Thoughts on 'run it twice'?

[SuuPRlim]

No idea where the needle you have has come from, going to assume you're just being grumpy

They refuse to accept it is totally neutral, because it isn't.

Wrong.

Same EV less volatility is superior to same EV more volatility. 

Wrong.

In your latest example, you of course take the £10k getting 7/1 on a 5/1 chance.  Just way clear of getting 1k at 16/1.

Wrong. (2x Wrong actually)

Hope you aren't doing the maths side of your bookmaking business.

If you want an account ever just let me know 

[EvilPie]

 

[Doobs]

No idea where the needle you have has come from, going to assume you're just being grumpy

They refuse to accept it is totally neutral, because it isn't.

Wrong.

Same EV less volatility is superior to same EV more volatility. 

Wrong.

In your latest example, you of course take the £10k getting 7/1 on a 5/1 chance.  Just way clear of getting 1k at 16/1.

Wrong. (2x Wrong actually)

Hope you aren't doing the maths side of your bookmaking business.

If you want an account ever just let me know 

It is maths, not opinion. 

I can have one £1000 bet at 16/1 on a 5/1 chance.  Mean expected return is 17000/6 - 1000 = £1833.33

or I can have 100 £100 bets at 7/1 on a 5/1 chance.  Mean expected return is 100*800/6 - 10000 = £3333.33

The rest is covered here.

https://en.wikipedia.org/wiki/Modern_portfolio_theory

It isn't perfect in the financial markets, because financial markets aren't perfect, but gere we are assuming something with known returns and known probabilities, so it works.   

Here is a better example.

You have a million pounds right now and no way of ever making any more money. 

You can take one bet of a million pounds, or a million bets of £1 each on separate 50/50 chances.  Each separate bet is on a perfect 50/50 chance that is independent of all others.

You take the million pound bet on a 50/50 chance you are taking a way bigger risk than a million separate £1 bets on different 50/50 chances. 

The max loss is still a million, but one you have a 50% chance of losing your million and living in the gutter, on the other your chance of ending up in the gutter is so close to zero that excel thinks it is zero.   

This example is extreme, but the more volatility you take for the same return, just means you are taking a bigger risk for no benefit.

Of course if your plan is just to grim your gamblers if you lose, take the million pound bet. 

Happy for anyone to correct the maths above if I have made a mistake.

[SuuPRlim]

No idea where the needle you have has come from, going to assume you're just being grumpy

They refuse to accept it is totally neutral, because it isn't.

Wrong.

Same EV less volatility is superior to same EV more volatility.  

Wrong.

In your latest example, you of course take the £10k getting 7/1 on a 5/1 chance.  Just way clear of getting 1k at 16/1.

Wrong. (2x Wrong actually)

Hope you aren't doing the maths side of your bookmaking business.

If you want an account ever just let me know  

It is maths, not opinion.  

I can have one £1000 bet at 16/1 on a 5/1 chance.  Mean expected return is 17000/6 - 1000 = £1833.33

or I can have 100 £100 bets at 7/1 on a 5/1 chance.  Mean expected return is 100*800/6 - 10000 = £3333.33

The rest is covered here.

https://en.wikipedia.org/wiki/Modern_portfolio_theory

It isn't perfect in the financial markets, because financial markets aren't perfect, but gere we are assuming something with known returns and known probabilities, so it works.    

Here is a better example.

You have a million pounds right now and no way of ever making any more money.  

You can take one bet of a million pounds, or a million bets of £1 each on separate 50/50 chances.  Each separate bet is on a perfect 50/50 chance that is independent of all others.

You take the million pound bet on a 50/50 chance you are taking a way bigger risk than a million separate £1 bets on different 50/50 chances.  

The max loss is still a million, but one you have a 50% chance of losing your million and living in the gutter, on the other your chance of ending up in the gutter is so close to zero that excel thinks it is zero.    

This example is extreme, but the more volatility you take for the same return, just means you are taking a bigger risk for no benefit.

Of course if your plan is just to grim your gamblers if you lose, take the million pound bet.  

Happy for anyone to correct the maths above if I have made a mistake.

OK....

In my initial example I was suggesting a scenario where you could bet the same amount of money with drastically different EV and drastically different win %s.
You can roll 1x dice 1 time 16-1 for £1000 and make 116% EV  with a win % of just 16.6% or you can roll in 100 times for £10 a go and get 7-1 make 16% where you are a big favourite to come out ahead and an enormous price to lose all your money.

