[POLL] Just-Dice INVESTORS: Do you agree with lowering the max bet?

[mughat]

If you want less risk simply invest less money. Simple as that. Think about it....

There is no good reason to change the max bet.

Investing less money doesn't change the risk of ruin (ie. of the site going bust).  Whatever you invested, you were risking 1% of it per roll.

I have temporarily set the max profit to 0.25% because with it at 1% nakowa was able to take huge chunks of the bankroll over and over, as he has demonstrated time and time again.  To make matters worse, each time he takes a chunk of the bankroll, more investors lose faith and divest, resulting in the bankroll shrinking even faster.

Risking 1% of the bankroll per bet turned out to incur a much greater risk of ruin than I had anticipated, and so I made the change to 0.25% in an attempt to prevent ongoing huge losses.  At no point was it intended to be "the solution"; it's a stopgap fix until I can put something better in place.  I need to seriously reconsider allowing investors to somehow set their own risk level.  If some want to risk 0.25% per bet and others want to risk 1% then I don't see why I shouldn't let them.

If nakowa continues to win, the 1% people risk going bust, while the 0.25% people gain an ever increasing share of the remaining bankroll.  Eventually all remaining investors (the cautious ones) will be risking 0.25% and we'll be back where we are now, and the 1% investors will have had the chance they're asking for - risk big to win big.

If nakowa does blow up one day, the 1% people get the lion's share of his losses, which they will deserve for 'keeping the faith'.

Of course, maybe he'll never return whatever happens; who can say?  I know he has said many times before that he has quit.

Setting your own risk % would be great! I can't wait for that to be implemented

[Rampion]

If you want less risk simply invest less money. Simple as that. Think about it....

There is no good reason to change the max bet.

Investing less money doesn't change the risk of ruin (ie. of the site going bust).  Whatever you invested, you were risking 1% of it per roll.

I have temporarily set the max profit to 0.25% because with it at 1% nakowa was able to take huge chunks of the bankroll over and over, as he has demonstrated time and time again.  To make matters worse, each time he takes a chunk of the bankroll, more investors lose faith and divest, resulting in the bankroll shrinking even faster.

Risking 1% of the bankroll per bet turned out to incur a much greater risk of ruin than I had anticipated, and so I made the change to 0.25% in an attempt to prevent ongoing huge losses.  At no point was it intended to be "the solution"; it's a stopgap fix until I can put something better in place.  I need to seriously reconsider allowing investors to somehow set their own risk level.  If some want to risk 0.25% per bet and others want to risk 1% then I don't see why I shouldn't let them.

If nakowa continues to win, the 1% people risk going bust, while the 0.25% people gain an ever increasing share of the remaining bankroll.  Eventually all remaining investors (the cautious ones) will be risking 0.25% and we'll be back where we are now, and the 1% investors will have had the chance they're asking for - risk big to win big.

If nakowa does blow up one day, the 1% people get the lion's share of his losses, which they will deserve for 'keeping the faith'.

Of course, maybe he'll never return whatever happens; who can say?  I know he has said many times before that he has quit.

In my book Nakowa just demonstrated he is up 13k, which is unlikely but very possible. And today he lost aprox. 3k.

Then, I'm assuming you calculated the "ruin risk" before taking such a decision: could you share with us what is the precise probability of Nakowa ruining the Casino, considering 1% max profit and a 23k roll for the Whale?

My "guesstimate" is that it's totally negligible.

Secondly, you speak about 1% max profit investors are "keeping the faith". In what?

In the site NOT being rigged, or in math and probability?

I've absolute faith in you, and therefore on the site. And you simply do not "have faith" in math and probability.

[tucenaber]

mechs, do you even realize why most people here are disagreeing with you?

Because they understand math.  They know that 1% is the most profitable bet size.  

For someone who claims to think 'long-term' you would know that mathematically nakowa would lose all his bankroll if we allowed him to keep playing.

So, you obviously don't think 'long-term'. You think short term because you are a greedy investor who wants the bet size dropped at the expense of other investors who are down and will never be able to recover their losses.

You're basically just greedy and selfish.  You want all the reward without the risk at the expense of other investors who now will never recoup their losses.

Once again, I have an opinion but I am not a deicison make.  And yes, I want to make profit just as the players do.

Well, you actually destroyed any short-term profit expectation for those of us who actually understand a bit of math and gambler's psychology, and also understood what kind of investment JD was, and therefore DID NOT DIVEST during the huge negative variance caused by Nakowa. You scared our best customer away by lowering the max bet he loves while he was playing, now we are roughly 30% down, and it will probably take many months to just reach b/e.

