Note: the percentage on my site is comparing the expected profit with actual profits. It has nothing to do with any timeframe. So it is meant to say that the players have been more unlucky and investors more lucky than expected (which is still the same conclusion as you "investors done well" ) Also, I don't actually calculate commission in there, so it should probably be around 200-300% (eg investors have been 2x - 3x more lucky than they should.)

Overall that statistic isn't really important IMO. It just show how much variance there can be.

Try all lowercase letters.

Yeh, with more decks it becomes a bit more difficult. For 8 decks I guess you will get something like this: card numbers will be 0-415, use 3 characters instead from final seed (= 0-65535), convert to number, then use the number if the number is 0-65311 (so skip 65312 and higher), with that number do a modulo of 416 so you will get a result of 0-415 = 1 card out of 8 decks. Again, if you got number already previously - skip it.

Video poker is like a slot reel where each reel is 52 cards, right? So I think that would be the same, but doesn't need the "check if double" thing because each reel can have exact same number than previous. After someone uses the "hold" option, it will use the 6th number, 7th, etc.

Normal slot reel is even easier, because it can just use 1 character instead of 2/3 (when you don't have more than 16 options per reel.) Crypto-Games also uses this for their slot games.





I wonder what OP, RHavar and others that participated in this thread think about above options rather than Fisher–Yates method. It seems easier to me because anyone can verify it with simple SHA256/512 and Hex-to-Decimal online tools - rather than running a Fisher–Yates and modulo-seed-loop-rng script.

4-10 was example. In reality there are 64 sets of 2 characters in a SHA512 hash. Some of those 64 numbers will be out of reach (208-255) or double numbers, but it should be still enough. One could do the math of odds of running out of numbers in biggest blackjack game (depending on "split rules" too), but I am bit lazy to do that. I do assume its enough even with all the splits.

If you have a multi-player blackjack game, it would probably not be enough. It could still easily just add a nonce to further loop or just use the Fisher–Yates shuffle. Still IMO the simple hash without Fisher–Yates shuffle is easier / more clear. So I would only use Fisher–Yates shuffle if it's really needed (=need lots of cards selected.)

Unique card: I mean, that Crypto-Games has 4 deck of cards, but actually its just numbers from 0-207, see: https://www.crypto-games.net/blackjackcards.html We will just pick numbers between 0-207 and each number represents a card. So any card, like a King of Hearts, appears 4x in total (numbers: 23,75,127,180.) When I say "unique" I mean that the specific number 0-207, like "75" cannot be repeated, since normally a specific card cannot be repeated either. However, the king of hearts can appear up to 4 times with those 4 numbers (23,75,127,180.)



edit: I guess this method is mostly nice up to 4 decks, since 2 hexadecimal characters are between 0-255. If you want more decks, you would probably need some ugly modulo thing with 3 hex characters. So probably better just Fisher–Yates shuffle with "seed RNG" for more decks. I am not sure if there is a site already that implemented that?

For Blackjack you can use the method I just described (which is used by Crypto-Games.) Blackjack only needs like 4 - 10 cards in normal game (even though 4 decks are used.) I like the way Crypto-Games does it, because it's very easy/understandable.

That method keeps looping looking for unique cards within those 4 decks. That's fine for like 10 cards, but if you would need like 40 cards, that probably gets pretty bad in performance (because it keeps looping for unique cards while not really discarding/removing them) and additionally that "final hash" would not provide enough characters.

So if you need to select more cards for other games:

It's not about the seeds used, but the MT RNG just has that seed input limitation. MT19937 is normally used and has a 32 bits limit, there is MT19937-64 which would be better with 64 bits but still not enough (although less feasible to brute-force results.)



Anyway for a lot of games, like blackjack, very easy solution can be possible. For example Crypto-Games.net does this:

1. Gets "final hash" (with "per roll" method but could be with "normal dice nonce" method.)
2. Has 4 decks so 208 cards: https://www.crypto-games.net/blackjackcards.html
3. Gets 2 characters from "final hash" and convert to decimal (= value of 0-255)
4. Use that as card (if within 0-207 range and not double.)
5. Repeat with next 2 characters till you have all the cards you need.

So that is easy/good imo?

For roulette it's even easier.. just loop those 2 characters from hash till its 0-36. (Crypto-Games does this too.) No need for "initial" stuff.