****I've just read it back and I said 100x £100 rolls which means that you'd be betting £10k not £1k which was not what I meant. I meant 100x£10 rolls so the stakes are the same but the win % is drastically different. So i'll apologise for the confusion there, and admit you were correct what you said******

My point is that you can't describe those are neutral, as it's clear there is a a big edge in picking one over the other. Just as you can't describe the other as not-neutral, when there is no edge to be had in picking one over the other.

In your examples, like the £1m coinflip or 1,000,000 £1 coin flips you're saying there is a clear advantage in one over the other, or running it twice....with no bankroll restrictions you're just wrong. There is no vacuum equity advantage in picking one over the other.  Portfolio Management is totally irrelevant to this, we're talking about isolated EV with perfect information where you know the upside, the downside and the win/lose%.  

If it's the guys last money on the table then OBVIOUSLY he should run twice because when he loses the pot he also loses the EV of all the future spots he might have if he was still in the game, spots he wont get when he's sat by the slots trying to nip the roulette guys for a taxi fare. Once you start bringing conditions in then obviously it changes, but that was never the discussion.

I'll put my money where my mouth is and bet on this.

[SuuPRlim]

Hope you aren't doing the maths side of your bookmaking business.

We each pick 3 guys with over 2000 posts on blonde that we know will be honest. You only need 2 of them to agree you're right to win.

Bet as much a you want.

[kukushkin88]

Hope you aren't doing the maths side of your bookmaking business.

We each pick 3 guys with over 2000 posts on blonde that we know will be honest. You only need 2 of them to agree you're right to win.

Bet as much a you want.

The opinion of high profile blondes wouldn't prove much though, it's just maths. You are right, following the qualification you posted.

[SuuPRlim]

Hope you aren't doing the maths side of your bookmaking business.

We each pick 3 guys with over 2000 posts on blonde that we know will be honest. You only need 2 of them to agree you're right to win.

Bet as much a you want.

The opinion of high profile blondes wouldn't prove much though, it's just maths. You are right, following the qualification you posted.

Kuk, i'll be picking you  

[Doobs]

Hope you aren't doing the maths side of your bookmaking business.

We each pick 3 guys with over 2000 posts on blonde that we know will be honest. You only need 2 of them to agree you're right to win.

Bet as much a you want.

It is maths, not opinion, as stated above.  1,000 hands with £100 pots produce less variable returns than 500 hands with £200 pots.  The higher the variance, the more likely you go bust.

If you want to believe it makes no difference, it is fine by me.   If OP believes you over me, that is fine too. 

[SuuPRlim]

It is maths, not opinion, as stated above.  1,000 hands with £100 pots produce less variable returns than 500 hands with £200 pots.  The higher the variance, the more likely you go bust.

If you want to believe it makes no difference, it is fine by me.   If OP believes you over me, that is fine too. 

I know you know what I'm saying is right, we both know that what your saying is also right and that's it's also not really relevant to the discussion...

I got hugely tilted by that bitchy comment about my business, didn't care for the tone. Probably wasn't meant like that so I'll just let it go though.

OP, you should probably run it twice if you can, it's quite a friendly and social thing to do and it is good for the game. If you're tilted/stuck/don't like the guy you're in the pot with /really wanna gamble/don't care if you lose/have a good feeling then just go once, can't go wrong really. Sorry for interrupting your thread with my nonsense.

Good luck to all :-)

[superwomble]

Thanks for the replies, I think I'll probably keep the box ticked when playing Zoom now.

It isn't that I'm 'worried' so much about my opponent sucking out on me, I just hadn't considered the reduction in variance correctly.

How does it work in live? Obviously in Zoom you don't get much choice - if your box is ticked and your opponent's is too then when it's all in you RIT. In videos I've seen of live games where it happens though I think the cards have been flipped before one of the players says "Run it twice?". Is it always the case that you choose to do it after seeing the cards or does it depend on the casino?

[SuuPRlim]

some, admittedly very few, casinos dont allow you to RIT, much more common is them not allowing you to run it 3-4-5-6 times (it happens!) because it's an absolute head-fuck for the dealers and mistakes are inevitable.

Most of the time, you get all-in, show your cards and the conversation will go something like "you wanna do it twice?" from one side, there's no reason why you necessarily why you need to show your cards before any decision is made, but in a live game, typically people would find it a bit anti-social to refuse to show your cards and then ask about running it twice, there's no logical reason for this it's just how it is. Like I said running it twice is kind of a friendly thing to do, typically.

If you want to run it twice i a live game...don't be scared to ask, quite often people just feel a bit awkward like they are asking for some sort of charity and don't bother asking. If you're  a regular in a live game I would personally advise running it twice wherever possible, as it limits the swings for everyone and improves the ecology of the game, as Doobs very expertly pointed out if you're there for the LOL-long-run it doesn't make any difference.

