$25----->25k Bankroll Challenge on Ignition

What part of my post does this HH refute? 😀

You had 0 equity on the river when you got all in, thus both lines went the same way. If one went a way and the other went another way, then my understanding of EVbb is flawed.

Trackers can't calculate EV vs range, only hand vs hand.

If I jam turn my EVBB goes through the roof.

It doesn't even out is my point.

I think tradition has always preferred EVBB but I've never been given a satisfactory answer as to why that is.

I don't want it to be true at all, I just came across it on X. I wasn't hunting for it.

It was probably a mistake to post it in retrospect but you reacted as expected.

If you would never jam the turn with the AJhh under any circunstances, but he would, I 100% agree with you. There are preflop all-in spots where you may have KK, villain AA, where the reverse can definitely happen and it will even out over the long run.

yes it should be expected to call out people who attempt using their platform to conspiratard - thanks for acknowledging that

I was actually trying to have a legitimate discussion but it was my mistake, I should of known that the people quickest to reply to that would also be the least likely to do so in earnest.

That's Game Theory.

lol, the copium is real

you spouted nonsense and got undressed, just own it

Everything I said is factual. Show me where it isn't.

But you really seem to care about winning forum battles. I actually come here to learn and help other people learn.

I'm still waiting for you to post something that helps people, everything you post is self serving.

Why do you use EVbb and bb/100 for those different calculations?

I think EV loss vs regs would be a great metric and then maybe EV win vs fish. Do you think that would be realistic to implement into a DB/HUD.

Not sure if we were thinking the same, but if you have the ranges and strategy of the average player, via MDA in a site with known holecards like Ignition, I assume you can get a version of what your true EV would look like? What do you think about it?

I think you would have to implement something like GTO Wizard's training program where you play against a solver and then compare it how you are playing spots.

GTO Wizard should make a tracking a program like this, I'm going to email them to see if what I am asking is possible.

Also I asked this question in the Theory forum as well.

Link here:

https://forumserver.twoplustwo.com/15/po...

One specific thing I just remembered is that a particular stable was victim of a collusion ring in apps a few years ago, and the graph of the students was consistently below EV. In that case, EVbb obviously was not measuring the reality, not at all, because the other players were sharing holecard information.

Do you mean seeing your frequencies from drilled hands and what “type” of mistakes you usually make?

I think in this instance jamming AA and getting called by AJ will obviously result in our EVbb being much higher than actual bb since we know the river saves him, but on many other runouts our EVbb is even higher by calling since we stack him on brick rivers (without the flush draw taking from our EVbb on turn). And I think it makes more sense to think about the EVbb/actual bb of our hand vs their entire range, jamming turn loses ev against their low eq bluffs but obviously is higher ev against his pair+draw and high eq hands like AJhh. I think it still mathematically evens out tho I don't have a rigorous proof for it hahaha

EDIT: actually sorry I'm over complicating this, even in this case where you jam AA, your EVbb and actualBB is still the same and both will be high, suppose his draw gets there 20% of the time for sake of argument, once he calls his EVBB is 20% of potshare and actualBB is the same in expectation (since we get full pot share 80% of the time and lose 20%), so I don't there is some way to magically diverge EVbb from actualBB

Yes something like this but without having to drill it yourself, so it is a filter built into the software. That would just be the first step. I'd also like a filter to see the average negative winrate of fish and then my specific winrate vs them.

My point with the HH is it is always going to get all in regardless correct?

But if I jam flop I'm 95%/I jam turn I'm 80%/I call river I'm 0%. So EVbb and bb/100 are very far apart based on my exact line.

They’ll be the same on average, but ev bb will have no variance

So lets say we always jam turn in this line, we can assume villain will always call, but we need to consider that we don't know what the river card is going to be

If we jam turn instead of calling turn, in this particular instance yes there is a huge gap between evbb and bb/100 because the river makes a flush and villain coolers us, but on 80% of other runouts bb/100 will be higher than evbb since we would win with AA on a non flush river and take down the pot instead, so over enough trials the evbb and bb/100 of jamming turn will still converge

I think the problem is we are assuming the river always coolers us here but we have to think about the aggregate of all possible rivers, correct me if im misinterpreting your argument here though haha.

