You cant compare 2/4 to 2/4 NL. A fairer comparison is if you equate bankroll/buyins.

2/4 Limit: $2000 roll, $100 buyins, 2-3BB/100 = $8-12/100
.5/1 NL: $2000 roll, $100 buyins, 4-6PTBB/100 = $8-12/100

I think I now understand what the other side is after here. Growlers, et. al. correct me if this is wrong. The question that you want answered is: "I want a win rate of $X/hr or $X/100 hands. I can get that rate in limit game A and/or NL game B. Which of these two games will result in lower variance in terms of $/hr (or $/100)?"

Assuming that is the question, then the answer would be to determine what the SD is in $/100 for you in game A and game B. Redhouse gave us a starting point (2/4 limit and .5/1 NL), so let's say we have SD numbers in BB/100 of 15BB/100 for the limit game, and 40ptbb/100 for the NL. In $ amounts that would be $60/100 for the limit and $80/100 for the NL. NL would still be more variance, but the gap looks much smaller, expressed in $/100.

I would think that the only players who have much of a shot at having more variance in limit are the TAG players who are good at NL, but kind of bad at limit. Maybe something like a 0.5BB/100 in the limit game and 5ptbb/100 in the NL with SD #s of 14BB/100 in limit and 30ptbb/100 for the NL. So....let's assume .5/1 NL for a $10/100 win rate...which means we need the limit game to be 10/20 (right?). Now in $ terms the SD is $280/100 in the limit and $60 in the NL, lol, yeah that's quite a difference. I guess we don't need such a disparity.

Growlers, I don't think we are disagreeing. I was treating risk of ruin as a variable for calculating bankroll given a particular win rate and SD. In that case you just pick what level of risk you are comfortable with, e.g. 1%, and then calculate for bankroll. In that case, SD does not affect risk of ruin. If however, you want to determine your risk of ruin given win rate, SD, and bankroll, then yes, SD affects the solution for risk of ruin. Similarly, you could solve for SD given the other variables. I hope I've put that more clearly than last time.

The other thing to consider is if that post by pokey in the 2+2 link is right, your SD number in PT will be skewed by the length of your sessions. If I understand his point correctly, the SD/100 in PT is the SD per session normalized to a 100 hands. So as he said, if you play 100,000 hands in a single session, PT will report an SD of zero. I just checked this in PT, and it appears to be right. (PT actually says not enough sessions to calculate).

Given this, say a guy plays only half hour sessions. Playing limit, his per session fluctuation will usually be pretty small since he cant lose a massive amount per hand and he isnt playing many hands. In NL however, it wouldnt be uncommon to be up or down a stack even playing short sessions, so it would be higher.

So basically the SD number in PT isnt a real SD i guess is my point.

And for what its worth, here are my thoughts in cliff notes form. I may do some real analysis if my NL game improves to the point where i seriously think about choosing it over limit as my main game.

Intuitively, the sources of variance in the two come from different sources. In limit you play more hands and play them further since (a) you get the right price more often and (b) your risk is limited by betting structure.

In NL, although you play fewer hands to the river, the $ variance for these few hands is much much higher. So the overall variance is going to be dominated by the manner in which you play these pots and the nature of your opposition. If you're a setminer for example, it'll be low since you arent shoving that often and you're shoving with good hands. If you like to shove draws etc though, it'll obviously be much higher. Similarly, against certain table compositions there will be more big pots than others.

So imo, variance in NL is much more sensitive to playing style, even if you measure it by PT's SD stat. In limit for example, most TAGs report an SD of about 15 BB/100. I play LAG and mine is somewhere in the 18-20 range, whcih is 15-20% higher. In NL on the other hand, most TAGs report about 45 ptbb/100. LAGs report a much higher number - close to 60-70 ptbb/100 which is about 35-50% higher.

I have no idea on what the SD in PT really represents. However it appears that some of the LHE and NL SD/100 numbers appear to be fairly close. We had a discussion about this on PFO and most of the LHE players in FR had SD/100 bunbers in the mid-teens. Between there and here we can't have everyone playing roughly the same amount of sessions in terms of lenght could we?

Where the hell is Marm when you need him? He knows a lot about the PT program, plus he's a pretty good with numbers. FWIW IIRC his SD/100 was closer to the 16/17 mark (altough I could be mistaken), but he dosen't tend to be ultra TAG either.