Eggy,
You are one of the best coders in SL and I'd like to see the artilce you refer to about real life gambling machines being on a schedule. Your post is the only reason I am replying to this thread and I most likely will never reply to another "casino/gambling" thread again.
I hate drama and would much rather stay out of it. I don't have the time for it and I believe people will figure out for themselves where they want to be entertained without a lot of help from other people.
Mostly I am interested in the article you refer to, not that I don't believe it exists but mostly because I'd like to read it for myself in order to improve on how I code my games to be the most compliant with real life games. Although SL is not real life I would like my SL experience I offer to be as true to the real as possible. If this article could glean how I could better do things then I surely would like to know exactly how it is so I could decide to or not to implement that.
At the same time (and mostly) I am going to defend random numbers.
Everything I have ever researched show that the only fair way is the random way. Of course I'm the first to admit that even at 99.99% payback if you continue to play in the end all your money is gone. But non-rigged == random IMO.
How Slots Work:
Refer to this article:
http://money.howstuffworks.com/slot-machine.htmOr skip to page three of that article:
http://money.howstuffworks.com/slot-machine3.htmA random number generator at the heart of the computer ensures that
each pull has an equal shot at hitting the jackpot.Here is some more ... from another article:
http://www.casino.com/videopoker/article.asp?id=43314th paragraph down:
Nevada gaming regulations require that all machines be random, but they also specify that each unseen card in video poker must have an equal probability of appearing while there is no such requirement for symbols on a reel slot.
You see with random you CAN guarantee in the LONG run a % payback and at the same time guarantee the same probability of hitting a "jackpot" on every spin/deal. That is the beauty of random numbers. Yes, the same machine "could" hit a jackpot twice in a row, but it's not very probable, yes the same machine "could" live it's whole life and never hit, but that's not very probable either. You see, over MANY spins/deals the numbers "approach" that percent. I don't see ANY other way to guarantee a % payback and at the same time guarantee that every spin/deal has the same chance as every other. The only way I know of to do this is with random numbers.
For instance you could flip a coin 100 times and have it come up heads every time, it's not impossible, it's not very probubal, but it is POSSIBLE. If you flip a coin a million times you might even have some runs of 100 heads in a row but over that million flips you will have close to 500,000 heads and 500,000 tails ... and every single time you flip the coin, no matter what came before or what will come after that flip still has a 50% chance of coming up heads. This is random numbers working IN THE LONG RUN and IN THE SHORT RUN. And this is the beauty of random numbers.
My slots are pure random ranging from the approx 80% to 95+% payback with no "schedule" ... every pull has the same odds as the last. Every pull is random. By counting the positions on the reels and caculating how often each pattern would show up when randomized and adding up the money in versus the money out anyone can calculate the the actual odds on a slot machine (if it has no weighted symbols). Even with weighted symbols which manufatures use to offer huge pots yet still set the odds yet still not schedule payouts (yes they use random) the % payback IN THE LONG run can be calculated. No not guaranteed but pretty damn close.
When I designed the slots I figured the odds on them or what they should be based purely on thier pay table, number of postions on the real, statistics on how often permutations should happen with a good random number generator etc. After they were complete I did in fact run simulations to run tens of thousand of spins and collect data. I ran many such simulations. Guess what, the simulations verified 100% that the odds I figured were actual odds realized in the machine.
I did not do this to verify I know how to set amounts to permutations to make a particular % payback machine. I did this basically to verify the random number generator used in SL is a good one for the purposes I was using it for and also to verify I didn't have any blatent mistakes in my logic (which is quite simple by the way).
All my poker games are pure random. The deck is randomized (shuffled), then cut (from a random spot), then every single card is dealt from the top EVERY SINGLE TIME. So yes you could hit a royal flush two times in a row. I have yet to see it happen, but it CAN happen. And yes, every single hand has the same odds of hitting as the hand before and the hand next.
My Jacks Or Better has a 9/6 pay table which is 99.5% payback if "PLAYED CORRECTLY" from the randomized hands. Actual data from actual casinos shows a 2% advantage to the casino over the 99.5% perfect play because humans are not perfect and or many do not know "how" to play a particular game.
My Joker Poker is a FULL PAYBACK machine if played perfectly, this takes into account the frequency at which natural royal flushes "should" occour in the world of random numbers, and playing it using an optimized strategy. Once again real data from real casinos shows the house to actually have about a 2% advantage here because of humane error or not knowing the best strategy.
