What language is that even?

1000 CLAM gets you 3 CLAM/day from stacking.

The bankroll has been hit when the players won 5700 CLAM two days ago.

I have mixed feelings about theoffline idea. Surely it is innovative, but is it fair? People are allowed to lie about their offline stakes which basically means that if they lose their investment, the max bet will decrease dramatically, as they cannot fill up again with their non-existent coins.

Secondly, math tells us that 1% max bet is the fastest way to increase your profits. Now people are allowed to increase their risk (when lying) to more than 10% against their own interest (the fact that some JD investors aren't proficient in statistics was already clear though.)

In the end I find myself making a guesswork of how much I should declare offline in order to reach a 0.5% of risk that I want to afford.

Who are you arguining that it is unfair to?

* the investor who is 'lying' can lose his coins pretty quickly - but he knows that going in, so that's not unfair

* the player who wins big sees that the max profit drops a lot after his win - but he just won big, so he's not complaining

* the investor won isn't lying sees his share of the bankroll diluted by 'lying' investors - but he gets to stay invested when the liar investors are forced out by a winning player, and gets a much bigger share if the winning player turns into a losing player after that


I think it tells us that risking 1% of your investment on every bet maximises the expected value of the logarithm of your bankroll - or something like that. To maximise the expected value of your bankroll when you have an edge you should risk everything on every roll shouldn't you? Your expected profit is 1% of the amount the player risks, so you want them to risk as much as possible against you. The reason we look to maximise the logarithm of the bankroll is that when you have 100 coins, losing 100 is catastrophic, but winning 100 isn't equal and opposite. 100 and 200 are similar amounts, one being only twice the other, but 100 is many times bigger than 0.


We can't tell whether it's against their own interest or not, or whether they're lying, since we have no idea of the size of their whole bankroll. Most of it won't be CLAMs anyway.

If you want to risk 0.5% of your bankroll per roll, calculate the size of your bankroll, work out what that comes to in CLAMs, subtract your onsite investment size from that number, and that tells you how much you should declare to be '/offsite'. For this you need to decide what constitutes your bankroll. You'll also need to regularly monitor the situation - perhaps the price of CLAM changes, or the value of the non-CLAM part of your bankroll.

You should ask me to translate, I don't use Google translator, my translation service link  https://bt.irlbtc.com/view/759494

This is in reply to a post on the CLAM thread.

I'm answering it here because it doesn't really belong there.


It's tricky. Suppose you deposit 100 CLAM and a minute later I withdraw 100 CLAM. I'm going to get the 100 you deposited. Should you have to wait 4 hours before you get staking rewards? I don't see why you should. The full onsite invested amount is now staking... I have things set up such that new deposits are used to service withdrawals, and so there's never a huge amount of coin 'waiting' to age.

It's almost always true that the amount of coin in the wallet that is waiting its 4 hours to start staking is smaller than the amount that is in people's balances but not invested. So you end up getting more than your fair share of the reward anyway.


I don't like the idea, and it would be a real pain to implement. Currently I don't track each investor's balance, I only know what percentage of the total bankroll they own. We can handle 100 bets per second because each bet only has to update the site's bankroll, not 200 different investor balances. If each investor had a different range of bet sizes they were willing to get involved with it breaks that scheme, and I have to update all the investor balances for all bets.

What's your motivation for suggesting such a scheme? How would it be fair for you to take all the nice low variance bets and leave the more dangerous bigger bets to other people? I don't think it would be.

First off, note that the word is "stake", as in "an interest or share in an undertaking or enterprise", not stack. By owning CLAM you have a stake or interest in the whole CLAM economy, such as it is, and get to earn rewards proportional to the size of your stake.


It's true, the whole wallet stakes, which includes onsite investments, balances, and any commission that has been charged by the site but not yet credited to an account. The 90% of each stake reward is shared between investors in proportion to the size of their onsite investment and the other 10% goes to the site as commission. The rewards don't count as profit for the purposes of calculating commission at the end of the week, since the 10% is taken off before the rewards are distributed.


I don't track it, but whenever I look it is between 2k and 4k, usually much closer to 2k. Right now it's 2037.72475057 CLAM. I expect somebody divested and forgot to reinvest their 2k. Players don't often leave coins in their balance - they either lose them, invest them, or withdraw them. There's nothing to be gained by leaving coins sitting idle in your balance when putting them in the bankroll makes them able to receive staking rewards.


I understand that some investors don't want the risk of bankrolling large bets, but there's nobody who ONLY wants to bankroll the large bets, or to bankroll more than their fair share of large bets.

It's clearly much better to bankroll a million small bets than one big bet worth the same amount. The small bets are the most desirable.

Allowing some investors to have an unfair share of the small bets necessarily means punishing the other investors with an unfair share of the bigger bets.

He is referring to balances not invested, just idle in users accounts.

I'd assume the 2037 figure is the total balance sitting idle in accounts.