Tilt is death in PLO, even more so than in NLHE. People are going to suck out on you. They're also going to hit weird runner-runner draws and you'll pay them off. You just can't tilt when that happens.

Honestly, I didn't misread your post. As I said (I won't bold my own words), we apparently play in different games. I rarely see HU pots: a single raise goes eleventeen ways, and a 3bet pot is usually still 3+ handed. So the idea of having 65% otf is unimaginable. I might have 35% in a 4 way spot and can ship my $300-500 stack in a 3bet pot, that's great in a vacuum but it ain't low variance.

confirmed

This.
If your 3! Pots are heads up your in the world's worst plo game

Sent from my SM-G935V using Tapatalk

Yes, you did. You talked about set value in plo which is not what i was talking about at all, and i clearly corrected for you. I also no where said heads up pot. Ive never seen a heads up pot in plo.

The fact remains that you can and should nut peddle in plo, and most often if you flop the nuts you have >60% single street equity.

Im done explaining.

Is it worth it to continue talking about the difference between variance and downswings? Being able to frequently get a lot of money in as a favorite leads to a big edge, which helps mitigate downswings. Frequently putting in lots of money also increases variance. Variance by itself only describes the range of results you can expect, not how much or how often you can expect to lose.

It doesn't work that way. Folding pre is zero EV and zero variance. Anything but folding will have positive variance. Hands where you fold pre still count as part of your sample unless you track hands in a very non-standard way.

If you play 5 all in pots with 70% equity or 50 pots with 70% equity at which point should your results be closer to EV? Obv are both lol sample size but my point is the more pots you play the bigger your sample size is and the more your results regress towards the mean.

Decreases variance was the wrong choice of words, I should've said increases sample size, so variance will play less of a role in your results.

I would disagree with this. Or at least, you don't feel the variance as much because you get through it quicker. But this is as long as you can still keep your edge. In 3 hours, If 1 guy has 3 all ins with 60% equity, reasonable chance he loses overall. If in 3 hrs, another guy has 10 all ins with 60%, much less chance he loses, as I'm sure you know.

4 way (as an example) to the flop, and you have over 60% How often do you think you're getting this spot? Once an hour, once a session? You have to contort the board pretty hard to make an example.

I done pretending you can explain.

He's talking about flop>turn.
Youre talking about flop>turn>river.

PLO is actually much, much easier to play and the difference between how people play and how they are supposed to play is much bigger. The fish are fishier. They are either nut peddlers who dont get the nuts often enough in hold em or they are the ppl who want to play every hand and will be wiped out in nlhe

There are multiple inflection points in Omaha so you can get away with playing solely to control variance (flatting AAxx from the SB, for example, never 3betting) and always take pot control lines, so your style can define your BR.

So, inspired by Angrist's charts, I decided to take my data and play with them in Excel yesterday. The key part is choosing the best trendline and I am not sure I did, so based on the feedback I get, hopefully, I can make the second versions of those charts more interesting.

First, because there was some conversation about how one's play is affected according to how many hours he's been playing, i did a chart on big blind profit based on game duration.



Excel offers 4 ways to do a trendline. Linear, logarithmic, polynomial and moving average. On this chart I used the polynomial because it sort of tells a story. As I play more hours I tend to win more. Then as I reach 10-14 hours my profits fall, only to rise again on longer sessions.

Since there are so few data after the 10 hour mark, I don't think cannot draw any safe conclusions. The data is too unreliable.

Thinking about game duration's impact on profit however, i think this is the crux of the matter.

Do I win less because I play long hours and I am tired, or do I win less, because I am stuck and I play longer hours so as to get unstuck. A lot of posters have said that it's the former. I tend to think it's the latter. It's far easier for me to show discipline and get up after 8 or 9 hours. I get more stubborn trying to make a losing session a winning one.

Then I did a chart that showed my profit in blinds by session.



This all looks pretty random (and that gives a bit of a perspective about what you should expect when you visit the casino each time) save for the fact that as time goes by, I tend to have sessions with larger wins and losses. That's probably misleading however. In the beginning of those results, I tended to play shorter sessions; while the past couple of years the tendency is to play longer ones.

That's why, I divided session profit by session duration to get an hourly BB/hr rate per session. This gave me this:



The discrepancy between the early and the late period isn't as big as it is on the previous chart. However, the trendline does show that as time goes by, average profit tends to increase. Hopefully, that reflects the fact that I have been improving as a player.

What I wanted to do is do the thing Angrist does by tracking my hourly based on the last 100, 500 and 1000 hours. Unfortunately, I haven't been able to find a way to do that on Excel. I am not even sure that this is the stat I should actually be attempting this with.

If anyone has any suggestions on how to do it, I will be happy to hear them.

Good job, DK.

As a plo reg I've got to say NLHE had way more room for exploit and I'd therefore way more fun in a good game. NLHE is just tougher that's all, at high stakes people are better. PLO don't necessarily offer but get edges inherently. Fish who move to PLO usually lose less or slower there (win sometimes to balance out).

Why not look at profit in BB/hr vs session length? It'll indicate if you're actually playing worse or better in longer sessions without masking the effect of simply having more hours at the table.


I make my trailing winrate figure in Matlab. All the sessions are ordered from earliest to latest in my log, so I first calculate a running total of cumulative hours played. Then I walk through the list of session and to a search for any entries with lower total cumulative hours to that entry, and within 100 or 500 hours of it. Once I know that list of sessions I just sum up the hours and the winnings for them. So the sample isn't actually exactly 100 or 500 hours, but it'll be within a session of it and should be only +- 2-4 hours in most cases.


