I'm interested in your opinions about the lowest stakes online vs live, which one do you guys think is harder to beat? I've never played live, so I'd very much like to hear what people who actually play live and online think. Thanks!

Online is harder. 2NL rake will rape you in the ass and the average skill level between the two games is probably similar, although it will depend a lot on which site/casino you're comparing.

Skill level is similar but the play is very different.

2NL you'll find preflop is generally more conventional, many hands don't go to the flop or are HU. But you will also get people doing crazy stuff you rarely see live, such as open shoving over limps or donk shoving every other flop.

Live is much weaker preflop, larger raise sizes and more callers, less 3 betting, more open limping. Paradoxically this makes postflop play significantly more difficult to navigate, since you end up in multiway pots much more often. There are many meh regs who have mastered some basic postflops heuristics, meaning while they are bad at poker, they are not going to call a x10 pot flop shove without a good hand and they will notice if you open 3bb for most of your range but 6bb for Kings and Aces.

Lastly there are a bunch of metaskills required for live that don't apply to 2NL. First, you are playing with real money, almost anyone can afford to get raped at 2NL, most people can't afford to blow off multiple buyins at 200NL on the daily. Second you need to handle the much slower pace of play without getting bored and frustrated. Third you need to be able to deal with playing vs. people you can see instead of sitting on the couch next to your piss bottle. Fourth there are hazards for degenerates in a casino, such as pit games and waitresses who will bring you alcohol.

Not sure what you mean by the online rake being too high. Not that I don't agree that it's high, to me the live rake seem even higher. I've heard of the live rake being about 10% with a $5 cap. On ACR for example the rake is 5% and you get roughly 15% rake back in combat points that you redeem for real money, so the effective rake is 4.25%. Not to mention that you pretty much have to tip the live dealer, on top of the rake and the money they take for high hands and stuff like that.

Promo drop should be about EV neutral, depending on the promo.

You make a good point. 10% is a lot, but if we say you always tip the dealer $1 live 1|2 rake is capped at 3BB. 2NL is practically uncapped. When you GII for a buyin that's 200*.05 = 10BB rake pre-rakeback. I guess it depends on how big the average pots are. If pots are small on average then 5% uncapped is better. If pots are big on average 10% capped at 3BB is better. I'd have to check my database but I think in really soft games it's not rare to pay 3BB+ in rake at micros. Soft games usually have bigger pots on average.

I see what you mean. I agree with the relative neutrality of the promo drop for the players who play on a constant basis, and partially with the difficulty of beating the rake. On ACR an all in between two $2 stacks is raked 17c, or 8.5bb. Depending on how may large pots like that one plays it is theoretically possible to exceed the live rake.
Regardless, here is my two cents about the online rake: while it is a tremendous money maker for the house because it continuously removes money from the tables, it should not be the main concern for any individual player. I am saying that because a 4.25% rake means you keep 95.75% of your winnings. Yes, 100% would be better and in that case one could have the expectation of breaking even while playing against opponents of equal skill. With the rake, the player has essentially a 4.25% handicap, so he needs to play 4.25% better than the average player in order to achieve the same result. That is not much mathematically speaking so what else is out there that makes so many players complain about losing money at micro stakes? IMHO the answer is the overestimation of our own poker skill level, with rake being a distant second.

I guess this depends where you play live.

There are different dynamics, fullring table vs shorthanded etc.

I only played live in spain and portugal and I'd say the skill level is comparable to NL25 regular tables online, zoom plays much tighter-agressive, if there is any rec on the table on the frespt market (there always was)

NL2c is much softer I think though I haven't played as low quite a long time, but from what I've seen people are always giving stacks away which they are more careful when playing live.

It might not seem like much but it's a very significant portion of your win-rate. Over about 22k hands I paid 18BB/100 in rake at 10NL.

I pay 10% rake capped at $15 for live poker in Australia.

Sorry, my level of understanding the poker lingo in quite basic, what do you mean by 18BB/100, what does 100 mean?

18 big blinds per 100 hands. A lot of players consider 10BB/100 "good" after rake. Someone with a 10BB/100 winrate at 10NL has more than half of their winnings taken by the rake.

100 hands

which is kind of meaningless as there's no real definition of how much rake "you paid"

What definition could there be other than the money taken out of pots you won?

I raise to $5 pre, you call, pot is $10, site takes its usual 5%. Who's paid the 50 cents at this stage? I certainly used to get rakeback on this back in the FTP days, how did I get that if I didn't pay rake?

That is a good point. Rakeback is given for the money you contributed to the pot, so in your example each player gets rakeback for 25 cents. This way if you win significantly more than you lose you get less than 15% rakeback overall, and you get full 15% only if you are a losing, maximum break even player. Actually, even a break even player gets a little less than 15% since he is technically a winning player due to the rake.

Thank you for the clarification, now I understand. It's true that lowering the rake would reduce the number of BBs paid over 100 hands. But I think the ratio was designed more as a measure of your own play style, rather than the rake itself. We all know that we need to select good starting hands, bet mostly when we are clearly ahead, play speculative hands when in position and it's cheap to do so, not engage in preflop wars and races. The ratio measures just that. For example, if I play 100 hands, fold 99 of them and in the one that I play I win a dollar from my opponent, I pay 10 cents in rake (5% of the two dollars in the pot, my opponent's and mine) and keep 90% or 90 cents as a profit. That at my stakes equals to 5bb/100 hands in rake. Now let's say I get involved in 25 hands out of 100, win 13 of them for $26 and lose 12 hands for $25. I won still $1 more than I lost, but I paid $1.3 in rake, so I actually lost 30 cents. The ratio would be 65bb/100.

There's no reason to consider the rake taken until the end of the hand, so nobody has paid the rake yet. The 50 cents is still part of the pot. If it wasn't still part of the pot it wouldn't be part of future rake calculations.

The easiest answer to the rakeback question is that the rakeback formula doesn't go by rake paid. Sites can calculate rakeback however they want.

Rake taken from a pot I've won is money straight out of my pocket. I couldn't care less about rake coming out of a pot I didn't win $0 = $0 no matter how much rake they take. 18BB/100 in rake paid by this definition is 18BB/100 straight off my winrate.