I did NOT calculate these odds, I found them throughout the internet from reputable mathematitions whom are themselves interested in calculating the odds in such things as these card games. I did not save URLs to all of the articles and research I did ... anyone with MANY hours on thier hands can go do their own searching and verify they will find similar data. Calculating these types of things is not easy, I won't even attempt it. But given a particular pay table and also given an optimal strategy for a game (these have been figured out) and also considering using a RANDOM sequence of cards being shuffled/drawn then the odds CAN be calculate and have for many games out there. All of these DO assume using RANDOM numbers. Those odds which have ALREADY been figured out by other people are the odds I trust and believe in and base my games on.
So no, your quote "A purely random game would have really horrible odds." is absolutely wrong. A purely random game's odds CAN be calculated both for slots and for card games. Not only that it is the only "FAIR" way to ensure that both the odds are KNOWN AND
each pull/deal has an equal shot at hitting the jackpot. ... each and every one.
That is each and every single pull or deal ... including the one just before or just after an actual jackpot. Every single one is EQUAL.
That is not to say a game using random numbers could have really horrible odds. It could! It all depends on how it's designed. It could also have really great odds. Random does NOT == horrible odds. But you CAN use RANDOM numbers and still KNOW the odds.
I have paid out 250,000 L jackpots (ok only one of those so far but I'm sure there are more to come). I have paid out MANY 100,000 + Jackpots ... yesterday I paid out 150,000 L in large jackpots and I don't know how many small jackpots. I see jackpots happen all the time ... I've seen several in a couple minutes then none for a couple days. I have NO IDEA when they will happen. They are RANDOM. One thing I do know is they WILL happen, and I don't have to freaking schedule them ... they just simply WILL all on thier own. And odd as it may sound, BECAUSE of RANDOM numbers I know more or less the frequency OVER THE LONG RUN that they will happen because that's the beauty of random numbers.
I'm sure people like Tony Tigereye has paid out even larger jackpots. I can't play my own games because that is no fun at all ... it's boring playing against yourself. There is no entertainment value. When I play I play Tony's. My game of choice is Jacks Or Better, even though Joker Poker has better odds. I don't know how his games are set up but I'm willing to bet they are RANDOM. If they're fair they are. If they aren't well they are still fun when I play them. And that really is the way I look at it. It should be FUN. When I go to the movies I know I'll sped X to see the movies. When I go to Tony's I know I will spend X to have FUN. I never go thinking I'll "WIN" I go thinking I will be ENTERTAINED. If I ever hit a Royal over there then that is HELLA ENTERTAINMENT. If I don't then I paid for and recieved HELLA ENTERTAINMENT anyway because I knew there was a CHANCE it could happen.
There is nothing more exciting then knowing that chance ACTUALLY EXISTS EACH AND EVERY SINGLE TIME I TAKE IT.
Now that is fair. Or at the very least == not rigged. I know the house has an advantage as well as anyone else. That is what I choose to pay for my entertainment, that house edge. That is what I'm paying. When I go I tell myself I'll use X dollars, if I double it I'll quit, if I spend it I'll quit. I don't know if that's a good strategy or not. That's just what I do.
Random is a beautiful thing. I trust it complete.
I will repeat one of the very first thing I stated in this post:
Of course I'm the first to admit that even at 99.99% payback if you continue to play in the end all your money is gone. But non-rigged == random.
Steelwolf may be wrong about anyone ever going to the Lindens ... I don't know of a case of that actually happening. I have offered it, maybe it's been done and I was un aware. I don't know. I do know I have offered at any time for my scripts to be inspected by any Linden for non-rigged == random with no pre-set outcome and that offer as long as the code itself is not taken and given away remains. I have never approached a Linden myself I would rather bug them about why mouse clicks aren't working or some other bug. I also think they would rather spend thier time fixing bugs like mouse clicks then being baby sitters.
I have been in the past accused of "rigging" when someone hit a bad streak and offered them the same thing "Find a Linden that will look, feel free". If they ever did or not, I really don't know and I really don't care and I really doubt that Lindens have time for any of that. If they do I welcome it. I know what my soup is brewed of and it's pure random. Look all ya like.
The actual cases that come to mind those same people have been back and been on "winning" streaks and apologized. I don't mind the criticism and I don't cherish the apologies, things are what they are, I know what they are and that's good enough for me.