If you need more/all cards, I guess Fisher–Yates shuffle with proper simple RNG (not MT) is faster than that long loop though :p

Yeh, bit strange Alex puts the new site online without a message here

But yes, this is the new site. You will need to register a new account with e-mail. After that you can go to settings and "Import Old Account" to import the old investment/etc. Note: username for importing must be lowercase (for me.)

Just imported account with x BTC so seems to work fine.

Username: NLNico

Thnx.

Not sure why I am a "cheap idiot", you are the one trying to get more faucet money.



The site(s) that you use restrict the bonus/faucet per IP. You are trying to bypass that. To me that is abusing their service since they clearly don't allow it.

If you think the site should allow it because you use it "legitimately", you should contact the site to ask for it. Rather than trying to bypass their restrictions.

Or maybe just don't abuse services that give you free money?

1. Ask for affiliate account
2. List them on your site
3. ??
4. PROFIT !

Added a verifier for Crypto-Games.net: https://dicesites.com/crypto-games/verifier (all games)

They also added a link on the bet details of each bet to directly go to my site with all the seeds - so verifying your bets is really easy. (Although for optimal provably fair usage, you should copy your serverseed hash and clientseed before each bet is made.)



Ps, still working on new site.. guess I won't make it this month (:

You realize BetKing has a crowd-funded bankroll, right? So if the owner would fake a 1012 bitcoin loss, he would actually have to pay 1012 bitcoins to the investors. In reality there will be only a few thread bumps on this forum, so this "promotion" isn't worth 1012 bitcoin. That would be pretty expensive marketing.

Yeh... this won't work.

First of all: < 49 = Result lower than 49 wins. > 51 = Number higher than 51 wins. You seem to switch them.

Secondly if I want to get a result lower than 49, obviously I would just select a number 0-4. Why would I ever chose 5-9? And because of that I got a huge advantage over the casino.

However, the casino can still easily cheat too. They can just generate a number 0-4 always when I click "over 51". How do you know that this 0-4 number was fairly generated? You cannot know that and you would lose literally every bet. So that is not provably fair.








The basics of provably fair aren't that difficult. Simple example:

Casino generates number + protects it with secret and makes a hash for it. You might find the hash confusing.. but the idea is that the casino gives "their result" in advance.. so they cannot modify it anymore. But it is hashed otherwise you know the end result in advance and could cheat.


Hash that you get in advance:

Now you make your bet and send in your result, let's say 76 and "over 51". Casino reveals their number+secret. You can check that the hash was really from that number+secret with any online SHA256 generator. So you can prove that the casino didn't change their number. (This is missing from your idea.)

For the number generation, they would just add up the numbers and if its over 100 remove 100 from it, so (76+43) = 119 - 100 = end result 19. Since you bet "over 51" you lose.

With this you proved: 1) the casino calculated the result with THEIR and YOUR input so neither party could know the end result in advance 2) the casino didn't change THEIR input since you had the hash BEFORE you gave YOUR input.





This is a simplified idea. The average "per roll implementation" and even the "nonce implementation" are based on the same idea with just a bit different result generation. IMO the nonce implementation is the best for individual betting because you don't need to verify the hash literally every bet. You can see a more detailed but still practical explanation about that here: https://dicesites.com/provably-fair

Yeh, only like 7 - 15 players (again.. fine for me.. I can actually play Stars MTTs if I want but sometimes I like small good valued MTTs more.)



Agree with rest though. Even just some SWC dude posting the daily good valued MTTs in here would help to get some more traffic. (Not even talking about development/android/etc.)

For me personally I can play 3-6 MTTs with 20+ BI every day within a 5h time spawn (8PM - 1AM EST), which are all good valued (overlay/rake/etc).. so for me personally that is good. But it depends on your timezone and obviously I could play a lot more on Stars.

I agree in general site is pretty dead though, but still some fun for specific games/times.

Depends what you call "a lot" probably

Personally I only like NLHE MTTs and SWC still has some pretty decent GTDs for that every night (overlay, good value, fun structures, cool players, etc.) Didn't play for months though, just happened to play few yesterday

That is 100% your point. You say - direct quote: "You think that this seed that you choose is the one you are playing with, but no..".

Obviously that is NOT the seed you are playing with, because you didn't click the "Change seed" button.

If you did click that button, you would be playing with that client seed.

Pretty sure thats just the message "hot wallet is empty" message.. so I wouldn't be too worried about it. But hope it will be fixed for you soon

Lol. Obviously if you change the client seed, you still need to click the "Change seed" button. OP are you blind or how did you miss that button?