If the bankroll was 100 Bitcoin and the house edge was 1%, the max bet would be 1 Bitcoin.
I'll remove all references to kelly criterion tomorrow morning to avoid any potential confusion and make the structure defined here and before in the thread that this is how it currently is.
We are in the midst of changing the whole system and re-writing all the areas, so this will be a fill-in until that time. When we announce the new changes, I will make sure that all updated changes and terms are clearly defined and available to see.
With the new changes, investors will be able to adjust their own level of risk. There are some that are in agreement that think anything more than half kelly is the right investment strategy, while there are others that are very willing to leverage their investment as much as possible. I think it is good that the investor has that personal choice.
That is entirely reasonable, and if its clearly communicated removes my reservations. FWIW: If you do that, and a gambler does max-bet that's going to be -expected bankroll growth for investors (like i said, earlier). I personally wouldn't invest, but if investors are aware it's entirely reasonable and their choice (and risky +EV for them) |
|
|
|
We apologize if we misunderstood that you were accusing us of being -EV from the statement of putting "investors into negative expected growth territory". From your admitted statement of our investment propositions being +EV, in its very nature, would mean that the expected value would continue to be positive going forward in the long run, so we're a bit unclear by what you really mean. Perhaps this was just an honest mistake.
Please read the kelly, and learn the difference between expected growth and expected value. You can (and probably are) +EV but -bankroll growth. This should be stuff you did 5 months ago. This is the second time you've called me out on my "mistakes" which were correct. I don't mean to be so hostile, but seriously.. To be completely fair, it is hard to call us negligent when the code remains the same as you have written it. We made one change which was to the commission structure to 20% of the house edge being wagered.
lol, the "one change" can break everything if done incorrectly. You can't throw in a random constant in on one line, and expect the previous safe-guards to not break very badly. We understand your concern about user fund security and thus hold no hard feelings over what we believe to be a misunderstanding. We take user fund security very seriously as well.
No hard feelings either. Just to be clear, I think the extra commission is entirely reasonable and important for everyone (apps, gamblers, investors) that there's a sustainable business model in place (which you seem to have). It's better for investors to have 30% of a large-number, than 10% of a small number. I get it. You just need to communicate it clearly and make sure you're not over-exposing investors (or at least, tell them they're no longer at a 1x kelly and at a Y). |
|
|
|
Moneypot takes 20% of the house edge, App owners take 50% of the house edge, 30% of the house edge is the advantage that investors hold over the players.
This is a very reasonable business model, and I don't think there's anything wrong with it. The important part is just being honest and clear about it, so everyone knows what they're getting into. Earlier in the thread you wrote the MP commission was 0.2%, which fine, was a mistake. Then you literally said the commissions are "no where near 70%" using a contrived example of how it's 7%. Now you're being clear, and that's great. Anyway, I'm glad we're in agreeance upon this. This will amount to 70% of the HOUSE EDGE of wagered volume as commission. It is not 70% commission directly taken from your profits which some may have been led to believe from the traditional use of the word commission.
70% commission of profits is better for investors than 70% of the house edge  (But worse for MP or Apps, who don't want to eat that variance) To onlooking investors, please take the time to understand and realize what is written. We encourage all investors to do their due diligence and are always standing by to answer any questions or concerns you may have. You have still side-stepped the most relevant question for investors, what risk level is investors money being risked at? From my reading of response implies a ~3.33x (?) kelly, but I suspect that is a mistake. Could you confirm what that is really the case? I'll ask the earlier question again for clarity: if the bankroll was 100 BTC, and someone was betting on 1%-house-edge binary bet ... what would the max-profit be? |
|
|
|
I'm not at all trying to be dishonest or disingenuous about it.
I stated what is an objective fact, your amortized commissions are 70%. I didn't mean that as an attack, or that it's bad (in itself, it isn't). I meant that as a fact. Which it is. To which you reply: I think you may have confused the commission as well. It's nowhere near 50 or 70%. If using the above example, the investors would be effectively paying a 7% commission on their profit.
