BigFiszh's avatar

Moderator

BigFiszh

1825 points

Makes sense for me. Can't see the problem, tbh?

March 9, 2019 | 4:54 p.m.

Fancy play syndrome? And why is it „2x pot jam“?! According to your information it‘s pot. If you find 50% better combos, fold. Who should ever know if 99 is in there?! See first sentence ...

March 7, 2019 | 1:59 a.m.

Good thing u only lost $60...

Wanna bet on it? ;-D

March 7, 2019 | 1:52 a.m.

Instafold. Highly speculative, two short stacks (that will spoil any draw chances on the flop), non-nuttish hand with big potential to make 2nd best ... did I miss anything?

March 5, 2019 | 12:31 p.m.

Comment | BigFiszh commented on Blockers

You are sure that you did not mix up suits? T9dd is higher, not lower. But as you argument with higher backdoor flushes it seem that you actually meant "higher" (instead of "lower")?

Reason should be that the small bd-value makes it a better bluff combo than the others and the nfd-backdoors are blocked.

March 4, 2019 | 12:06 p.m.

Testing a given preflop solution is no rocket science! Model a tree for the given scenario (i.e. BB vs. BTN-open), add Q7s and Q5s and run a sim over 256 trees. Then create an aggregated report and compare the EVs for Q5s, Q6s, Q7s. Either you'll find an answer or the solution is bad ...

PS: If you don' t know what I'm talking about, I'd be willing to help out for free (against my coaching ToC :D), as this one interests me, so just PM me.

Feb. 26, 2019 | 4:55 p.m.

ot considering exploitative adjustments in poker is extremely counter productive and misleading

Agree, but then we don't use Pio. We work with Sims (and Nash) if we don't know our opponents. And analyzing if we would sacrifice a lot of EV (in terms of exploitability) with simplified strategies is a very intelligent way of working.

It is as well depending how often that spot occur and how small pot is.

First YES, second. NO. :)

Feb. 23, 2019 | 4:16 p.m.

When OOP 3-bets, IP has a bunch of hands that can't call. They either got to get raised - or folded. If IP is not "allowed" to raise, all those hands have to be folded, which is a great EQ win for OOP, as IP loses his equity share for those hands. Should lead IP to bet less though (and save EQ share more often).

Feb. 17, 2019 | 2:12 p.m.

The spot is ultraugly, definitely. I'd rather rule out AA ... that would not really fit (flat 5-way pre with AA?).

Reasons why this did not get much love yet likely is that it's extremely difficult to assess from remote, w/o any reads or additional hints.

You say "recreational" player, which likely means, he will NEVER (ever) bluff. That means, we got to ask which combos he might overplay? Do you feel he's capable of risking 400bb with 77? 88? Would he play 98o pre? Did he signal anything with his action? Did he shove fast, hesitantly, happily, thoughtful, shy ...?

Feb. 13, 2019 | 9:46 p.m.

I'm sure there are people that read the thread and didn't respond to it, simpy because there is not much to add to the discussion

Broad statement! ;-D

I've just stumbled over this post and feel challenged. I've done (and still do) dozens of database analysis for my coachees (/ad :D). This just ahead to show where I am coming from.

OP did not ask for how to "fix" those leaks, he primarily asked for how to identify those. And actually this is quite easy - even with limited knowledge about theory.

For example, take the cbet defense% oop in SRP. It's X%. Is that too much? Too little? Just about right? Should we raise more - or call more?

The answer can be found pretty simple - by intelligently using filter functions:

1) Define a baseline for this scenario. That means, setup a filter for all comparable situations (SRP, marginal hand, 95 - 105 bb deep, etc.) where you found yourself oop facing a cbet and folded! Now check your loss in bb/100. This is the one decision we can always fall back to.

2) Next, substitute the "did fold" filter by a "did call" filter. Check your win/loss-rate again. Is it better than folding? In that case, you fold too much!! Simply said, because in any single case where you had called instead of folded, you would've gained money.

3) Next, substitute the "did call" filter by a "did raise" filter. Check your win/loss-rate again. Is it better than folding? In that case, you again fold too much! Is it better than calling? Then you're calling too often and not raise enough.

Simple, right? Beware though, these are NO fixes yet! Knowing that we fold too much, call or raise too much / too little is just a leak so far. We still don't know what the fix is. This is comparable to a Nash solver that can precisely tell us how far he is away from the Nash strategy, w/o having a clue how to get there or how the Nash strategy looks like.