Man, do you realize that guy loves to make 300BTC bets? And he thinks he actually has a system that beats the house edge? And he says he is not a gambler and he just kept coming back? What would you think that guy would have done when he started to lose big? I'll tell you: he would have lost more, faster (just look how fast he lost the last 1,000BTC he sent to the site today, compared to the first 1,820BTC). That's basic human psychology at work. Nakowa is the prototypical gambler, he has EXTREMELY DEEP POCKETS, and you scared him away just because there was a minuscule chance that he could have won the whole roll?

Anyhow, anyone better than me at math could calculate what was the exact probability of him winning the whole bankroll, considering that:

• he has aprox 23K BTC
• house bankroll was aprox. 30k BTC at that moment
• house edge is 1%
• max profit is 1% of house bankroll

Mechs, I saw you in the chat: while we were all commenting how "percentage/nakowa" was gambling, you were whining. "I don't want to see this", "i feel bad", "i know how this story ends, he always starts losing big and then he goes up +6k btc", blah blah blah (I quoted you very loosely but you know that's the spirit of what you were saying). Shortly after Doog lowered the max profit.

You know what? You were just crapping your pants, you were overemotional and you didn't act rationally, triggering a -EV decision (lowering max profit, pissing off the whale). "i know how this story ends"? YES, YOU SHOULD HAVE KNOWN. MATHS HAVE NO FEELINGS. THE STORY ENDS NAKOWA ENDS UP LOSING EVERYTHING IF YOU LET HIM PLAY LONG ENOUGH. Unless you crap your pants and scare him...

I agree completely with you.

To answer your question:

There is no way to completely beat the house since only 1% of the money can be taken in each bet. But he could take e.g. 90% or 95%.

assuming he bets max all the time, he can lose log(23000/30000+1)/log(1.01)-1 = 57 bets in a row before going bust
and he can win log(1-0.9)/log(0.99) =229 before taking 90% of investments.

So the risk of him winning 90% is (r^57-1)/(r^(57+229)-1) = 0.7%. (r=0.495/0.505)
The risk of taking 95% is 0.17%

(I think)

[RationalSpeculator]

Risking 1% of the bankroll per bet turned out to incur a much greater risk of ruin than I had anticipated...

If nakowa continues to win, the 1% people risk going bust ...

 

Dooglus, please explain how high the risk is to go bust when 1% max bet?

From the simulations done the worst case scenario was losing 80% of bankroll.

Never did it happen that the bankroll went bust and in the end the whale always lost everything.

How high are the chances for this to happen, if any?

[TheFuneral]

As many of you know, Just-dice investors lost a lot of BTC to Nakowa.  He currently has 13,000BTC of investor money in his hands. As a result, dooglus lowered the max bet to 0.25%

Nakowa has stated that he will never play again since the max bet has been lowered.

As an investor, Are you happy with lowering the max bet?


EDIT: Some things to consider

PROS:
-- Less variance for investors.

CONS:
-- We must assume for now that Nakowa will not play again (this might change, but he said he won'y play again at this bet size). This means there is no way for investors to recover the 13,000BTC losses.
-- As BTC exchange rate rises bet sizes will drop. As time goes on without nakowa, its possible that the site will never hit +6000 profit making it impossible for some investors to recover their losses.
-- Changing bet size without notice wasn't good for gambler confidence.

If you want less risk simply invest less money. Simple as that. Think about it....

There is no good reason to change the max bet.

isn't that the whole reason why the site was made this way? So it could be left alone? Looks like emotions are sneaking in.

[Peter R]

PROBLEM 1: We have mis-treated our best customer by changing the max bet as he was playing, without evidence that he ever cheated.

Is there some way we can spin this into a positive?  Admit that we mistreated him and then give him VIP status (like his name in the data stream in a more bling-bling colour, his name on a Wall of Fame or Greatest all time Gamblers, some other idea to stoke his ego???)  He asked earlier "don't you think I proved a point?"  Well, he did: he proved that we were not as confident in our advantage as we thought we were.  He deserves some cred here... 


PROBLEM 2: Are we 100% confident that 1% max bet does in fact optimize profits over the long term given fair dice rolls (e.g., does the Kelly criteria consider the dynamic nature of the max bet size and the size of the gambler's bankroll)?

If we do know this as fact, then I don't understand any rationale to make the site "sub-optimal": wouldn't you just be trading expected profits for a more stable income?  If you want more stable profits, then diversify! 

