Bankroll is definitely relevant. Actually because the BR is so relatively small at peerbet your expected return is relatively high since with 1 coin you will have a bigger percentage of the winnings (and losses.)
Incorrect, The size of the bankroll and your % of the bankroll is irrelevant, if your ฿1 investment is 0.1% of a bank that is turned over 100 times per year with an investors edge of 0.9% you expect the exact same return as if your ฿1 investment is 10% or whatever% of a bank that is turned over the same amount of times and has the same investors edge.
What you might be getting at is that you could assume a smaller bank will be turned over more often and that would increase your expected return.
You are joking, right?
It is something like this, right?
Yearly expected return = ( yearly wagered * edge = expected profit ) / virtual br *100%Obviously bankroll is relevant since it's in the formula.....
Sure: if wagered is low, your return will be still bad despite BR. Sure: EV can be same even w/ bigger BR if wagered is bigger (eg: bank turnover is same.) But
BR is relevant, just like wagered and edge are relevant.
Just because you like to use the term "bank turnover" and do something like this:
Yearly expected return = 365 / [(virtual br / (tracked wagered / tracked days)) = bankturnover] * edgeDoesn't change anything, since "virtual BR" is still in the formula and therefor relevant.
Since most sites allow investing on margin, bankroll shouldn't be used though but the "virtual bankroll" should be used. I tried to explain it before to OP but he seemed confused and I can't care too much to explain further. Without the "wagered" numbers it's unclear if the given numbers are correct but you should probably assume the "turnover / expected annual return" numbers are incorrect.
When I checked most sites did not provide a virtual bankroll figure, I am not confused by what it is or what figures should be used. Obviously the wagered numbers are used to calculate the turnover and the expected annual return.
The question is if you calculate turnover with: bankroll/wagered or virtual bankroll/wagered. It should be the latter one and "bankroll" is completely irrelevant. Since you do list bankroll (and not virtual bankroll) it is still not clear which one you use. Calculating the virtual bankroll is btw dead easy, most times just max profit * 100 if edge is 1%.
Example:
Site has 246 BR, but 245 on 10x and only you are on 1x. So virtual BR is 2451 BTC. Your return is 1/2451 of winnings and losses. Let's say wagered is 13529 coins in a year and investors get full 1% edge. We get:
EV = (13529*0.01)/2451*100 = 5.52% per virtual coin invested.
So for the 1x guy his profit would be .0552 per coin. For the 10x guys that would be actually 0.552 per coin. This adds up: (0.552*245+0.0552*1) = 13529*0.01 = 135.29 BTC profit.
If we use just normal BR:
EV = (13529*0.01)/246*100 = 55% per coin. That is wrong. Because the profit is: 13529*0.01 = 135.29 and you as 1x investor get only 1/2451 of this = 135.29/2451 = 0.0552 - not close to the promised 55% per coin. I am assuming you are doing this which explains the high wrong numbers for BitDice.
It is even pretty clear you use just bankroll:
BitDice has a virtual BR of 10,010 BTC. You say they need 10 days for bank turnover which is: 10010/10 = 1001 a day. 1001*195 days that you are tracking = 195,195 wagered. BitDice is currently at 102,132 BTC wagered in all-time of it's existence so you are using "normal BR" - which is wrong as explained above.
If the "wagered" numbers would be there I could calculate the exact correct bank turnover and expected annual return, but now I cannot. I am still sure they are incorrect though. I hope you understand what I mean now. Correct me if I am wrong.