With random numbers I don't know when I will lose or win, I don't know when you will lose or win. All I do know is there will be a whole lot of both going on ... it will be RANDOM ... and depending on the pay table there will be a % in the long run that the house ends up with. Whether I am the house or I am playing the house. It's that simple.
It really should be for ENTERTAINMENT purposes only.
The house is a movie theater, come watch the movie ... set your own ticket price ... watch as much and as little as you like.
I don't know what the odds are on all my card games. I do know I have taken the pay table of REAL LIFE games when I could find them. I shuffle the deck, I cut the deck, and I deal every single card from the top after that ... fall where they will.
I believe Hang Ten is near the worst around 93% payback. I know Jacks Or Better is 99.5%, Joker Poker is full payback, Duces is close to full payback but I haven't found an article or break down for my exact pay table. Kings Of the Deck I have no idea because I can't find full stats on it nor somone who has posted the odds nor even exactly how the pay table and the progression on it should work but the pay table I have is based on the only one I COULD find so I suppose it isn't horrible. I think I toss one or two percent back in for the progression or there abouts. If/When someone does hit five kings of clubs they will be very very very happy and that has the same odds of happening EVERY TIME. I don't know what all the "best" strategies for playing each game is but many of them are posted on the internet.
My slots I do know and they are better then most sold or in game, the Christmas slot is my most gernerouse one so far. I forget exactly as I'm sitting at work right now but it's something like 93% payback or better. No this is no where near 99% as some in Vegas but it's also no where near 75% as MANY in game. NONE of my slots are as bad as 75% unless I've made a stupid mistake. None of my slots so fare have weighted positions although some do have teasers ... which is when a losing round comes up I'll tease by showing a close miss to a pot BUT I don't EVER change a winning to a losing or a losing to a winning spin nor change the odds EVER ... the tease is for excitement only and is frequently done with slots.
I may make some with weighted reels eventually to offer huge pots but the % payback will always remain generous. I'm also thinking of making very very very generous payback even better then the Christmas slot because why not? In the long run even 1% is a gain to the house.
I give a LOT back to the community. One event costs close to 5K at my club of cash given back. I NEVER use money/camping chairs. I NEVER have pure freeplay clicking for an hour. We mix it up and jump around doing diffent things and SOCIALIZE, have fun etc...
I never come out slugging trying to hit top 20 or get incentive awards. I have events to HAVE FUN, give back, and promote my place. Although I have hit top 20 and gotten incentives, my events have never been for those purposes.
I most likely will never post to threads like this again. I'm more interested in the scripting threads and I'm more interested in creating new content.
That's my RANDOM cents on this subject. And my reasoning why I believe that Random is actually THE WAY to do it right.
Please show me another way that's just as much fun?
--Chilly
Here are some other interesting links:
Old but good from SL forums:
/120/85/34811/1.htmlSome real life stats:
http://www.casinospage.com/casino_returns.htmJacks Or Better Strategy:
http://www.casino.com/videopoker/article.asp?id=2584Duces Strategy:
http://www.casino.com/videopoker/article.asp?id=2656Another Look at slots and pots:
http://pilarski.casinocitytimes.com/articles/5990.htmlSomeone Actually Doing some casino math: OUCH HEADACHE!
http://gaming.unlv.edu/research/subject/casino_math.htmlOne last thing to note, in my opinion making open source is not so much a fear that other people will see how to make slots or card games. They are both EASY. Slots ... have some positions, grab three random numbers, spin to them and count the score.
Cards are, shuffle, cut, deal, allow holding, deal, count score.
SIMPLE
Open source to me is giving away my optimization techniques for how to make MANY games run at once and quickly using the very least amount of sim resources by techniques having to do with texture optimizations, script optimizations, score counting optimizations, and communications optimizations. Those are the things I hold dear and sweat and lost sleep over implementing. The card shuffing or reel spinning. Anyone can do that! You want to see those it's simple, it's something like this:
LIST = [0,1,2, ..., 52]
llRandomizeList() <<< or what ever the call is
CutDeckPosition = llFrand( between 0 and 51 )
Cut the deck.
deal
deal
deal
deal
deal
hold?
hold?
hold?
hold?
hold?
deal?
deal?
deal?
deal?
deal?
count score() <--- top secret super fast way to correctly count the hand score! Figure your own way out!
DONE!
-----------------------------------------
here is slots:
llfrand()
llfrand()
llfrand()
SPIN SPIN SPIN
count!
that's it .... jesus!
Forgive my script function torture/misspelling/incorrectness I don't have the wiki open but you get the idea...
cheers