Quote from: Keyser Soze on January 14, 2014, 02:35:11 AM

By investor risk I was referring to the max profit per roll percentage, currently 0.5%. I would not want to create different betting pools as dicenow has, for example.

I dont see that this is needed too. The house advantage at just-dice can stay at 1% like it always was, no other pool needed in my eyes. The kelly value is different from the house advantage.

Quote from: Keyser Soze on January 14, 2014, 02:35:11 AM

Right now the site takes the results of any bet and distributes it based on each investor's portion of the bankroll. Only investing or divesting causes that percentage to change. If investors were able to adjust their risk (their portion of the max profit per roll), then the site would have to recalculate investor's percentage of ownership much more frequently. Not to mention implementing a different method to properly account for wins/losses.

Im not sure if im wrong now but i believe that wouldnt be a problem. At the moment there is only one value. The house. Every investment and divestment is calculated to or from it. When someone would invest at 1% kelly then the house would grow by the investment in total. The max profit wouldnt be a fixed value then anymore, taken from the house value, it would be its own value. So when someone adds an investment at 0.5% then the max profit is rising 0.5% of that investment. If someone invests at 1% kelly then max profit is rising 1% of the new investment.
While playing the house value and the max profit value is adjusted accordingly to the results of the bets and when someone is divesting the change of max profit value in comparison to the value of max profit at the time of investment is taken for calculating the profit or loss. I think thats just like it works now and it should even work with fully free adjustable kelly values. Only that the calculated max profit is independent from the kelly value then.
I might have a thought error i didnt see though.

Quote from: Keyser Soze on January 14, 2014, 02:35:11 AM

Imagine we had many investors all with various amounts of risk (0.1% to 1% profit per bet, for example). Now think of how the site would handle bets of different sizes, not just max profit bets. As far as I know, there is no easy calculation to solve this problem. If you do have one I'd love to hear it.

Now I am not very familiar with server management, but I could see how a significantly more complex calculation for each bet could greatly increase the server's load.

This issue was discussed in great detail awhile back in this thread. I assume nothing has happened yet because it is not an easy change to make, nor is there an consensus in the implementation.

Edit: I've looked at dicenow, but don't really see how they are calculating the variable risk. It mentions leverage and I don't know why. The variable risk should have nothing to do with leverage.

I guess! that the server at the moment not even is using 5% of the CPU. I wouldnt know otherwise what its used for. I think the variable kelly wouldnt need a much higher CPU-Usage because the calculation of max profit out of the house or max profit calculating out of the last bets shouldnt mean a difference that is so much away from current status. (Under the premise i mentioned above...)

Quote from: GeoRW on January 14, 2014, 05:23:10 AM

This would be disadvantageous for current investors who want to stay at 0.5% Kelly. For the same amount of variance (or slightly higher) they would get less profit. Max profit is hit very rarely these days and only there you would experience full Kelly variance. Moreover, daytrading at these high-variance days has hit investors who did not divest greatly, which is not good for Dooglus and the site.

So you dont want to take a higher risk but dont want that others take it too? I dont see the problem. Either you want the higher profit and take the risk or you play it safe. But at the moment its mathematically nonoptimal because some didnt want to take the risk of variance. Even though the past has shown that stochastics work pretty fine like it should.
Its correct that max profit bets are played seldom but then again... thats what a risk setting would be for. Setting your personal comfort zone between risk and possible gain.

Daytrading? For me it looks more like people invest when the profits are on top and divest when its below the value it should be. But i didnt observe that long enough maybe.

Anyone knows when the charts at https://bitcoinproject.net/ are back?