Quote from: dyask on January 16, 2015, 12:14:32 PM

-snip-
Not really ... let's just say it is possible to write a random number generator that gives clues to the next roll. The clues could be coupled to the server seed and another function. If you know what the next roll is like, you have a lot of options for cheating. However that is just one approach, there are others and I'm sure many that I've never thought of.

Knowing the next number would give an advantage to the player not the operator. I dont see how this works, besides against the investors, which is a completly different topic and we are allready way off the OP (are ponzis fair).

Quote from: dyask on January 16, 2015, 12:14:32 PM

I'm not going to give you any details on how to make a random number generator that cheats. I'll just say it is extremely difficult to make a random number generator that is truly random. Additionally if you can run it and get the same sequence with the same seed, it isn't really random.

Truly random is not the goal, as doog once said:

Quote from: dooglus on October 12, 2014, 04:40:19 PM

-snip-
I have tried many times to explain how truly random numbers can't be *provably* fair, and so we use secret seeds, public hashes and nonces to generate a sequence of pseudo-random rolls which are provably predetermined and which allow us to prove that we didn't change them depending on the user's bets or deliberately pick "bad" sequences.
-snip-

Quote from: dyask on January 16, 2015, 12:14:32 PM

If you have the source and you can really understand the source then the hash keys might mean something. Believe me, that is a tall order.

The source is mostly very simple, e.g.:

Code:

function getLuckyNumber($serverSeed, $clientSeed, $incrementalNonce) {
   $seed = $serverSeed . '-' . $clientSeed . '-' . $incrementalNonce;
   do {
   $seed = sha1($seed);
   $lucky = hexdec(substr($seed,0,8));
   } while ($lucky > 4294960000);
   return ($lucky % 10000) / 100;
}

Quote from: dyask on January 16, 2015, 12:14:32 PM

I wrote my own sha256 routine using AVX instructions along with parallel paths in the processor and understanding the simple hash math well enough to do that isn't something that can be done in a few hours. There is a great deal of open source code that has had major bugs found only have years of a great many people going through the code. If you don't believe me, read some pull requests in git repositories for your favorite Linux.

I have no idea what this has to do with the topic at hand.

Quote from: dyask on January 16, 2015, 12:14:32 PM

The point is that these proofs are only valid if the site is honest in the first place.

I have yet to see an argument why this would be so.

Quote from: cloudbets on January 16, 2015, 08:19:51 AM

You can check this site, the game called "the list" is similar to a ponzi

www.bitnety.com

Is interesting because rules and data is public and you can withdraw your coins whenever you want, without waiting time. As far I see, there is also another intersting feature: when players withdraws money also gives some of this to players still inside the "list"

Its a bit difficult to understand at first sight, but i tried the test mode and works!

I dont know, they got provably fair wrong and wrote "probable fair instead. Ill take a deeper look when I have more time.