[integrity42]

The number of people disagreeing with Dooglus outnumbers those agreeing by 2:1.

I hope he reconsiders what he's done.  He listened to mechs instead of listening to 30,000BTC in investors who left their money in at 1%.

[mughat]

The number of people disagreeing with Dooglus outnumbers those agreeing by 2:1.

I hope he reconsiders what he's done.  He listened to mechs instead of listening to 30,000BTC in investors who left their money in at 1%.

+1  
Setting your own risk % is the solution to this madness.

[RationalSpeculator]

Dooglus, and all others that are for lowering the max bet:

Are you aware that expected returns are lower than 50% per year without the whale?

[Peter R]

The number of people disagreeing with Dooglus outnumbers those agreeing by 2:1.

I hope he reconsiders what he's done.  He listened to mechs instead of listening to 30,000BTC in investors who left their money in at 1%.

+1  
Setting your own risk % is the solution to this madness.

What would happen then if I always increased my risk % to something high like 2.5% when the fishes were playing, and then decreased back to 0.25% when a whale comes along?  

[integrity42]

What would happen then if I always increased my risk % to 25% when the fishes were playing, and then decreased back to 0.25% when a whale comes along?  

What will happen is that smart investors with bots will take a majority of the winnings when small fish are playing, and then will drop down to 0.25% when the whales play. Fucking the low-risk investors like mechs who will be stuck at 0.25% by diluting them.

This is exactly the bot that I will be coding.  If it is released as open-source and everybody uses it, you will see wild fluctuations  of max bet size when whales start playing.  This will have the effect of scaring away whales and large bettors who will see the max bet size drop as soon as they start martingaling or playing large.

Of course it would be fairer to all investors if the max bet was static at 0.50%  or 1%, but now investors will be gambling against each-other by diluting each-other at the right times.

[Oleander]

PROBLEM 2: Are we 100% confident that 1% max bet does in fact optimize profits over the long term given fair dice rolls (e.g., does the Kelly criteria consider the dynamic nature of the max bet size and the size of the gambler's bankroll)?

There is some disagreement on this.  I've argued that the JD situation is different enough from the Kelly model that the 1% max bet is not necessarily optimal.  And I think that Nakowa is evidence of this.  Others think the Kelly Criterion is applicable, and all we need to do is wait for Nakowa to go bust.

It's a difficult question... probabilities are notoriously counter-intuitive.

Since we have people who believe in the Kelly Criterion despite our recent experience, maybe the best solution all around is to let investors set their max bet percentage individually.  If the Kelly supporters are right, they'll prosper in the long run.

[tucenaber]

The number of people disagreeing with Dooglus outnumbers those agreeing by 2:1.

I hope he reconsiders what he's done.  He listened to mechs instead of listening to 30,000BTC in investors who left their money in at 1%.

+1  
Setting your own risk % is the solution to this madness.

What would happen then if I always increased my risk % to something high like 2.5% when the fishes were playing, and then decreased back to 0.25% when a whale comes along?  

You will lose the big wins. Think of it as a less extreme version of divesting when the whale comes along.

[Oleander]

What will happen is that smart investors with bots will take a majority of the winnings when small fish are playing, and then will drop down to 0.25% when the whales play. Fucking the low-risk investors like mechs who will be stuck at 0.25%.

This is exactly the bot that I will be coding.

This is no different in principle from divesting when the whales play.  It's not a new problem.

And really, switching between 1% and .25% is less of a problem than switching between 1% and 0%.

[jgarzik]

+1  
Setting your own risk % is the solution to this madness.

Indeed.  Let investors (the market) decide.  It is their money being risked.

[Deprived]

Either he was cheating or he wasn't.

If he was cheating (which I still believe VERY unlikely) then the solution is to fix whatever needs fixing to stop the cheating working - lowering max bet would be shuffling the deck chairs on the Titanic.

If he wasn't cheating then it's a pity to have changed J-D from being the site whose motto could have been "we welcome whales" to "we shit our pants if a whale swims anywhere near".

And mechs not wanting him to play there because he thinks the whale is an asshole is stupidity climbing to a new peak.  If I offer -EV bets then I'd FAR prefer to have assholes betting against me than people I actually like.  But I'd happily take action from either - that's what the business is for.

The problem I see here is predominantly one of fear from people who want to dump all/a large part of their investment in J-D and treat it like a bank.  Rather than treating it as just one high-variance/high-expectation item in a diverse investment portfolio.

