Quote:
Originally Posted by Kelvis
Isn't that what a bankroll is for? What do you mean by "low variance approaches scale better"? And by nature of being less profitable, they are less sustainable. Also who says we're going to start bluffing off stacks? If you want to make any decent money in live games you absolutely cannot afford to quit a good table being deepstacked against the fish. It's lunacy.
Sorry if I've been a bit vague, I try to explain better. This is going to be a bit of a wall of text
1. Bankroll:
Yes you are 100% correct, that's why you have to have a good BRM cause your bankroll is supposed to deal with the variance.
Unfortunately, variance is not fixed but conditional.
So lets say your average standart deviation for FR live games is like 80$/100hands while your average winrate is 5$/100hands. You accept a risk of ruin of 5%, so you need around 2000$ as bankroll.
Now you get into a wild game, you vs 8 whales. Lets adjust for that by increasing standart deviation to 120$/100hands. Of course you win more, so we adust in a linear fashion (both up 40%) to 7.5$/100hands.
To have the same risk of ruin, your required bankroll for this game just went up to 2700$
I always wonder, why people scream and shout when a noob uses the same BRM for NLHE and PLO, but find it perfectly natural to play with the same BRM in nitty and wild games.
As a matter of fact, you cannot play every NLHE game with your 20BI rule. When variance goes through the roof, you have to either play way under your bankroll, play very nitty or stay out of the game or your risk of ruin increases dramatically.
2. Bluffing off stacks.
Look through the beginner section of the forum. The majority of post whining about not being able to beat donks are written while on tilt. We are not robots and every player has a point at which he just cracks. Get your balls kicked a couple of times in a row will have an effect on your game.
In online low limit games it's not uncommon to get sucked out 1-2 times per hour, which doesn't matter since you play so many hands that you can grind it back up within the same session.
But imagine you sit at that whale table, they are splashing money left and right, but you are card dead for one hour. Then you get 77 on a J73r flop, he bets, you shove, he calls, shows JT and the runout is TT. Then half an hour later your KK gets cracked by Q9o and then you lose with 99 cause counterfit. All that happens while you sit there and "know" you are the better player, they have fun and laugh their asses off.
At a certain point you will tilt and play worse, even the pro's do. There are more than enough youtube videos from high stakes games with Ivey, Antonius etc. tilting off their stacks.
3."I'm the better player":
This is always the most stunning point of rant posts. "they play so bad, while I play so well but I'm so unlucky"
With a winrate of 5$/100hands and a standart deviation of 100$/100hands your 95% confidence interval is -1.3 to 11.3 over a sample size of 200k hands. Estimating 30hands/hour and average session lengh of 5h, you don't know in 1300 live sessions, if you are a winning player or not with 95% confidence. Thats almost four years IF you play every single day.
The key fact here is: The higher the variance, the wider the confidence interval meaning you can be less certain if you are lucky or not the higher varaince goes.
4. Low variance approach is more sustainable.
This expression was a bit vague. But without going into more complex math/ portfolio theory I try to give the following example.
Let's say you have 100K in your account and you are looking at two investments:
a.)average 20% return per year with a volatility of 10%
b.)average 80% return per year with a volatility of 40%
Which investment would you chose if you wanted to quit your job?
Which investment will make you more in the long run?
So while its probably easy to say "well I take a.) if I wanted to quit my job and b.) if I wanted to make the most money" think about risk of ruin and confidence intervals and you realise, that you just cannot be certain when it comes to b.)
If you lose half of your account in the first year and make 100% next year, you have made zero moneys. You could make 200% 4 years in a row and be a millionaire... or you can blow up within the first 8 months.
Going back to a.) it's pretty certain that you are not broke within the first year and you will stay to play for a pretty long time.
This allows you to let the compounding interest effect work for you and - which is the most important thing - you can quit or shift towards another investment because you know early when a.) doesn't work anymore without going broke.
Let's relate this to poker:
With a low variance approach you can be more certain that you are a winning player at current stakes. Nobody will play for a living with 5BI's won in 1 out of 5 sessions and losing 1 BI in 4 sessions although it's +EV.
If you can rely on your stats more, you can move up in stakes earlier to make more and you can react to changes in the metagame earlier and lose less.
Ultimately, you can put more volume into your game the lower your variance/higher certainity is, either by playing more tables (online), quitting your job and play fulltime or moving up in stakes faster.
Ergo low variance (or better low sortino ratio) scales better.
TL;DR:
Wow, what a wall of text this became...I hope it's a little clearer now.
After all, it's understandable that you want to play as much and as deep as possible against a whale because it's intuitive. Idiot plays trash, never folds and is all in every second hand. But think about the fact that people thought of JT as the best hand for a long time, before online poker was a thing. Now you come along, shove AA pre and they mark you as the idiot, although you make the correct play.
So why do you think the "whale" is bad...he could play a high variance but overall winning strategy and you just don't get it.
The question is HOW you play against this guy and HOW your BR looks like. Will you be a nit, play fit or fold every street and only shove when you have the stone cold nuts?
Or will you also play flips (i.e shove AKs pre), shove combo draws (i.e. Kd2d on 3d4dKh)?
While the later option can make you more money and fit/fold is lower variance and the better option.
Also when you play usually 2/5 on a 100K$ BR and see a wild 1/2 game, of course you can to push for the maximum, cause you're playing under bankroll and the dollar variance will probably be less than in your 2/5 games.
But it is lunacy if you play 1/2 on a 1k$ BR and suddenly a drunk whale sits down to splash. In this instance you are definitely playing over your bankroll if you don't nit up by a serious amount.
Key take away: In wild games, always chose the low variance route since variance and bank roll requirements go up and certainty about your winrate goes down...untill you are playing well below BR and don't give a fk
Last edited by Foldelinio; 10-13-2018 at 05:49 AM.