Fixing it needs a lot more theoretical knowledge, comparison of proven winners, working with a coach or simply discussing here in the forum. But identifying leaks just needs a bit of craftmanship.

BigFiszh

PS: A "word" of warning along - regarding sample size!!

Be aware that scenarios ideally have to be fully comparable. If you find yourself with A-high + mid overcard on a 7-high board in a SRP oop with 100bb effective stacks, BB vs. MP, you should filter for that exact scenario, same holecards, same board structure, same positions, same eff. stack sizes.

It's very likely though that you quickly get to a sample size that just is insufficient! Don't ever underestimate, how much variance can influence results, if we speak about very tiny fractions of stacks.

Example: you have a sample of 1,000 hands in a given scenario. Accidentally, in one single hand of this sample you won or lost one entire stack (not visible in EVbb). 100 bb won or lost makes a difference of 200bb. Over the given sample this makes a swing of 20bb /100! This can easily lead to completely misleading interpretations.

Obviously this is even magnified if we talk about scenarios that are uncommon in and of itself, i.e. x/r as 4b oop.

The solution is easy. S-A-M-P-L-E-S-I-Z-E!!!

100,000 hands is not a database to work with seriously at all. 200,000 hands allow very basic scenarios with broad filters. 500,000 hands let you analyze most flop scenarios and a couple more complex things. 1,000,000 makes it interesting to dig deeper, up to the river. 3,000,000+ starts allowing analysis for unusual lines, i.e. x/r on the river and stuff.

Now, I hope I actually did contribute something substantial to the discussion.

Feb. 8, 2019 | 10:42 p.m.

Exploitability is most likely secondary. EV difference between pure mixed strategy and pure 66 / 33 strat is almost negligible. That's not the reason.

Intuitively the results should not surprise though. Say, Villain bets 0.01% of the pot. Would we x/r our entire range now? No. Why should we? Villain could just check behind, so it's kind of "dumb" to punish him. We deny ourself the chance to just call and see a very cheap turn ourself.

What does that mean? => When V. bets small, we are way more incentivized to call! Very few weak hands are weak enough to not at least call. And for many strong (non-nuttish) hands, calling in this scenario is better than raising as well. That means, the raising part shrinks significantly. When he bets big though, we are less motivated to call. All "marginal" hands either get flushed down the toilet - or will be taken as bluffs. So, we call less, fold more and the rest will be raised (polarized).

Makes sense?

BigFiszh

Feb. 8, 2019 | 10:08 p.m.

You should do two things:

1) Make sure that you work with a big enough sample size.

2) Set a reference point: filter for all river spots, where you folded instead of calling (so all filters are identical, only the very last should be different). Then note, what your EV is. THIS you should compare with calling. If you lose more by calling - you identified a problem.

Jan. 20, 2019 | 5:35 p.m.

Comment | BigFiszh commented on 50z is this a cooler?

Happy new year. :)

If you say, there won't be any bluffs, why do you include AQo? JJ? Do you think those are value-4-bets oop against UTG+1 open + BTN-3-bet?

Jan. 1, 2019 | 8:41 a.m.

Comment | BigFiszh commented on 50z is this a cooler?

What range do you assign to Villain?

Dec. 30, 2018 | 5:14 p.m.

Comment | BigFiszh commented on 50z is this a cooler?

MOD: Moved to Low Stakes

Dec. 30, 2018 | 5:12 p.m.

Mostly irrelevant. EV for "check" includes x/r, so if x/r would yield some higher EV, it should show up in the check-option.

Dec. 29, 2018 | 10:17 p.m.

I'd like to put down some thoughts, that my students already know by heart. :-) And believe me, I don't want to belittle you (or anyone), and if my words seem harsh, it's completely unintended. It's due to me not being a native English speaker and due to my passion for the topic. :-)

OK, that said: solvers seem to spark the phantasy that nowadays the "hard and dry theory" can be spoonfed by simply downloading some prepared charts.

If it were that easy, to just download some charts and learn those by heart - how can anyone expect that poker could still be worth it? Because there are still some players left that simply don't know where / what to look for the "solution"? Unlikely, huh?

I won't get tired of stating the following:

A solver is a pocket calculator!