Which is both insulting, and a plain lie. Even how you derived the 7% commission is using the same definition of commission I am, so you can't even pretend it's a terminology mistake. It kind of seemed like you're setting these questions up so that you can knock us down when we respond.
That's not my intention, I just wanted straight answers. I have no quarrels with you and I don't know why there seems to be tension between us.
I think when you're holding other peoples money, there's a certain duty of care which you're not showing. At least in my books, when you hold other peoples money, tell them you're risking it according to a certain schedule, and don't -- that's negligent. I honestly don't even think 90% commissions is unreasonable, and there's many investors that would line up for that if you can give them enough turn-over and risk management. If bettors were all to hypothetically bet to the point where they reach the long run with their bets and with that much volume, I think any investor would be significantly pleased knowing that the greatly reduced volatility over a much larger sample greatly benefits their return on investment position even after including the app fee and house fee payouts.
Of course, that's why investors invest. The more bets the better. I'm of the firm belief that our investment is +EV.
Well obviously. That's pretty hard to mess up (although I can name two casino that did). You seem to have accused us of otherwise
Huh? I have never said, or implied such a thing. If anyone told you I said that, they are lying. And if your investments were -EV, it's probably more likely that I'd start gambling on your site than saying anything. (for reasons I do not know, especially since you're the one that set it up this way)... Can you please explain why you feel this way and why you feel this way so strongly that you're saying it disqualifies us from being able to run a casino?
I'm saying that if your previous answer to me is correct, you're risking >3x as much for investors as you should. When I originally wrote MP, if the bankroll was 100 BTC the max-profit on a 1% game would've been 0.5 BTC. Which is the fully kelly. You added an extra 20% commission, and are saying that max-profit is 1 BTC. When in actuality you should've decreased the max-profit to keep the bets kelly-compliant. See my earlier point about negligence. |
|
|
|
I think you may have confused the commission as well. It's nowhere near 50 or 70%. If using the above example, the investors would be effectively paying a 7% commission on their profit.
I hate to be blunt, but that's just both disingenuous and dishonest. Yes, in the absolute most extreme case where someone only does a single giant bet, the commissions might be 7%. In reality, even when people lose money they turn over their money several times. So what's typical? Well, how about some maths. A gambler who's got 100 BTC to lose, will on average he'd be able to wager 10000 BTC (on the standard 1% house edge site) before running out of money. Which means investors are going to pay: 10000 * (0.5 + 0.2) in commissions, or 70 BTC. The remainder goes into the bankroll (30 BTC). Which is exactly what I said. Investors are paying an amortized 70% commission. The max bet is still set to the previous settings of the original Moneypot settings, 1x Kelly. If the house edge is 1%, you can only win a maximum of 1% of the bankroll. If the house edge is 5%, you can win up to 5% of the house bankroll.
*sigh*. As I suspected. You're off by over a factor of three, which means you're putting investors into negative expected bankroll growth territory. I would've thought understanding half a page of basic casino maths would be a prerequisite to running a casino. Although to be fair, you're not the first investment site owner to struggle with it. It's pretty sad, but I think Dooglus is the only guy running a site that really gets it. |
|
|
|
Yes, it is 20% of the house edge. Not 0.2%. My mistake. I meant to say to calculate it, you would take House Edge * 0.2% * wager to see the house commission.
The documentation on MP needs to be fixed too then. It's not a particularly small mistake, considering it's the difference between investors paying amortized 50.02% commissions vs 70%. Also, would you be able to answer my earlier question about the max bet, which is probably the most important.  |
|
|
|
Did that bet not violate the kelly?
Nope, it was an aggressive series of kelly-compliant bets. Or think about it this way, there's no point having a 1000 BTC bankroll if you're not risking 1000 BTC. If the biggest loss was 100 BTC, you'd only need 100 BTC bankroll. |
|
|
|
Everybody keeps saying that there are risks involved with investing, but I'm not sure I understand where they're coming from. Is it that they might lost money in the short run, or because they could leave with your money?
