Kalupso's avatar


2938 points

Yes, only betting with Kh or Kc. Blockers make a big difference in this exact situation and especially when betting 200% pot. The lower fold than MDF (do we have a better term?) gives means the worst air will be minus EV to bet and the better air indifferent between bet and check. KQ won't show down very often or with much equity although both ranges have around 70% unpaired hands. Like 30% of IP's range is A high and the remaining 35% unpaired hands will bluff quite often and also improve some of the time. Maybe we will have around 5% equity if checked all the way down?

Oct. 19, 2020 | 6:34 p.m.

I know you are not big on solver strategies but I thought we could use it to highlight the point that certain hands juts works better as bet and certain works better as check. When you have both hands that clearly wants to check and clearly wants to bet we have to split the range to not commit a large mistake with a part of our range.

For reference let's look at "GTO" strategy as a baseline:

So it's basically a very polar approach with big size because the boar is great for IP (used preflop range from DB for IP caller, weak reg alias).

Let's simplify the tree to 75% pot or check to see which hands are clearly better bets or checks for a "standard" size:

The green parts are the hands that strictly prefer to check. Those are unpaired hands without a FD or a high FD blocker (typically blocking A or K high FD). All the lower FDs that check (not paired or A-high) always XR after checking and EV as XC are much worse than XR for those. There aren't actually any value hands except for KhKc that is much better to bet than check but sets and AA always goes for XR after checking.

All of the above assumes that IP underfolds compared to what MDF formula would give. Frequencies against 3/4 pot which according to MDF formula would required a fold 3/7th of the time (about 42%).

In practice, we can see a fold frequency that's even higher than 42% and then a strategy of just betting a lot without having a hand with good properties can be fine.

Oct. 19, 2020 | 2:06 p.m.

Comment | Kalupso commented on BTN vs BB 3bet pot

Also be conscious about not being too much in your own head. It's very common among poker players in particular and me included quite often (at least in the past). Good poker decisions aren't made by thinking as hard and as much as possible. Things actually have a tendency to get more removed from reality and distorted the more we think. It's the same in real life too and taken to the extreme it becomes serious mental disorders. Sam Harris in a podcast said that all people sort of are on a spectrum between insanity and completely healthy. I mean on a spectrum between being in the head and believing all thoughts to seeing reality exactly how it is with no distortions. Meditation certainly is a great antidote for this thing and can help us move further towards the healthy side of the spectrum.

Oct. 18, 2020 | 10:05 a.m.

Comment | Kalupso commented on BTN vs BB 3bet pot

Appreciate any opinions and criticism of my thought process.

You asked for it :D

Villain has low c-bet and very high fold to 3bet stats, but only 150 hands sample

Don't use these stats with such small samples as it takes between 100 and 200 samples for a stat to start to be accurate but they can give useful information about extreme tendencies before that. We can't do much better than VPIP, PFR, 3bet, and some overall stat for postflop aggression with this small sample. Getting those stats would be useful for the analysis ;-)

Think it's standard.

I'm getting a bit critical but standard is maybe the worst word you can use in any poker analysis and it's not a reason for a play. Your hand is actually extremely close to between call and fold in games with rake to the point where small differences in 3bet size, 3bet range, and flop SPR will shift it between call and fold. Anything but standard!

Here I was thinking that his range will be very strong given his apparently tight nature and low-cbet stat so I'd have low fold equity.

Once again, those stats are useless here. Bet or check are both good and there isn't a strong reason I can think of to not have it in both lines, so it looks like a good spot to mix so you can hit a variety of hands in both bet and check line and you don't want to always bet or check your other hands either.

Here if he had checked I'd bet and I think a raise is ok but I like that my hand is kinda "disguised". Also think that since he's tight on c-bets I couldn't raise a sizing that would make him fold strong flushdraws as his range is so protected by overpairs, strong top pairs and JJ. Not sure my thinking makes a ton of sense here tho. Also with my draw being dominatable don't want to bloat the pot.

There is a ton of thinking here and most of it is things that is very different from what experts think about. Things I'd consider is:
- Small bet size on draw heavy board.
- Checked flop so weaker range
- Likely a reg player
- A lot of unpaired Ax in his range

All of those indicate a weaker than average range. Like you say, there is a lot of good FDs in his range so just a turn raise might not get all the folds but maybe raise turn and barrel river will? An option can also be to call turn and expect a very high EV river bluff when he checks. Both options seems fine and it's difficult to quantify which is best. If you raise, you likely have to fold to a 3bet and will lose some equity realization but might also get too many folds.

River you need to bluff everything without SDV as he will fold way more than MDF formula will tell you. Size is probably fine but it's a bit in the middle as you want to go bigger with the best hands and smaller with Jx. I think bluffing bigger will get extra folds and it's not so inelastic given small turn bet which means he will have some thinner bets like 88, TT and 9x (now trips).

Oct. 18, 2020 | 9:05 a.m.

I can smell PokerSnowie :D

The guys above gave good answers 🙂

Oct. 18, 2020 | 8:32 a.m.

