I'm always interested in learning more about how to play (any of the three games) but I don't understand what you mean by swapping "strats." How is this different than just hanging around here discussing strategy and asking questions?

One thing that IMO would benefit this forum would be more theoretical discussion of how to assess certain situations, perhaps using a range of related situations, rather than everybody insisting all discussion has to be about one real-life hand. Here's what I mean. To take an example at random, "how to play backdoor low draws."

To an extent it's just a math problem, but there are also some parameters and assumptions that change how you think of the math problem. The naked BDLD will come in (24 * 20) / (45 * 44) or 24% of the time, but just having 8 bets in the pot (getting 4:1 on half) isn't enough. Getting quartered occasionally really hurts your odds on a longshot, and often you'll have to put in a turn bet only to fold on the river.

So the value of a BDNLD depends greatly on its equity besides your nut low draw. Do you have any remote outs to win high? Do you have a backup low card that could win low if you get counterfeited?

Let's make up some action....

2 limp, we raise button with AJ92, blinds call, limpers call.

Flop (10 SB)

LAG BB leads out, one fold, two calls.


The value of our hand is a lot different on:

3QT
3QT
6QK
3KT
69K
69K
and so forth. How would each of those change the value of our draw?

?

back door low draw

i question the equation not the abbreviation.

it troubles me that no one else seems to.


in 2 of the 3 circumstances where you may encounter a BDLD, 24% is an adequate approximation for the percentage of board run-outs (turn river combinations) when you don't account for an opponents range however it is never arrived at by the equation (24 * 20) / (45 * 44)

additionally, 24% isn't the percentage that applies to the subsequent discussion within the post.

It's best to express that using words, not punctuation.

Silence is not acceptance. There's generally someone on 2+2 chomping at the bit to jump all over something like this. I'm surprised it took so long.

It's because ngFTW is the only one here that has any understanding of mathematics.

Local casino's rake structure for a $5/$10 LO8 w/ full kill is: $1 on flop, $2 at $10, $3 at 20, and $4 (cap) at $30. Does this sound standard?

I like the $4 cap, but it hits pretty much every hand...

It's good to troubleshoot our own thinking so I'm trying to troubleshoot mine here. If we have naked A2 (or A3 if that's BDLD), all else equal:

six nonpairing low ranks remain before the turn
24 cards remain of those low ranks
seven cards are visible in Hero's hand or on the board
45 unseen cards remain

Prob(nonpairing low rank on turn) is 24/45

If that should happen:
five nonpairing low ranks will remain before the river
20 cards remain of those low ranks
eight cards are visible
44 unseen cards remain

All else equal, prob (nonpairing low rank on river conditional on nonpairing low rank on turn) is 20/44.

P (A intersection B) = P (A) * P (B | A)
P (np low turn and np low river) = P (np low turn) * P (np low river | np low turn) =

(24 / 45) * (20 / 44) = (24 * 20) / (45 * 44)


I must admit, I still don't understand what I'm missing. I hope someone will take a shot at providing a real explanation so I can learn from the holes in my own thinking.



*I'm not considering card elimination from ranges we expect opponents to be playing, etc. (How much that matters is a separate interesting discussion.)

What does that have to do with anything? I was trying to give an example of how to discuss a generic theoretical topic and then move to specific synthetic hand examples. Clearly it wasn't a very good example. Sometimes one misfires.

It's up to you how you want to allocate your time, but your posts would be more meaningful if you'd bother to add a sentence or two explaining what you're referring to occasionally.

to have a backdoor low draw, BDLD, requires that your hand consists of atleast 2 low ranks and that the board contains exactly 1 low rank.
a deck consists of 8 low ranks.

with exactly 2 low ranks in your hand and 1 on the flop, the turn-river runouts that complete a low are the combinations of 5 non-pairing low ranks and the subsequent 4 non-pairing low ranks.
10 combinations of 16 each =160 combinations of the total 990 possible board runouts (where order doesn't matter) (and opponent's range is not taken into account). 16.12%

with exactly 3 low ranks in your hand and 1 on the flop, the turn-river runouts that complete a low are the combinations of the 4 non-pairing low ranks and the subsequent 3 non-pairing low ranks plus the combinations of the 4 non-pairing and pairing one of the 3 low ranks in your hand.
6 combinations of 16 each + 12 combinations of 12 each =240 of 990 board runouts. 24.24%

with exactly 4 low ranks in your hand and 1 on the flop, the turn-river runouts that complete a low are the combinations of the 3 non-pairing low ranks and the subsequent 2 non-pairing low ranks plus the combinations of the 3 non-pairing low ranks and pairing one of the 4 low ranks in your hand plus the combinations of pairing 2 of the low ranks in your hand.
3 combinations of 16 each + 12 combinations of 12 each +6 combinations of 9 each =246 of 990. 24.84%


when you hold A2J9 and you have a BDNLD and your opponent can have any hand, you have a ~16.2% probability of making a low at showdown. consequently, 8.1% is the upper bound of your avg. pot equity from low. if pot odds are such that you need 26% equity to make a plus EV call, you need atleast 16% equity from high.


i would think if you've glanced at what i posted above you will have recognized your error, but as it seems you think i should be very explicit in my posts,
is wrong. its 5. 8-2-1 =5
5 nonpairing low ranks
20 cards remain of those low ranks
is wrong. its 4. 8-2-1-1 =4
4 nonpairing low ranks
16 cards remain of those low ranks

^^^ Should note that in a more realistic scenario, your true low probability/equity is a bit lower even.