Nothing else. If you (seriously!!) want to learn about algebra, you won't use a pocket calculator either, right? And you wouldn't download some charts and learn them by heart. Instead, you'll take the hard route, grab some (good) books and learn about the basics. Once you've managed the basic stuff, you'll take a calc to get the more complex stuff done. Still, your intention of using it is to save some time when taking the next step, not to learn anything from the results!

Just my 2 cents. Maybe you (or anybody else with a similar mindset) wanna think about it. :-)

BigFiszh

Dec. 19, 2018 | 1:14 p.m.

Comment | BigFiszh commented on PIO on Mac

Got a PC at home? Install Teamviewer (or any other remote control software), run Pio on the PC (let it turned on while you're travelling) and remote control it from the Mac.

Dec. 1, 2018 | 6:08 p.m.

Comment | BigFiszh commented on PIO on Mac

It's not only easier to run - Pio ist a heavyweight resource consumer. Installing it on a windows platform on a mac would be like driving a sports car on a car transporter. :)

Nov. 24, 2018 | 6:16 p.m.

Agree with everything my foreposter said.

Additionally, ignore your "3-stack-read". :) This might be correct in many times, but it's completely worthless. Even moreso, as labelling someone as "fish" or "no fish" is worthless in and off itself. You can / should derive absolutely nothing from this assumption. It will just confuse your mental decision process. Go with specific reads (might be pool reads as well) - or ignore them at all. Just my 2 cents.

BigFiszh

Nov. 24, 2018 | 8:13 a.m.

Hi buddy, welcome to this playground. :)

HUD reading is quite difficult, not only in general, but even more as most have severely individualized their HUDs for their own special needs. Without any explanation (or abbreviations at least), it's almost impossible to fully understand any stat that will be shown. Most will - inevitably - show or even start with the basic stats, namely VPIP (how often does player get involved voluntarily, namely not by just paying the blinds and then see a flop or win preflop), RFI (how often does play raise first in, namely opens the betting round preflop when all have checked to him / he's first to act), 3b (how often does player 3-bet vs. open-raises) etc.

In general I'd suggest to "ignore" the HUD stats in the video. If it were important what you see there, I would expect the coach to explain how his decision is based on what can be seen on the HUD.

Does that help?

BigFiszh

Nov. 24, 2018 | 8:07 a.m.

With no stats given, I don't see how we can fold this. Like ever.

Obviously it "feels" ugly, but that should not affect our decision. If the "feeling" is valid, we had made mistakes elsewhere (i.e. by 4-betting preflop, betting too big / betting at all on the flop), but once we take the line up to the flop bet, I'd say our fate is sealed. What do you ever wanna call down with? QQ only?

Nov. 18, 2018 | 4:33 p.m.

Agree with Resolve on preflop!

Disagree about last sentence postflop though. Turn is a pretty "easy call" as played. The headaches connected with that root from preflop. If Villain has ANY range that justifies the preflop play, betting flop (bigger though) and x/c turn seems valid.

Nov. 15, 2018 | 7:40 a.m.

Yeah, obviously we HAVE a raising range, although it's very tiny. But fun to play with TT IP in that spot. :)

Nov. 2, 2018 | 9:03 a.m.

That's just not correct. Positions will shift, but ranges overall should get loser.

Contra-example: There's no blind at all. What now? You'll play 0% - from any position. That contradicts your idea.

Oct. 19, 2018 | 5:39 a.m.

No reason to bet less than all-in on the river imho. He reps a set, not sure if 2-pair should even raise the turn. Every valuehand that raises the turn should be close to nuts, so less than shove makes no sense (in theory).

That said, his riverbet leaves me kind of clueless. It might be that he tries to incentivize you to call, it might be that he's just not aware of the math stuff. Or he might try to "cheaply" continue his bluff. Which again, makes no real sense in general.

Oct. 18, 2018 | 5:47 p.m.

It's completely okay that deltaEV is negative!! That's the "right" of the aggressor. B's only target is to keep the "negative number" still as big as possible. That's all he can do.

Oct. 18, 2018 | 5:36 p.m.

Hey buddy, you're partly right, partly wrong. :) Right in having caught me, my explanation is a bit fuzzy and in parts simply wrong.

Wrong in claiming that the book is mistaken.

But as I am coaching that stuff, let me restore my honour and take another attempt. :)

So, grab a cup of coffee, free your mind and fasten seat-belts, we are starting the engines …

There's 3 bets in the pot. Player B has 20% equity (hot-cold), means, if it's just checked down, he will win 0.2 x 3 bets. That's the baseline. Let's call it "calculated EV".