I kind of hate full range check OTT. There are too many hands in your range that is strictly better as bet than check. Low FDs for one always wants to bet here and checking them always is not good. Range splitting doesn't have to be so complex always and it's not like a full range bet or check is a balanced strategy either.

Oct. 17, 2020 | 4:12 p.m.

Population at 100nl and lower cold calls too wide and will be overfolding spots like this (both regs and fish). I see no problem betting a ton merged either if the calling range is wide enough. In spots you have more medium hands, you can just size down a little to like 50% pot or something to not overplay them. I don't think you will get away with this approach at 200nl+ so looking to learn to mix a few spots it's more important to do so is likely a nice way to improve if you have ambitions of moving up n

Oct. 17, 2020 | 3:20 p.m.

App games on both US facing and Asian facing platforms have FR games up to like 5/10 running with regularity. MTTs are also close to exclusively FR but of course another beast.

If you look at interviews of the best live players, you'll see that a surprisingly large proportion of them did battle HU online for at least one year of their poker career and I don't think that's random.

Oct. 17, 2020 | 1:10 p.m.


Oct. 17, 2020 | 12:21 p.m.

There is FR in some app games but even there it's easier to get multiple tables if you can play short handed. Might as well learn short handed and HU if you want to become great at this game 😃

In HU especially you get the largest repetition of difficult spots and can improve general poker skills faster. Think Brazilian futsal vs football. Players develop technique and skills for football quicker in futsal because of the nature of the game. Only playing the most slow paced poker format will likely keep your growth slow.

Oct. 16, 2020 | 7:38 a.m.

This a sim I made where I tweaked pre and flop a bit to fit low PFR fish but left turn and river with equilibrium sub-game. Turn probably plays somewhat close to the equilibrium sub-game which is folding most made hands weaker than Ax, calls all AX or better and some Qx, and calls the best draws. Also XR the sets and some combo draws. Flopzilla is the more traditional software for this kind of range analysis where we are more concerned with combinatorics than the strategy itself.

The range contains 34% second pair or weaker and 66% top pair or stronger.

Those are the mixed hands that we need to get a fold from to profit when shoving and below for 67% pot.

Those are the "objective" specifics I'm able to come up with about the spot. Can we use this to make a strong conclusion? I'm not sure but I believe we at least have a better understanding of what needs to be true for it to be a good or bad bluff. To me, it seems like not the worst double BW board to triple barrel because there is a decent amount of "snap folds" like Qx and missed draws but low PFR fish tends to underfold on both Axx and 2 BW flops to 2 and 3 barrels in my database overall. Low PFR fish is the worst player profile to fire multiple barrels against because they arrive postflop with stronger ranges and slowplay more postflop while at the same time not floating too much junk. If I were to bluff, I'd much prefer a smaller size like 62% pot or something to target more of the snap folding hands. I think that would work out better against a range with close to 20% 2pair+ but with a decent amount of missed draws but overall I try to not bluff a ton on boards where low PFR fish have many decent hands.

Oct. 2, 2020 | 9:56 p.m.

I'd probably mix flop vs that and not pure bet range. After betting flop, give up turn with most worse bluffs (you have a pretty good one) and always give up river. These players tend to get very strong ranges by the river on boards that hit them and then underfold because their range is so strong by the river.

I do bluff a ton vs weaker players to give some context but my work has found these players and spots some of the worst to bluff (passive, tighter rec on a board that hits their range well).

I strongly disagree with the notion that the guy is capped as his range is fairly tight to be a rec and he'll have close to all AQ, A4s, 44 in addition to some A4o, AA, QQ, and AK (yes, passive recs do limp-call AA, QQ, and AK some of the time). Low PFR recs just calls more of their nutted hands than other profiles too leaving more 2p+ in their river range than any other player profile I know of.

Oct. 2, 2020 | 2:29 p.m.

Oct. 2, 2020 | 9:05 a.m.

Comment | Kalupso commented on JD 2020

suggestions about improvements on either the technical or mental side

I'd suggest that you start trying some meditation. It'll help with emotions (we learn to not control emotions so I don't want to say emotional control), focus, disconnecting from poker, and overall happiness. There are many good resources out there but maybe the easiest way to get going is The waking up app

Oct. 2, 2020 | 9:04 a.m.

Do you have the VPIP/PFR and over how many hands?

Oct. 2, 2020 | 8:57 a.m.

Preflop rake?

July 26, 2020 | 6:38 a.m.

Fantastic video! I hope you can make more videos of this format in the future.

July 23, 2020 | 6:14 p.m.

Thanks and congrats to everyone!

June 4, 2020 | 9:58 p.m.

Comment | Kalupso commented on Is Zoom "real" poker

I haven't played on Party for the last 3 years but there are some differences between the sites. Zoom does play tighter than reg tables on every site I've played on.

June 4, 2020 | 6:13 p.m.

Comment | Kalupso commented on Is Zoom "real" poker

The title is another discussion... but zoom can be great for focusing on fundamentals. For me, most money is made at reg tables or in new emerging Holdem variants outside PokerStars but when I don't have a good game to grind I often play zoom and mark a ton of hands for review. Being able to play a ton of hands and take breaks for HH reviews at any time makes it a great learning tool.

