BitBet open beta.

[cosurgi]

(There's also an FAQ.)

What is a bet's weight?
The earlier you bet the more of the total pot you will win in case your side prevails.

What formula do you actually use?

Let's say that I won a bet. I placed A bitcoins on this bet, and the total bitcoins placed on my side was A_sum. The total sum was B bitcoins (the opposite bet was C_sum = B-A_sum bitcoins), and I placed this bet at percentage of time left till end equal to T, expressed as: T= 100%=1.0 is when bet starts, and T=0%=0.0 is when bet ends.

Then what is my reward?

R(A,B,T)= ?

Explained here: https://bitcointalk.org/index.php?topic=147914.msg1572084#msg1572084

Or in other words, the losers side is split between the winners by coin*weight from which 1% house fee is taken.

It's only explained qualitatively. I am asking about exact formula used.

For example this one formula is NOT used, even though that's what you are suggesting:

Reward = TOTAL * 0.99 * MY_BET * WEIGHT / Sum_all(winning_bets * WEIGHT)

The formula which I found to ALMOST match the results is:

Reward = TOTAL * 0.99 * MY_BET * WEIGHT^(1/3.5) / Sum_all(winning_bets * WEIGHT^(1/3.5))

Where "^(1/3.5)" means raised to power (1/3.5). This formula however becomes increasingly wrong for WEIGHT smaller than 10000.

So what is the formula?


EDIT: I found a formula which is almost correct for WEIGHT > 3000, it is:

Reward = TOTAL * 0.99 * MY_BET * SQRT(LOG(WEIGHT))*WEIGHT^(1/6.0) / Sum_all(winning_bets * SQRT(LOG(WEIGHT))*WEIGHT^(1/6.0))

But I am still asking what is the formula actually used.

[kakobrekla]

(BetAmount + (LosersBtcSum / WinnersWeightedBtcSum)  * ( BetAmount *  BetWeight ) )*0.99

[cosurgi]

(BetAmount + (LosersBtcSum / WinnersWeightedBtcSum)  * ( BetAmount *  BetWeight ) )*0.99

Thanks, it agrees perfectly. Using my notation, that would be:

Reward = 0.99 * (MY_BET + LOSERS * MY_BET*WEIGHT/Sum_all(winning_bets * WEIGHT) )

EDIT: if you change the formula, please let me know https://bitcointalk.org/index.php?topic=147914.msg1576743#msg1576743

EDIT: so that winning bets do not lose money:
BetAmount + ((LosersBtcSum / WinnersWeightedBtcSum)  * ( BetAmount *  BetWeight ) )*0.99 or 0.98

[Monster Tent]

Under .US nexus requirements .US domains may be registered only by the following qualified entities:
Any United States citizen or resident,
Any United States entity, such as organizations or corporations,
Any foreign entity or organization with a bona fide presence in the United States

As you are located in Romania how can you possibly use a .us domain ?

[MPOE-PR]

Under .US nexus requirements .US domains may be registered only by the following qualified entities:
Any United States citizen or resident,
Any United States entity, such as organizations or corporations,
Any foreign entity or organization with a bona fide presence in the United States

As you are located in Romania how can you possibly use a .us domain ?

I fail to find the "CANNOT BE LOCATED IN ROMANIA" clause in your quote. Why do you find thinking so difficult?

[MPOE-PR]

BitBet FAQ updated.

[MPOE-PR]

Since its beginning BitBet has been tolerant of bets placed following an event that effectively ends the bet, refunding all such late wagers upon resolution. This policy was intended to be forgiving of the occasional errant bettor who might not've noticed that a currency had already traded at a certain price, that a verdict had just been read in a court case, or who had otherwise just missed a cutoff.

As users have gradually shown over time and have exemplified in the extreme today, BitBet's tolerance is taken for a chance to swindle the operators and legitimate users by spamming bets after the resolution criteria has been met.

Henceforth, and in the interest of fairness and uniformity, all wagers received, whether refunded or not, will be subject to the same 1% fee.

The FAQ will be updated presently to the same effect.

[globalvillage]

Ha Ha Ha  legitimate affiliate for bits, LOL

[kokojie]

I think the bet approver need to take note of the "close date" not too close to the event resolve date, fore example (this is a real bet on the site currently):

A bet that is: Bitcoin difficulty will fall in 2014, and close date is... 2014-12-15

This is not good, since any big bettor can just wait until the last minute to make sure difficulty will not fall, and bet big to compensate for weight.

A sensible close date should be something like 2014-10-01, so that there's 3 full months of uncertainty, and will not be unfair to the early bettors.