Now, player A starts betting. He starts with value-only and then adds more and more bluffs.

Player A wants to develop a strategy where - whatever player B does - the profit for A gets maximized (and minimized for B).

OK?

Let's call the EV that results from B's actual strategy the "realized EV". So, A's target is to maximizing the difference between calc.EV and real.EV.

Now, if B would fold anytime A bets, B would "lose" the pot anytime A had actually bluffed. In terms of current EV B's EV when folding is zero, but actually he gave up his share, which had been +3 (the pot), in case he had 100% equity (against a bluff). And here we are not talking about gaining even more, if B had called the bluff - we are simply talking about B not winning the naked pot which he originally was deserving.

That means, B's loss when he folds is 3 times the percentage, A bluffs. If A never bluffs, B's loss is zero (0% x 3 = 0). A valuebets 20% (B folds), A never bluffs - so B everytime takes down the pot, when he has the best hand. If A bluffs 2%, B will "lose" (= give up) 3 bets in 2% (as he always folds to a bet) for a total of 0.06. And so on. Still with me?

That leads to the first formula. As long as B sticks to the "always fold"-strategy, his EV-difference between "realized EV" and "calculated EV" would be (bear with me that I will substitute the "X" from the book by "B" - as bluff percentage - as it would otherwise collude with the multiplication-sign later on):

3B (-> according to 3x in the book)

We can "prove" that by subtracting "calc.EV" from "real.EV":

real. EV = (1-0.2-B) x 3 (-> The bet% of A is 20% + B, and as we are always folding, we only win the pot when A does not bet, which is 100% - 20% - B)

calc. EV = (1-0.2) x 3 (-> B's original share - namely hot-cold-equity = 80% times pot)

delta EV = (1-0.2-B) x 3 - (1-0.2) x 3 (-> simple subtraction)

Now we do some formula work to simplify, which eventually leads to:

delta EV = (3 - 3x0.2 - 3B) - (3 - 3x0.2)
= 2.4 - 3B - 3 + 0.6
= 3B

Nice, huh? It gets better! :D

OK, that was the warmup, let's go to the more complicated part. What happens if B always calls?

B wins the pot, when A checks, wins the pot + one bet when A bluffed and loses his bet when A valuebet.

Let's put up the entire EV formula:

EV (real.B) = ((1-bet%) x pot) + bet% x (EQ x (pot + 2 x bet) - bet)

Now let's define some of the variables before we continue:

bet% = V + B (-> bet% = value% + bluff%)
EQ = B / (V+B) (-> B's EQ is the ratio of bluffs to total range)
V = 0.2 (-> 20%, the value% of player A)
pot = 3
bet = 1

OK, here we go:

EV (real.B) = (1- 0.2 -B) x 3 + (0.2+B) x (B/(0.2+B) x 5 - 1)
= 3 - 3 x 0.2 - 3B + (0.2+B) x 5B / (0.2+B) - 1)
= 3 - 3 x 0.2 - 3B + (0.2+B) x (5B / (0.2+B)) - ((0.2+B) x 1)
= 3 - 0.6 - 3B + 5B - (0.2 + B)
= 3 - 0.6 - 3B + 5B - 0.2 - B
= 3 - 0.8 + B
= 2.2 + B

Now, what's the delta.EV?

Remember, delta.EV = real.EV - calc.EV:

delta.EV = (2.2 + B) - ((1-0.2) x 3)
= 2.2 + B - (3-0.6)
= B - 3 + 0.6 + 2.2
= B - 0.2

Now - substitute my "B" by "X" (as it is used in the book) and you'll recognize the formula from the book. Yieeehaaa!

So, what you have now, are (brutally) shortened formulas to calculate the EV-loss for B, that A's bluff-strategy results in, for each counterstrat of B, always calling or always folding.

Obviously, B will always take the strat (call or fold) that yields the highest EV in, depending of the bluff% of B. That leads to the strat-switch at 5% bluff frequency.

And against B optimally defending (switching from folding to calling just in the right moment), a bluff frequency of 5% leads to the biggest delta.EV (= EV-cut) that A can hope for.

And that was what the authors wanted to proof.

Everything "clear" now? :)

BigFiszh

PS: Everybody who reads this line - and did not skip the rest - earned my serious respect! :D

Oct. 18, 2018 | 4:19 p.m.

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