I have recently deposited and started playing online again. Previously I have played online and the option to cash out / run it twice or thrice didn't exist. You put the money in and run it once and take whatever the variance gods have planned for you. Exercising good bankroll management and not playing above your limits you hope for the variance to even out and run according to expectation.

My main game is PLO and as you may expect there are a lot of spots where it's 60/40, 65/35, 70/30 and even 85/15. After being at the end of some unfortunate results it has got me thinking whether I should be changing my tactics. Currently I do not opt for the EV cash out (I believe the site takes a 1% fee for this) and I have been mostly running it once. I can think of 1 or 2 times occasions where it was almost 50/50 for a big pot and I accepted to run it twice.

Running it twice

I understand the theory that the more times we run the deck, the closer to our equity we realise. The next bit is probably fish logic i'm sure, but if I accept to run it twice - I give my opponent additional cards to hit and risk chopping the pot with them. Conversely, if we run it once and they hit - I end up with 0% of the pot.

Running it three times doesn't seem much more optimal than twice, other than in a spot where my opponent has one out and hits on a run. I will at most lose 33% of the pot vs 50% of the pot when running it twice.

The pros generally seem to run it twice so it must be an optimal strategy. Although I imagine it somewhat reduces the financial swings and variance too.

Question
1. Should I always be running it twice / thrice?
If no / maybe to the above:
2. Are there scenarios where I should run it once, and scenarios where I run it more times? Such as 60/40 spots multiway, 80/20 heads up, when i'm behind and want to give myself more opportunities to hit.

EV cash out

The other option the site gives is the opportunity to do an EV cash out. The money goes in and i'm 70/30 - I can cash out 70% of the pot and the site takes 1%.

It would appear a great way to be realising equity, but means that you're giving away a % of your winnings + effectively paying more rake. Conversely, it is better than taking 0% of the pot.

Personally, I think if you play within your bankroll and aren't going to go on monkey tilt then it's probably better to not pay for an EV cash out.

Question
Is the EV cash out really a sucker's bet?

Running it twice makes no difference to your long term expectation, but reduces variance. This allows you to either play slightly higher or have a lower bankroll. If you are properly rolled it means you are likely to stay closer to your ev expectation, so it reduces swings.

The cash out lowers your win rate, and reduces variance. Whether that is worth the tradeoff will depend on your risk of ruin with or without it, your loss aversion level, and mental factors. If you tend to tilt easily it may well be +EV.

In a perfect world, yes, it is a sucker bet. How perfect is your play and mental game?

On that basis it sounds like I should be opting to run it twice when offered.

Although when in a spot where i'm the dog with 30% equity, I should welcome the variance and run it once? If running it multiple times gets me closer to my ev expectation, I have more to gain running it once, than I lose. ie. I can take 100% of the pot if variance is on my side or lose the 30%. Alternatively, the more times I run it, the closer I get to realising the 30% (which would be a loss).

The cash out feature has some appeal because it would be nice to walk away from sessions winning rather than losing / breaking even. At the stakes I play it doesn't annoy me enough to put me on tilt. Although after a succession of beats I feel some frustration. The downside of using the cash out is when my hand holds and watching 30% of my profits not be awarded. It's a "damned if you do and damned if you don't" scenario.

This is the wrong way to think about it. Running it once is an ev-neutral proposition that is simply a variance reduction. Whether you run it once or run it twice (or however many times), your ev in the spot is identical. Yes, if you're a dog you will have the chance to win the whole pot, but most of the time you won't. On the other hand, if you run it twice, you will quite rarely win both runs, but you have a pretty decent chance of chopping the pot. Getting half the pot back is something that basically never happens when you get it in as a dog.

Either way, you have neither gained nor lost money by electing to run it twice, you've simply reduced the amount of variance in the game. Running it twice is basically trading the variance from winning/losing the entire pot into more situations where you chop, making the game less swingy. Of course, you will still scoop or get scooped sometimes.

You also mention that you can feel some frustration after a succession of beats. It's possible that RIT can help here (because a lot of "beats" will turn into chops instead of losing the whole pot), but also possible that it can hurt (if you RIT and lose the second run, you might feel like you "would have won" the whole pot if you just ran it once). This is a personal psychology thing, and you'll have to consider what works for you.

Cashing out is strictly worse than running it twice because you are paying an extra fee to the pokersite, and while the fee seems small it's only happening in your biggest pots so it adds up decently over time. There might be some rare situations when you want to use it (for example, if you're shottaking or if you're super deep with a whale and don't want to risk losing your stack), but generally speaking you should stay away from the cashouts. Online is a competitive environment with narrow winrates as is, you don't want to give any more $ to the sites than you have to. Incidentally, a couple sites out there actually double rake RIT hands, and on those sites you should likely avoid running it twice as well. But on most sites RIT does not cost additional rake, making it a decent option.

I played a session now and opted to run it twice when given the opportunity. Only once did I run it once when I was pretty much drawing dead to 1 out. It may have saved me on one occasion and on another I chopped the first run and won the second run - which would have been a better outcome than running it once.

I did not opt to cash out in some spots where I was 70%/85% and once again took a beat. Maybe it's a case of happening fairly regularly in a small sample but for now I will avoid paying the extra fee and not using this feature. Maybe playing higher / shot taking it is worth using.

I believe GG used to double rake for RIT but stopped doing that.

while running it twice obv does not change ev, it does have some important metagame implications - while in theory reducing variance shouldn't really be important to a winning player, it will affect weaker players and their behaviour at the tables. For example, you always want to run it twice vs a hit-and-runner, but run it once vs shortstacks or title-prone players.

All that being said, trying to assign some "strategic" importance to how many times you run it is a sign of a very serious leak, it means that you don't understand some very basic terms like EV

as for cashing out, afaik apart from a brief promo at GG, it is always raked. Don't opt-in to pay extra rake, period