Both. And short run can also be long run. There's no guarantee that a site will have continued volume, or after the bankroll drops you might get diluted out. Here should be a sobering graph of some of the risks involved:  ...and there's sites that have done a lot worse (like been hacked). If you're investing more than a few percent of your money, you're just gambling yourself =) |
|
|
|
If the share is fair, you should get 10% of whatever people lose in total.
Something for Dogedigital to answer, but as I understand it, with 10% of the bankroll you will get 3% (amortized) of what people lose in total (the rest of apps and MP) |
|
|
|
Commission is the house fee taken to pay for servers, hosting, work and services to maintain the website.
It is 0.2% of the house edge.
For example, if you're playing at BetterBets at a 1% House Edge and bet 1 Bitcoin: 1 Bitcoin (bet) * .01 (house edge) * 0.2 (commission) = 0.002 Bitcoin (Total Comission taken).
Ugh, is it 0.2% (what the FAQ and your post says) or 20%? -- I'm going to assume it's 20%, because 0.2% is negligible and doesn't make much sense. Also the total amount of commissions should include what is paid to apps, no? So someone bets 1 BTC @ 1 house edge. The apps get 0.005 BTC, MP gets 0.002 .. and investors have all the risk? Also just to make sure it's clear for investors, let's assume the the bankroll is 100 BTC and someone wants to bet on a binary 1% house edge bet. What exactly is the max profit of that bet going to be? Because if investors are going to take 100% of the risk for 30% of the EV, it's pretty damn important they are risking under a kelly. |
|
|
|
Which site is better to invest in?
Is almost certainly the wrong question to be asking. The kelly explains why: https://en.wikipedia.org/wiki/Kelly_criterion . If that's not making sense, I'd strongly suggest not investing in a casino. But the short of it, is even if it's a good investment you should only be putting a tiny fraction of your money in (to achieve +expected growth), so it's rather irrelevant which is better. The relevant question is along the lines of is X worth putting money in, and if so, the % of your total bankroll should be so small it will have almost no bearing on your ability to invest in other things worth investing in |
|
|
|
Something doesn't quote add up: Site Bankroll: 389,800,231.98 bits
Wagered against bankroll: 14,589,464,076.97 bits
Investor Profit: 87,654,258.75 bits
Your stake: < 0.01 %
Your amount: 100.86 bits
Today: Site Bankroll: 389,934,103.16 bits Wagered against bankroll: 14,772,889,421.86 bits Investor Profit: 99,064,021.05 bits Your stake: < 0.01 % Your amount: 103.90 bits So collectively the gamblers lost 99,064,021.05-87,654,258.75=11,409,762.30 bits, which is a 13.0% increase. The site bankroll barely budged, making only a 0.03% increase, meaning the investors collectively divested 11,275,891.12 bits. Ergo, gamblers lost a ton of money, the other investors divested a lot of money, and I only get a 3% increase? Note: I did not divest any of the 100.86 bits that were already in there. So either moneypot is faulty or the formulas are more complicated than that (unless I'm wrong)... It's probably the last one Let's illustrative purposes lets the bankroll was 100 BTC, and you own 1% of it (1 BTC), and the investor profit has been 0.1 BTC. Now let's say a gambler comes, and loses 1 BTC. That means the investor profit has gone up from 0.1 BTC -> 1 BTC (1000%)! How much would you expect to make? Well you don't make 1000% of your money, that's for sure, because if so you would've made 10 BTC (that's more than the gambler lost!). But rather, your profit would be { your stake} * { increase in bankroll } or 0.01 BTC (You own 1% of the bankroll, so take 1% of the loses and gains). So back to the moneypot example, the change in investor profit is actually meaningless. What you care about is a gambler lost ~11.4 BTC on a ~390 BTC bankroll which means the bankroll should have increased by ~2.92%, and thus your stake should be worth the same amount more. Which matches your observation. BTW a very probable reason the amount divested almost exactly matches the amount lost, is because the person who divested the money is the same person who lost it (Disclaimer: I'm not an investor and haven't been following MP, I'm just going purely based on the info in your post) |
|
|
|
Sometimes I see the "site bankroll" fluctuate. Is that just a result of other investors investing or divesting and a combination of gamblers winning and losing?
