On your example, if I needed the money out after one year, the fact that if I can wait out 5 years I'm guaranteed a 7% return isn't completely comforting. To get back to poker, it's different if you're only rolled to lose 5 or 10 of those 2:1 shots. You may never see the long run.

Not saying I do or that anyone should play this way. It's just how I see avoiding variance in practice.

It's an entirely different scenario.

If you are putting money in a 5-year term investment and more than likely have to withdraw in shorter term, then clearly there are additional variables to consider.

I don't know what we're disagreeing about. I think we both agree that your stock example and my poker example are on average positive returns. I think we both agree that there's variance around both examples, such that while they may generate a positive return on average, there's no guarantee that over the short run they'll generate a positive return, i.e. you could lose all your money in a month in the stock market, and you could miss your draw in poker. I'm also saying that someone could conceivably decide that this variance is not worth the return, and prefer something with lower variance and lower returns (say investing in treasuries as opposed to stock or folding as opposed to calling with a slightly +EV draw).

Would anyone mind humoring me here to help settle a debate? Let's say you
Have a 10k life roll and expenses were 1k per month living in Vegas. You
Are a long term winner at 1/2 for 8bb/hr. At what points would you
Jump from 1/2-1/3 and 1/3-2/5? Thanks

In that scenario I would jump to 1/3 now. Its not that big of a skill difference and there's not a massive difference in bankroll needed. The jump to 2/5 is much much bigger skill difference in my experience so I would want to either take occasional shots or have a much bigger roll first.

Just pointing out that it isn't avoiding variance. You cannot avoid variance...

Sure it's a minute detail, but that's kind of the point that everyone keeps missing in these discussions.

In your response of short-term effect, problem with adding such caveat is that when variables change, so does EV calculation. What may had been +EV is now -EV. In other words, you are not avoiding variance, you are avoiding -EV spots.

Thanks for response. I know skill difference is minimal just was factoring that 1/3 does play bigger than 1/2. But based on your response at what point would you start shot taking 2/5?

What do you mean by "life roll"? Do you have an actual poker bank roll that is separated from other funds and wont be touched no matter what happens in your personal life? How much of that 10K is actual dedicated poker money?

Also, do you know what your standard deviation is?

By life roll saying that all $ is poker money and life expenses come out of that same 10k so no they are not separate. Sorry for being vague don't have deviation numbers atm was just looking for a broad guideline on when to start playing different limits based on the wine rate and monthly expenses given, again appreciate your responses

Really, what you need to do is crank out some hours to get a bigger roll because your roll is tiny. If you consider $6k is what you need for 6 months of living expenses then you only have a $4k poker roll which isn't sufficient for 1/3 or some of the bigger 1/2 games in Vegas.

Sure, you could take a chance and just allocate $3k to liferoll (3 months of expenses) which gives you a $7k bankroll which would allow you to take shots at 1/3, but it sounds like you haven't played much 1/3 in the past, so why not just put in tons of hours at 1/2 and run your bankroll up a bit so you can take more shots at 1/3?

I agree that's the way to go, I prob didn't word original post that well. So let's say we put three k aside for three months expenses, with the 7k remaining for 1/2 when would you suggest trying 1/3 and how many buyins?

Actually if you fold here you are giving up $210 worth of ev. Not sure how you got $7. Folding here would be lol bad. And that is assuming he was all in ott, if he has more money behind for implied odds... even better.

But even if it was somehow only +$7 in ev, you still should not fold. It is like taking $7 and burning it in a fire. Poker is inherently a game of variance, embrace the small edges, that's how vegas was built.

If you separate $7000 just for poker, you have 23 buy ins for 1/3. If you are really an 8BB winner at 1/2, that should be plenty. It would be nice to see your standard deviation though because that helps with figuring how consistent your results are. If you have lots of big wins and big losses, you are more likely to need a bigger roll in case several of those big losses come back to back to back. If you have smaller wins and losses and win a higher percentage of sessions, you will tend to need a smaller roll IMO.

If you have a long track record (500-1000+ hrs) and your biggest downswing is 4-5 buy ins, then you can have move confidence that 23 buys ins in plenty. Some guys have the same 8BB win rate but have regular 10+ buy in swings. They would need a bigger roll.

1500 hour sample for 1/2. Appreciate all of your help thought I would be fine with 20 buy ins for 1/3 just wanted to get some feedback

That is not necessarily true. Variance is a tangible cost. That is why equities prices on the stock market are heavily influenced by volatility. Or you can look at the insurance industry where people pay to reduce variance.


What value would you put $7 eV if right now we coin flip: you get either -$100,000 or +$100,007 at 50/50 chance? Your answer will probably change if the payout changes to $-5 or +$12.



But I do agree that $7 is a lot to pat for variance. If we fold in similar spots once per hour, we will lose out 2.33 blinds/hr at 1/3.

I wouldn't keep just 3k as a liferoll though. Do you have a car? What happens if something major happens to it? Can you go without it? What happens if you have a major medical emergency that prevents you from playing for a period of time? Can you cover the expenses and survive without playing for a bit? What happens if your laptop breaks or your phone gets stolen, or you get jacked for the money on your person.

Of course there are many other potential pit falls as well (pit games, strippers, girlfriends, loaning money to friends, home burgaries, ballin' out of control, etc).

The EV of either one is $3.50.

Def wouldn't keep three k life roll and be comfortable with it, but for a short term thing till I build my roll I think it's ok. Will for sure add to it as time goes on my main concern was making sure with my win rate that 20-25 buyins would not be sufficient.( obv not ideal I know more would be better)

Whops, nice catch. Meant what would you pay for the option to gamble in each scenario. You'll probably pay more for the latter even though both eV are the same. That is the price of decreased variance.

Not all EV is created equal. You can compare this to the Sharpe Ratio in finance which measures risk-adjusted return.

Sharpe ratio = (Mean portfolio return − Risk-free rate)/Standard deviation of portfolio return

Which can be adjusted to poker as simply: EV/St.Dev

Since most people are under rolled for whatever games they are playing in, I think it is worthwhile to chase all +EV situations that are negligible to your BR while passing up marginal EV spots that will have a detrimental effect on your BR (like already being stuck for the night while 300-400 BB's deep and finding yourself in a very marginal but +EV stack commitment situation).

My stretch of runbad really forced me to reconsider the notion of risk/reward at poker and decide that the long-term approach is best for achieving my goals. Sure it would be nice to be playing 10/10 again right now and I could certainly shot take on weekend nights, but the downside risk is significantly greater than the potential rewards right now, aka a low Sharpe Ratio for me personally.

There's a really good post by Aaron Brown (it pains me to say that because I worked with him and found him just insufferable) somewhere on 2p2 where he responds to people trying to equate poker to Sharpe ratios. The gist is that unlike investing where you are trying to harvest return in a noisy market, variance is actually a weapon in poker. If anyone could find it they should post it ITT.

If you can make someone fold a +EV situation by forcing variance upon them then you are doing it right.

Is there anywhere I can look up downswing length and dollar amounts for different levels?

Different levels? It depends on way more than that.

Different stakes I was referring to not skill levels not sure if I was clear, just wanted to know if there was a thread on it already that I may have missed

It really isn't the stakes that make the DS length, nor the BB amount, it is player edge (or lack thereof) and standard deviation. Stakes play a roll in $ amount, but in ways only knowable if we know approx expected winrate and standard deviation.