumption deep)
Just stoved it and you have 41% vs (2 combos of AA, two combos of KK, all sets, 87s 65ss AKss AQss) which i don't think is a majorly optimistic range so calling is okay I guess.
Dec. 8, 2012 | 2:55 p.m.
I think someone we need to determine if he is ever 4bet bluffing. If anyone who 4bets to this size is ever 4bet bluffing it's a big leak and something we can take advantage of.
However, I think generally people who 4bet to this size at 25nl are unbalanced towards value. The sample size is fairly important too. Generally I'm still getting it in here though because he will bluff a non zero % and 3bet folding QQ is just sucky without really good reads.
There is definately situations you can do it (eg live vs really bad players who will 4bet super unbalanced and flat 3bets really wide), but I don't think we have enough information about this player to put him in that category yet.
Maybe I'm just bad though, would be interested to know if anyone does fold here.
However, I think generally people who 4bet to this size at 25nl are unbalanced towards value. The sample size is fairly important too. Generally I'm still getting it in here though because he will bluff a non zero % and 3bet folding QQ is just sucky without really good reads.
There is definately situations you can do it (eg live vs really bad players who will 4bet super unbalanced and flat 3bets really wide), but I don't think we have enough information about this player to put him in that category yet.
Maybe I'm just bad though, would be interested to know if anyone does fold here.
Dec. 8, 2012 | 2:46 p.m.
Also minor point but go $2.50-$2.75 on the flop because their calling ranges are going to be fairly inelastic here.
Dec. 8, 2012 | 2:40 p.m.
Load more
Feb. 14, 2013 | 5:50 p.m.