In Excel that's a little more annoying IMO.

You can easily calculate cumulative hours. Then you need a second column that's just "cumulative - trailing distance" (100 or 500 or whatever).

Now you need to use the "sumif" function like this:

=SUMIF($L$2:L3,">"&Q3,$F$2:F3)

where column L contains my cumulative hours, column F contains my session winnings, and column Q has my calculated "cumulative-limit" value.

That calculates your total win/loss for the trailing limit hours. Then do the same thing for the trailing hours themselves:

=SUMIF($L$2:L3,">"&Q3,$G$2:G3)

where column G is the number of hours in each session.

You'll see the trailing hours float around a little bit as the last session included changes length, but most of mine were < 105 hours for each entry.

Then you can plot the trailing winrate vs session number or total hours played or whatever.



I hope that helps. I'm really interested in seeing such a figure from another player.

Sure, I can do that.
Ok, I need to read this carefully to get the hang of it, but i ll give it a try.

Ok, some questions on some things I need to understand better/i m having problems with.
You don't mean manually, do you? because that would be a pain.
Ok, I did a column with cumulative hours by creating the formula (=E2+E3), where basically, each cell is the sum of the current cell and the previous cell.

but then I tried to do a new column with the formula (=F20-100) to figure out the trailing distance, but I get either gibberish or numbers that don't add up.


I think I understand the first two, but the last one is the one I am trying to figure above and I can't correct?

I used to be a ftp reg at 5/10 plo for about 3 years. Not that this is all wrong (sol nails the first sentence), but I agree with very little of it. Name dropping is *the worst* but at a WPT event I was talking with Anthony Zimo who has a 25k plo bracelet while playing some 5/10/20 plo and he reiterated how much better holdem games are to exploit fish.

JB really gives very, very poor advice in every thread I've seen. It's a bad look.

^ and yet I was called "rude" for suggesting he drop down in stakes after reading his travesty of HH's

I've grown up enough to be happy for anyone who wins, or claims to win at poker. When I was in my teens/early 20s I was definitely more jealous of players who had more success, but rather than being an ******* I actually improved my game to win at medium/high stakes. Maybe he wins at his stakes. Good for him. But, this forum has been infected with quite a bit of poor advice in the past few weeks.

No. Way too much of a pain. Just a for loop. (Couldn't see a quick way to do it with matrix math :shrug

This looks wrong and circular.

Using the column notation that I had, session hours is in G, cumulative hours is in L.

So L1 is : =G1

L2 is : =L1+G2

Then you can drag L2 down. and get stuff that looks like:

L3 is: =L2+G3


This one should be super easy once you have column L working right.

In column Q you just need:

Q1 is: =L1 - 100 (then drag/fill the rest of the column)

or -500, or whatever "trailing" period you're interested in.

Every entry in Q is 100 lower than the corresponding entry in L.the actual calculation of the trailing distance itself (which should always be close to your target 100) is the result of one of the sumif calls.

I think your problem is the cumulative hours column. It should be monotonically increasing.

No, it's not working. I did it your way for cumulative hours and I get the same results as before - which did make sense. Maybe it's the cell format?

I am using [h]:mm.

PS. Now that I notice, if you deduct 1 from L1, it's actually deducting 24 hrs. However, the format on the cell is also [h]:mm, so I don't get it.

Oh ... I just have "hours"

So a half hour is 0.5, maybe make that conversion in a new column.

I wont disagree with sol, tbh i dont fully understand the exploitability argument.

I assume its along the lines of equities run farther apart in nl so its possible to play educated with a higher edge. Also the ability to have unlimited bet sizing which adds an interesting dynamic to the game.

While I'd agree, i believe live low stakes plo sees such immense errors that the edge still favors the intelligent player relatively well. I saw bare bottom 2 pair get all in in a $4k pot last month. In 1/2. So this isnt exactly a game of "thinner edges".

I would say that live plo is an easier game to play overall. I tried to convince my gf to learn it over nl. At the lowest stakes, it is mostly a binary decision tree.

Do you have the nuts?

If yes -> pot
If no -> fold

Id disagree that the fish lose less in plo. I understand they can have bigger wins, and that closer equities means smoother win/loss amongst players, but a poor player's losses can be substantial and can trigger a true quit (which is the biggest reason why plo is actually a pretty poor game format for longevity imo)

Probably about 80% of the nl players i play with are the same regulars from 5 years ago. Plo is a revolving door.

I figured it out. You need to subtract 100/24 because for some reason Excel substracts time as a fraction of a day.

Regarding PLO, when the money gets in *early* in a hand, like when fish are re-potting each other preflop, the equities are generally close and they're not really making *huge* mistakes getting it in. So they win and lose a lot of pots in nearly coinflip situations and selectively remember the wins more than the losses.

When the money starts getting in later in a hand, like on the turn, that's when we can have massive edges on them with dominated and counterfeited draws or freerolls. But that gets us back into the deep stack situation where you need to be very well rolled to even get that far in a hand. Bottom two pair for $4k ... or 20BI for NLHE.

I think you need a little more than that binary tree in a small player pool because they *will* adjust to you only putting money in with the nuts. But that additional adjustment can be as simple as flatting pre and checking the flop to let them bomb it for you.

Many non-whale fish can't handle the swings of PLO and stop playing. I've seen it almost kill an entire room because half the players went broke and couldn't play at all until the first of the month.