The main downside would be that the games play much tighter preflop and there are more folding and less bluffing than you see at reg tables or live, so you learn to beat games with different player tendencies than live poker. It's also more difficult to develop reads and exploit individual players.

June 4, 2020 | 4:58 p.m.

Comment | Kalupso commented on BTN vs BB question

Need to overthink

Haha. Poker the game of overthinking things. I tend to do better when I just hammer away at stuff and try instead of overthinking and figure it all out in advance. Just try what you think is correct and filter your DB for win rate with groups of hole cards after trying it. Probably need like 150k hands to start to see a trend.

May 22, 2020 | 12:59 p.m.

Comment | Kalupso commented on BTN vs BB question

RaoulFlush I've done the sims and looked at where the EV comes from so it's not like I'm commenting out of thin air. Solver goes around 30% in my model with fish limping 40% of hands with a capped range. I use lower rake to simulate postflop edge. With strong limp/3bet strategies you see much lower isolation frequencies and more overlimping.

Against a 40% range you need like A8o, K9o, JTo, K3s,65s T8s to be ahead which puts you at around 30% of hands.

Against fish I really do like database data because it's difficult to weigh all the factors accurately. The model can also get weak when there are players left to act but I believe the model I used here didn't have any major flaws.

May 22, 2020 | 12:10 p.m.

Comment | Kalupso commented on BTN vs BB question

RaoulFlush I agree those probably open but about as wide as I'd go. I still think you miss one of the most important points which is that you get fewer preflop folds. Yes postflop is better than in solver or against a reg, but preflop is a ton worse. There isn't just one factor here (fish) but many factors that influence the EV.

May 22, 2020 | 8:01 a.m.

Comment | Kalupso commented on BTN vs BB question

HaAbDe We want to have some calls but still be quite tight because of the 4x size. Think hands like AJo, KQo, 22+, and good suited connected hands.

May 22, 2020 | 7:57 a.m.

Comment | Kalupso commented on BTN vs BB question

CO open limps who is a fish

Isolating a fish at 40%+ frequency seems very excessive to me but I'd be interesting to see database data on how it's working for you. I've been playing all kinds of games and sites the last two years so I don't have DB data and rely on sims and logic.

I do think you have some misconceptions though as close to all hands in your range mostly want a preflop fold so getting more action pre by a fish calling you generally reduces EV of fringe hands like A8o, K9o, K4s, and 76s.

May 21, 2020 | 5:09 p.m.

Comment | Kalupso commented on BTN vs BB question

I don't think the word defend is a good word to use or thinking about defending against anything. You just want to play the hands that are +EV to play and not defend against anything. The iso range should be substantially tighter than a common 40% BTN opening range, and the odds are much worse than normal opening size, and there is both the BB and the fish left to act. All of these things contribute to tightening up your range and increasing the requirements for how strong of a hand you need. I think iso ranges on the BTN will fall between like 20% and 33% on the BTN against a fish when BTN isn't over limping so 3betting like 8%-9% should be reasonable. Probably not a good idea to flat a 4x with both BB and fish behind that can reraise.

May 21, 2020 | 1:37 p.m.

Comment | Kalupso commented on Solver based redline

2) Rake

The baseline redline for the entire population as a whole is rake. If everyone were playing GTO, then everyone would lose with redline at the rate of rake.
Edit: This is wrong because all-in preflop and check down to the river both have negative net blueline for all players combined but like 65% of rake should be in the redline (WTSD less than 35%).

Doesn't bet size also have an effect on the redline and especially on the river? Example: You bet 1x pot with a bluff and you get 50% folds and this leads to winning the pot 50% of the time (EV 0,5 pots) as redline and losing pot 50% of the time in blueline. If we change to 2x pot and is now getting a fold 2/3rd of the time, we win redline 2/3rd of the time for pot and lose blueline at 1/3rd of the time for pot. The "redline EV" is 0,5 with 1x pot and 0,667 with 2x pot.

May 20, 2020 | 8:29 p.m.

hariadam No, it's a graph from CardrunnersEV. I just made a simple tool myself to use it with solver ranges.

May 20, 2020 | 9:08 a.m.

I looked at the Q2 hand at minute 6 and it seems fine. It's indifferent but because of double FD turn with many SDs there are many better bluffs to choose from so it's barely used by Pio. It seems like it's among the first hand you can look to add if IP overfold.

Strategy for A-high and highcards without a draw:

Same for Tc turn so it seems like the only reason Pio doesn't use it much on Th is that we have so many better hands to choose from:

This is the distribution of ranks/cards in the folding range on the Th turn (doesn't take into account suits):

There doesn't seem to be a problem with blocking folding range.

May 18, 2020 | 5:29 p.m.

I found the commentary especially strong in this video. Your thoughts flowed naturally and were insightful.

May 16, 2020 | 11:54 a.m.

Load more
Runitonce.com uses cookies to give you the best experience. Learn more about our Cookie Policy