[VM]

This was a good read and some interesting points made.

It looks like Dave / Doobs both agreed that EV is the same if you RIT.

I must have got lost in the maths somewhere, I think I get it if you're talking dice / coins etc but if you RIT in Hold'em, say if you hit your draw on the 1st board with a straight draw similar to an earlier example (AK vs 56 on a 47A flop), you now have 1 less out on the second board or you could hit 2 of your outs on the first board and have 2 less outs on the second board.  Likewise, if you brick the first board, you now have 8 outs + runner runner outs on the seconds board, with 2 more known cards. Does this not mean that the win % on 1st boards vs win % on 2nd board is not exactly the same when you RIT on every occasion? Or have I missed something? Some insight into this would be good.

[SuuPRlim]

Does this not mean that the win % on 1st boards vs win % on 2nd board is not exactly the same when you RIT on every occasion? Or have I missed something? Some insight into this would be good.

Hi Mate,

Yes, it does mean that, with every card that's exposed the win %'s change, the example you gave is spot on. AK vs 56 on A47 you have 8 clean outs from 45 unknown cards  then if you brick you have 8 clean outs from 44 unknown cards. On the next run you have 8 from 43 then 8 from 42.  Every card that's exposed the % chance of winning changes, and this is the same even if you don't run it twice, or run it 5 or 8 times!

The big point is that at the point of making the decision, you don't know this information, you don;t know which cards are coming and in which order (if you did know for example that you'd be hitting your straight on the first card...you could just run it once ), so in the above example if on the first run it comes 3, then 8...like you said the first card of the next run you have only 6 outs from 43 (14%) instead of 7 from 43 (16.3%). Big difference...but you have no way to know this when you have to make the decision, so as the saying goes, you pay ya money...you takes ya chance

Hopefully that clears it up before Doobs comes in and starts babbling about some complicated economic theories 

It looks like Dave / Doobs both agreed that EV is the same if you RIT.

We actually didn't quite agree on this... Basically what i said was that deciding whether to run it twice or not is a NEUTRAL EV decision, so makes no difference to the maths of the hand, in the OP Womble was saying he didn't want to run it twice because he didn't want to give his opponent the chance to outdraw him twice, I understand EXACTLY how that feels but the reality of the maths is that it doesn't make any difference to your EV. If you have 60% equity in a £1000 pot, then you have £600 worth of equity and you can run it 1,2,5, times and that won't change.

Doobs' point, was that actually running it twice is more +EV than not running it twice because you reduce your volatility and therefore the chances of going broke. Doobs' is absolutely correct that the most +EV thing in the world is not going broke. I'll give you a good example using a coinflip (god help me!!)

I give you two options.

1) You can flip a coin with me, you get 11/10 and the minimum bet is 1% of your NET Worth. Minimum amount of flips is 10 and we can flip as many times as you like.
2) You can have 30/1 on a coin flip, min bet is 100% of your NET Worth. Minimum mount of flips is 10.

Option one, you make 10% you can flip forever, betting 1% of your roll every time, chances of going broke is not 0% but it's so close to zero it's fair to say its' pretty much impossible to go broke and even starting at £10 a flip if you flip 24/7 for 2 years you'll be up millions.

Option two.  If you start at £10, and win all 10 in you'll have won £8,196,282,869,808,010 !!! Problem is that chances of winning them all is ~1/10,000. Soon as you go broke you don't just lose your £10...you loose the EV of all the other flips you could have had. Option 1 man won't make a trillion pounds, but he'll sit and print forever.

Even at 30x the ROI, option 2 is not the correct option, and as Doobs says that's mathematically proven. Now change the odds to make them the same and option 2 is even more reckless.

However, the reason why that ^^^ isn't relevant to this discussion is because in the OP we have no information about the size of the pot nor it's relation to wombles Net Worth. If the pot is £1000 and Womble is worth £2000 then running it twice is slam dunk the correct mathematical play, if hes worth £1,500,000 then it's exactly the same.

Also, we don't know that he has a positive expectation in the game (no offence Womble! Just an example) so for example if he has a negative expectation then the maths changes completely, if he has a negative expectation then there is no future EV to protect (just further losses to incur.)

Hope that helps

[atdc21]

Excellent write ups Dave.

[superwomble]

Also, we don't know that he has a positive expectation in the game (no offence Womble! Just an example) so for example if he has a negative expectation then the maths changes completely, if he has a negative expectation then there is no future EV to protect (just further losses to incur.)