If we always call turn, then the EVbb line and bb/100 line still stay together since on river there's no more cards left before showdown

Yeah so the "enough trials the EVbb and bb/100 of jamming turn will still converge." I know during the pluribus bot experiment Noam Brown and company made some sort of filter to simulate more than the 10k hands that were played. We would need something like that but for variance.

I just used the AA hand as an example because I was thinking about this subject today. My overall point is we have been using the same old metrics since the inception of online poker and I was wondering if any software personality types had any idea of how we could improve upon it.

I will need to re-read Noam Brown's paper to ask better questions. I won't pretend to understand all of it but is it possible to basically combine a solver into a DB for a home pc?

Paper here if you want to take a look:

https://arxiv.org/pdf/2007.13544

I think you would need an expert in computer software but also an expert in statistics to understand what type of variance shortcuts you could reasonably take get a better metric than ev BB.

This is a bit outdated, might want to take a look at this - https://arxiv.org/pdf/1612.06915

No matter what though it's impossible to model since table conditions change hand to hand each one is a different spot. evBB is a better estimate than BB but there is literally no way to know and no point in sweating over it since your goal is to increase your winrate over time anyway.

I'm trying to learn this node since these blitz games play deeper. River is a jam or fold in a solver!

Hand History driven straight to this forum with DriveHUD 2 Poker Tracking Software

NL Holdem 0.25(BB)
BTN ($42.47) [VPIP: 19.1% | PFR: 16.3% | AGG: 31.6% | Hands: 376]
HERO ($32.89) [VPIP: 27.6% | PFR: 23.3% | AGG: 39.7% | Flop Agg: 43.9% | Turn Agg: 36.3% | River Agg: 40.7% | 3Bet: 10.2% | 4Bet: 12.2% | Cold Call: 9.5% | Hands: 72850]
BB ($31.30) [VPIP: 21.1% | PFR: 18.1% | AGG: 34.9% | Hands: 385]
UTG ($34.85) [VPIP: 19.2% | PFR: 19.2% | AGG: 23.1% | Hands: 81]
HJ ($26.62) [VPIP: 18.9% | PFR: 15.4% | AGG: 18% | Flop Agg: 11.1% | Turn Agg: 26.5% | River Agg: 19.4% | 3Bet: 6.8% | 4Bet: 11.5% | Hands: 729]
CO ($38.01) [VPIP: 24.6% | PFR: 21.9% | AGG: 27.7% | Hands: 545]

Dealt to Hero: 3♥ 3♠

UTG Folds, HJ Raises To $0.50, CO Folds, BTN Folds, HERO Calls $0.40, BB Folds

Hero SPR on Flop: [20.9 effective]
Flop ($1.25): 9♥ K♣ 5♥
HERO Checks, HJ Bets $0.37 (Rem. Stack: $25.75), HERO Raises To $1.66 (Rem. Stack: $30.73), HJ Calls $1.29 (Rem. Stack: $24.46)

Turn ($4.57): 9♥ K♣ 5♥ 9♦
HERO Checks, HJ Checks

River ($4.57): 9♥ K♣ 5♥ 9♦ 4♣
HERO Checks, HJ Bets $3.88 (Rem. Stack: $20.58), HERO Calls $3.88 (Rem. Stack: $26.85)

No you weren't. If you were, you wouldn't be posting a timesofINDIA article. You're also a month late and this has been debunked with proof but that is besides the point.

You should apologize to Rickroll, as insufferable as he is, he's obviously right.

I think tradition has always used both with BB/100 being the more ''used'' stat, as you should. Not relevant if it satisfies you or not.

Good discussion about this topic over here if people are interested.

https://forumserver.twoplustwo.com/15/po...