so here are a couple things that I have read about all in ev:

-unless you are playing HU, it doesnt accurately reflect your real EV in ring games
-it only applies to all in situations, and your true EV in every aspect of the game is a different story

so i am wondering:

-why do many ring game players put such a large precedent on AIEV?
-does this stat have anything to offer non-HU players?

AIEV is pretty accurate. Over small samples it predicts your "true" win rate more accurately than the Net won stat, even in ring games.

None of those two stats are all that important though, if you ask me. The one stat that really matters at the end of the day is...

Pots where you get all in are quite rare (in cashgames, at least), but they have a significant effect on your overall winrate. It's therefore useful to use the EV line to estimate your theoretical winrate, since it "ignores" the real life variance of those significant pots.
For example, KK and AA alone make more money than all other hands combined, so if you run badly with KK+ over a small sample, your net results will be atrocious. The EV graph can restore your sanity, as it will often 'prove' that you'd be winning if only you ran better. On the flipside, if you're actually running above EV in all in pots, and that's leading to a monster winrate, the graph can stop you from getting too carried away, as it serves as a reminder of your luck.

It's just as useful for ring games as it is for heads up.

^ The problem with using all-in EV for ring games are card removal effects from folded opponents. And no, depending on your play style this isn't always negligible.

When you see those graphs where people are running over/under all-in EV over several hundred k hands and the gap between the lines is only increasing over time, that's a strong hint that the all-in EV calculation is not consistent with the actual game EVs.

It could also imply collusion.

If there is a non-zero money advantage to be had from using card removal effects (and I think there is), your AIEV line will necessarily be drifting away from your data-EV line (as opposed to theoretical actual EV).

It does seem like a relatively negligible effect if you aren't colluding though.

Example for people not following: Solid player shoves allin in PLO, two loose players fold, you have a marginal flush draw. If you call, your AIEV graph will consider the loose players hands as just eight random cards, but since you know the players are loose, they likely would have called if they had flush draws themselves. So they're mostly folding other suits, meaning the deck is more rich for your flush draw outs, which might tip your decision to a call instead of a fold. As long as good players employ this type of thinking sometimes, their data-winrate should be above their AIEV-winrate over large samples. My guess though is that it would take several millions of hands before the accumulating small advantage of card removal decisions becomes more significant than just straight variance.

About the card removal effect, its negligible enough that it doesnt affect things. unless collusion.
About OP initial statements, they are false. Where did they come from?

That's the conventional wisdom, sure. If you actually look into card removal effects then they don't seem negligible though.

e.g. The all-in EV of small pairs vs highcards changes considerably for blind-vs-blind battles when folded to on a full table. This is a fairly frequent situation and if your playstyle puts you on one end of that matchup more often than the other, then this may quite possibly cause a significant drift in the all-in EV line.

1) how much does using the ev curve instead of the net winning reduce our std? There's another thread right on this a forum about this, I made made a very rough estimate that it reduces it by 10%, so about 6 for a FR player.

2) let's say on average 4 players have acted before our all in with another player. We have 66 and he folds enough for us to want to shove, but calls AA. I don't want to complicate this so let's say we only win if we get a 6. Let's say previous players only fold if both their cards are lower than J. This reduces the 6 in the deck from 2 to 1.5 approx. And cards in the deck from 46 to 38. If our chance for a 6 was 20% now it's 20% x 1.5 / 36 × 48 / 2 = 18%
So we lose 4bb on our shove.
This means that for the error induced by blockers be bigger than the std reduction of the all in adj ev, we need to find ourselves in a situation where we can shove and blockers are enough to make us not to or viceversa, once every 67 hands. I doubt I even shove 1 in 67 hands.
And this is a best scenario, a hand like AJs cares much less about blockers. And of course, we are not the only ones that understand
blockers effects, some of out opponents should too. And also, the times that this makes us fold cancel out with the times that this makes us call, altough those are rarer I think.
If we ignore the unpaired hands and only calculate what happens with our pairs we need to have a borderline profitable shove every 67 hands with 22-99 or whatever. That is, the shove needs to be 0-4 for us to then not do it because of blockers
Is there anything wrong with all I've said?

I think 2) is missing the point a bit, as the blocker error being smaller than the std reduction is not sufficient at all.

If you reduce std by 10% while introducing a 0.5% bias, that's still a pretty terrible result. Players expect results and a/i EV lines to converge in the long run, if bias is introduced this will no longer be the case and that can be extremely misleading.

A player may be running below all-in EV and conclude he's playing fine and just running bad, when in reality the difference might be completely due to a biased all-in EV calculation. If that's the case he'll just continue "running bad" for eternity unless he changes his game.

Let's look at an example:

Regular SD: +/- 100$ per 100 hands
AIEV SD: +/- 90$ per 100 hands, with a +0.5$ per 100 hands bias

A 0.5$ bias to get a 10$ reduction in SD might look like a reasonable tradeoff on first sight, but let's see how this works out once you play more.

On a 10k sample:
Regular SD: +/- 1000$
AIEV SD: +/- 900$, +50$ bias (50$ bias for 100$ SD reduction)

On a 1mio sample:
Regular SD: +/- 10,000$
AIEV SD: +/- 9,000$, +5,000$ bias (5k bias for 1k SD reduction)

etc.

You are right, my reasoning was wrong.

So if we are winning 0.5bb/100 (which I find a high estimate tough) due to blockers our aiev becomes less accurate than net winnings after 40k hands.

For 6 max, it takes about 160k hands, not counting the fact that as the effect gets smaller we also have a harder time taking advantage of it

I find this discussion about removal to be very interesting and i hope it continues.

Is it possible that the effect of considering removal would be more pronounced in PLO where more cards are removed, and in theory it can be easier to hone in on the kind of holdings your opponents fold '

BTW lezalias here are the articles these first questions came from:
https://www.pokertracker.com/blog/20...-all-in-equity
https://www.pokernews.com/strategy/o...roll-23141.htm

First article is from 2011 talking about PokerTracker3 software and potential bugs with hand histories, not card removal effects. No idea if the bugs still exist for modern software but I doubt it.

Second article seems like more of a critique of being results oriented in general even though the author is specifically blaming AIEV.

it certainly makes sense that card removal would have greater impact in Omaha, especially in ring games. I have heard this discussion about removal influencing AIEV but never seen any consensus on the subject. If the edge from collusion can express itself this way then I cant imagine that a skilled ring-plo player wouldn't be able to get an edge from considering removal. In NL i feel like the effect would be more negligible

Has anybody seen any "research" into the effects of card removal in PLO? Seems like it could be an interesting topic.