Plus when PrimeDice still had public stats, this is months ago if not last year PrimeDice already had 4000+ profit. That is at least a billion bets ago.
Are you sure PrimeDice ever had public stats?
I remember they used to show the number of bets and the total amount wagered, but I don't think they ever showed their profit.
They stopped publishing the amount wagered about a month ago:
Quote
Tue Sep 30 11:14:09 PDT 2014 BTC WAGERED: 479,855.2881 TOTAL BETS: 1,165,821,154
Wed Oct 1 09:26:44 PDT 2014 BTC WAGERED: 481,527.3520 TOTAL BETS: 1,173,764,666
Thu Oct 2 11:09:28 PDT 2014 BTC WAGERED: 484,587.2089 TOTAL BETS: 1,184,083,585
Fri Oct 3 12:36:05 PDT 2014 TOTAL BETS: 1,194,817,217
Sat Oct 4 10:53:32 PDT 2014 TOTAL BETS: 1,203,248,935
Sun Oct 5 10:05:42 PDT 2014 TOTAL BETS: 1,213,617,864
but given a house edge of 1%, we would expect a profit of around 5000 BTC at that point. Of course, the actual profit could be vastly different in either direction.
He probably figured a way to determine when the middle part that loses either way, above 33.0 and below 66.99, has passed its cycle and just bounce between <33.0 and >66.99 depending which one is going into positive variance.
According to Wikipedia:
The squared term means that variance is always non-negative. So what is "positive variance"?