Actually that one didn't seem to deviate much but them random trials tho :/

^^^

select count(madeLo(hero, river))
from game="omaha8", hero="A2J9", villain1="[a-5][a-5][2-k][3-k]", board="[9-K][9-K][3-8]"
Results:
Trials COUNT 1
600000 97039 (16.17%)

IMO abstraction and intuition are probably better concepts for a newcomer, rather than the raw math anyway

now with basic alternate reads sorta, def not perfect but at least not any 4 lol

low-ish suited/connector player image:
select count(madeLo(hero, river))
from game="omaha8", hero="A2KQ", villain1="[a-5][2-5][3-8][4-k]", board="[9-K][9-K][3-8]"
Results:
Trials COUNT 1
600000 93308 (15.55%)


sorta high-ish image or maybe decent player but more of a plo/nlhe background(?):

select count(madeLo(hero, river))
from game="omaha8", hero="A2KQ", villain1="A[2-5][t-k][t-k]", board="[9-K][9-K][3-8]"
Results:
Trials COUNT 1
600000 108398 (18.07%)

Appreciate it. I had a brain fart and got six remaining ranks instead of five. I probably just forgot to eliminate the one on the flop. That's all I was missing, but I appreciate the detail.

I hadn't thought about solving it with combinations (permutations seems more straightforward to me even though the numbers are bigger) but both give the same answer

(20 * 16) / (45 * 44) = 0.1616 repeating = 160 / 990. Not a wrong method I was using, just a messed up parameter. No big deal.

So if I have A2 on a 345 board for the nut low and wheel. If a 2 comes on the turn do I still have a low?

when you get counterfeited you always still have a low, just generally no longer as good of a low.

a wheel is the best low, so you will still always have the nut low on future streets. however, after the turn, A3/A4/A5 are also nut low.

The only thing you need to worry about when this happens is whether or not your high hand is good enough for the high half of the pot.

Frequently when 4 or 5 to a wheel appears, the 5 high straight will not be enough.

The only reason this is a problem is that it is very easy for someone else, sometimes even multiple other players, to hold the same low (54321) as you. The essence of that problem is that you will lose chips unless you hold or tie the best high hand.

When you tie one half and lose the other and the tie is with one other player, this is called getting quartered because you receive a quarter of the pot.

Sometimes you tie with two players and this is called getting sixthed.

In fixed limit games quartering is of little consequence and little should be done to avoid it except in very extreme situations.

In pot limit or no limit games it is quite devastating because you will lose half of any bets and it can mean losing half your stack when quartered and two thirds when sixthed.

Hope this is insightful.

Sent from my GT-I9505 using Tapatalk

True but more generally people in O8 use the term applying to either made hands or low draws. So if you have A2QQ on K83 and a deuce turns, you'd generally say your low draw got counterfeited. But this statement is 100% correct when it's already a made low that subsequently gets counterfeited.

"A wheel is like a college degree. They can never take it away from you...."



Critical point. Beginners, take note!


I'd amend that--certainly considering folding is rarely of any consequence. But getting quartered or not may factor into your decision to take aggressive action. (I know you know this, Lucius.)

Most often, mediocre players miss opportunities to win the high half of the pot through aggression because they're terrified of getting a fraction. Sometimes the other quarter will scowl and berate you for losing you both money. Laugh it off in a way that implies you're terrible at O8 and didn't know what you were doing.

(Don't you ever be the one berating.)

I can cook up some examples of this theme of "using our nut low, even a quarter, as a foundation to help us win high," if anyone would find it helpful.

Never forget--in limit poker you happily "waste" bets for a decent chance at winning pots, or half-pots.

you never lose 2/3 of your stack with a winning hand. with a winning hand your stack at most gets reduced by 1/2.

Look to start a bankroll 4-8 live full kill . To play 20 hours a week is 2000 a good roll to start

How can we possibly answer a bankroll question if we don't have an estimate of your winrate?

I will say this--unless "grinding up a bankroll" has non-economic value for you, you would do much better to practice a bit at 4/8, set aside money (or borrow if you're comfortable), hire a coach, make very sure you're playing well, and play at somewhat higher stakes when games are good. 4/8 is just not a good use of a sharp player's time. Full kill helps a lot though.

What stakes are available in your location?

8-16 and 10-20

OK, what's your motivation for wanting to "start a bankroll"? For your goals, how does that compare playing a little 4/8 to get your feet wet, then (with appropriate controls to make sure you're +EV!) jumping up to 8/16?

I am a little confused here: I thought I scooped a pot but only halved it.

My hand:
A65A

My opponent's hand:
4T23

Board: J7289

I thought my high was A6 for the flush and my low was A5 for the A2578 low, but my opponent won with 23478

Is the Ace not counted as the low hand here?

Low hand should be read from the highest card down so you both have 87 lows but his next lowest card is a 4 (vs your 5) so his 874 is lower than your 875.