I personally divested yesterday for the first time - but only because I needed more liquid BTC immediately after BTC-TC announced closure (to try to grab some bargains - the few hundred BTC I keep as cash having been used to buy my own investors on BTC-TC/LTC-Global out immediately at full value).  Those funds would have been returning to J-D within weeks - they still might, but if the site loses its big action then EV is going to fall a lot (I don't care about variance - my own bankroll management handles that).

EDIT:  I have no problem with the idea of investors setting their own risk level.  I wasn't keen on it - but it's far preferable to the house using a non-optimal kelly that we can't compensate for.

[Peter R]

The number of people disagreeing with Dooglus outnumbers those agreeing by 2:1.

I hope he reconsiders what he's done.  He listened to mechs instead of listening to 30,000BTC in investors who left their money in at 1%.

+1  
Setting your own risk % is the solution to this madness.

What would happen then if I always increased my risk % to something high like 2.5% when the fishes were playing, and then decreased back to 0.25% when a whale comes along?  

You will lose the big wins. Think of it as a less extreme version of divesting when the whale comes along.

OK that makes sense.  But what if I also increase to *over* 1% when the fishes are playing?  Like Oleander pointed out, would I not be diluting all the other investors that don't actively manage their account?

[integrity42]

The number of people disagreeing with Dooglus outnumbers those agreeing by 2:1.

I hope he reconsiders what he's done.  He listened to mechs instead of listening to 30,000BTC in investors who left their money in at 1%.

+1  
Setting your own risk % is the solution to this madness.

What would happen then if I always increased my risk % to something high like 2.5% when the fishes were playing, and then decreased back to 0.25% when a whale comes along?  

You will lose the big wins. Think of it as a less extreme version of divesting when the whale comes along.

OK that makes sense.  But what if I also increase to *over* 1% when the fishes are playing?  Like Oleander pointed out, would I not be diluting all the other investors that don't actively manage their account?

What stops a big investor with 10,000 coins taking their risk to 5% when small fish are playing. Then when whales play, their bot automatically drops the risk to 0.25%, diluting investors like mechs.?

So large investors with a bot that actively manages the risk will fuck smaller investors?

Either way, I think if the max bet is constantly changing so drastically, then whales won't play.

[Deprived]

The number of people disagreeing with Dooglus outnumbers those agreeing by 2:1.

I hope he reconsiders what he's done.  He listened to mechs instead of listening to 30,000BTC in investors who left their money in at 1%.

+1  
Setting your own risk % is the solution to this madness.

What would happen then if I always increased my risk % to something high like 2.5% when the fishes were playing, and then decreased back to 0.25% when a whale comes along?  

You will lose the big wins. Think of it as a less extreme version of divesting when the whale comes along.

OK that makes sense.  But what if I also increase to *over* 1% when the fishes are playing?  Like Oleander pointed out, would I not be diluting all the other investors that don't actively manage their account?

Easily fixed - only allow people to change their risk level once per day.  Then people can't game it - and have to pick the risk level they'll take for ALL action.  The house doesn't need backing for the small bets - so if you want a share of those you should also have to take a similar part of the big action.  People could still divest/reinvest when whales were around - but there's no stopping that whilst giving people control of their money.  And the people who routinely divested when he bet before at 1% risk will likely do the same at .25% risk if a whale starts winning again - so it's not a new problem.

[Peter R]

The number of people disagreeing with Dooglus outnumbers those agreeing by 2:1.

I hope he reconsiders what he's done.  He listened to mechs instead of listening to 30,000BTC in investors who left their money in at 1%.

+1  
Setting your own risk % is the solution to this madness.

What would happen then if I always increased my risk % to something high like 2.5% when the fishes were playing, and then decreased back to 0.25% when a whale comes along?  

You will lose the big wins. Think of it as a less extreme version of divesting when the whale comes along.

OK that makes sense.  But what if I also increase to *over* 1% when the fishes are playing?  Like Oleander pointed out, would I not be diluting all the other investors that don't actively manage their account?

What stops a big investor with 10,000 coins taking their risk to 5% when small fish are playing. Then when whales play, their bot automatically drops the risk to 0.25%, diluting investors like mechs.?

So large investors with a bot that actively manages the risk will fuck smaller investors?

Either way, I think if the max bet is constantly changing so drastically, then whales won't play.

Haha!  So the logical investor strategy would be to actively manage.  Just-Dice would be a tease: advertise 500 BTC max bets, and then as soon as we attract a whale and start to dance, the max bet plummets.