When a gambler loses, your stake stays the same but the bankroll goes up (aka you make money). When a gambler wins, your stake stays the same but the bankroll goes down (aka you lose money). When someone invests, the bankroll goes up but your stake decreases (aka you don't win or lose). When someone divest, the bankroll goes down, but your stake increases (aka you don't win or lose). Also, how much of that belongs to moneypot and how much belongs to investors?
Good question, something for them to disclose Secondly, I assume "wagered against bankroll" is the total amount of money gambled, so it will always increase?
yup Finally, and this is where I am the most confused, what is the formula for "investor profit"? Also, is it proportional to "your amount", and does it change when invesstors invest or divest?
Investor profit is just how much gamblers have lost in total. |
|
|
|
Congratulation to endless_pain for an awesome finish to the night:  Not only working his way up from 8 BTC in the hole to 38 BTC ... but ending with a 15.1 BTC win! |
|
|
|
But if the seed is set. You can't change it. The roll has already been determined. The house knows it. It is a number which has already been determined. This does not fit the definition of random.
It seems the only "aspect" of the randomness is you can set your own seed. But they still know. And you're limited to changing it once every 10 minutes or so.
So the rolls are not random.
Correct. All provably fair schemes aren't random. Dilbert explains why:  So we use the next best thing, provably predetermined. The house **does** know all bets in advance, however it's not a problem as the house can't do anything about it. It just has to honor your bets and keep increasing the nonce by 1. If it deviates from that, they've detectably cheated. If the casino feigns down-time, you should be able to continue playing from where you left off etc. So the house knowing the games in advance isn't actually a problem. |
|
|
|
Is it possible to crack the hash site side/server side, perhaps not immediately but within time being they know all three parts up front? SHA512 is possible to brute force/crack, even long random passwords can be cracked rather quickly if you have a huge gpu setup to do it. Its been done. (Theres articles about the time it takes to bruteforce sha512 all over the net, and with a huge gpu farm, its possible to crack 1000s of passwords/outcomes in minutes).
Brute forcing a password doesn't exploit a weakness in sha256, it exploits a weakness in the password (simply there's not much entropy in the typical password, so you can try them all). For a casino to (undetecably) cheat on the other hand, they would need to find two (valid) inputs that hash to the same thing. This is known as a collision attack, something that sha256 is believed to be secure at. If you for instance believe that PrimeDice might have cutting edge theoretical crypto to do this, you could protect yourself by "pre-rolling". A collision attack would be completely thwarted by making a couple of bets, to force them to use one of their two preimages. Now for the casino to cheat, it's going to need to be able to pull off arbitrary preimage attacks, but at that point bitcoin is going to be broken anyway. And even so, you could still defeat it by prerolling a couple dozen or so rolls. tldr; It would require breaking crypto believed by everyone to be secure for a casino to cheat. As a casino owner, I doubt it even makes financial sense to cheat even if you can (after all, a casino is in the business of giving you variance for EV). |
|
|
|
Good time to start playing bustabit:  73 BTC won thanks to a bunch of high multipliers last night (and some beginners luck). Quite a shock to wake up and see the house down 125 btc compared to when I went to bed |
|
|
|
Back online, sorry about that! |
|
|
|
Looks like the game server exploded  I should have a new server built and ready in 30-60m. Sorry for the inconvenience. In the mean time, all deposits and withdrawals are functional |
|
|
|
|
Show Posts
