LG John

Ok, leading on from the probability experts thread, the casino gambling thread and my poker probability thread I put this question to you:

How probable is probability?

I'm being serious. If we know something is mathematically 70% likely to happen, will it happen 70% of the time always if played out over time.

Ok example time: Lets say you (the reader) is the 'house' and I am the 'punter' You have divised a computer gambling game that has the odds mathematically set 1% in your favour. I have devised a 'computer programme' that plays the game for me perfectly according to the maths so that theoretcially I'm 1% disadvantaged but no more, no less. Now. We agree to put up £1m each but in my case its just a drop in my vast financial ocean but to you its EVERYTHING you own. The bet per hand/turn/spin is just £1 and my bot and your system can play at 100,000,000 hands/turns/spins a second. We play winner takes all - would you play me under those conditions? Or would you fear that somehow the mathematics would go wrong over all those billions and billions of hands/etc required to wipe you out?

Katana

Hey, maybe you can teach me how to play Poker Kenny. Lets start of with a warm up game to teach me how to play followed by some intense gambling. Call it baptism of fire.

LG John

LOL, I'd love to teach you to play poker - how many years do you have

ProperCharlie

yes, otherwise it wouldn't have a probability of 0.7

maths is a way of describing reality, not the other way around.

LG John

That's pretty profound......I think

ProperCharlie

all i meant SB, is if something doesn't behave in the way that you have described, then you haven't described it right. if you say something has a binary outcome, and it turns out to have three possible outcomes, then it doesn't have a binary outcome after all.

you dig?

V45DSM

The more hands you play the closer it will end up at 51/49 in my favor. So, if played over billions of hands, I would bet my life on it. On average I would win £1 for every 100 bets, so, 1 second of play (100 Million hands) would see me relieving you of your 1 Million Pounds Sterlin.

[Alan Partridge]

Back of the net!

[/Alan Partidge]

[Edited by V45DSM - 1/8/2004 6:18:25 PM]

LG John

Yeah I see what you are saying. So given that its mathematical probability (calculated perfectly beyond all doubt) we are looking at here it simply cannot go wrong save for forces we don't/can't understand. God, alien interference, etc.

Katana

I've got like next weekend or something.

LG John

LOL, I can [i]maybe teach you what beats what and the betting structure in that time

ProperCharlie

if it doesn't turn out in the way that the maths suggests, then there must be an additional factor at work that hasn't been taken into consideration when the calculations were done. some people might call this God, others might call it a magnet under the roulette wheel.

V45DSM

God died years ago and the odds of another world that's able to sustain life, being close enough and within our timeframe, with inhabitants intelligent enough to get here, where ever ‘here’ is, are very small.

[Edited by V45DSM - 1/8/2004 6:12:07 PM]

LG John

What are you suggesting

imlach

I'm being serious. If we know something is mathematically 70% likely to happen, will it happen 70% of the time always if played out over time.

Hehe, you've surpassed yourself now!

As someone said, it's easier to think of probability if we think of binary probability. ie, heads/tails. Unless there was another factor (ie, one side of the coin being more weighty than the other), then yes, probability of getting heads at time=infinity would be 0.5, and probability of tails at time=infinity would be 0.5.

Apply that principle to ALL probability, and you have your answer.

QED.

eClaire

What length of time? Infinity?

Fast_Blue_Scooby

Saxo in your case, if you have calculated for there to be a 1% edge and the maths is correct, then it will always be 1%. And yes over a long period of time, the results would get closer and closer to this 1%.

Now in my game, horses and sports trading, I can only 'guess' at what the percentage chances of each horse winning, events occuring etc.

However by recording results over a long period of time, and so far this involves almost a thousand bets so far, I have found to have an edge of 2%. This is proved by carrying out maths on previous results but I do have to rely on this being correct for future events.

So far this year, I have a 2.15% profit on turnover relying on my figures from last year. I have already had a few hundred bets this year and already my system is working out fine.
I do bet thousands a day and a 2% target on turnover is bringing in a tidy profit each day.

So if you have proved by maths that the system has a 1% edge over you, then there is no way you can beat the system. Over time, it will slowly grind the money from you.

If however you can prove you have the edge and you have 100% faith in your system then go for it.....

SteveV-WRX

sorry to sink your boat V45DSM,

but the probability of the being alien civilisations at least as advanced as our own is expessed as drakes equation.

N = R*fpnEfLfifcL

IIRC in a galaxey our size the was likely to be 10's of 000's of world with intelligent life.

S

so Saxoboy, your going to be quids in!

ProperCharlie

Steve - don't think so.



The aliens aren't on Kenny's side, you know.

eClaire

Probability theory is the mathematical study of probability.

Makes interesting reading actually (if you are sad enough to actually read it!!!)

ajm

Yes - it will be exactly 70% with an infinite sample size!

ProperCharlie

you forgot about the aliens, ajm.

dnb

How were the game random numbers chosen? If it's by computerised random number generator, then they aren't random - they will be pseudo random.

Therefore it is possible to design an autoregressive algorithm to predict the next "random" number the computer will choose. This does work in practice, and very quickly cracks most pseudo random number generators. More than that, I can't & won't say.

Of course, this won't work if the numbers come from something like thermal noise, as this is properly random.

LG John

Whats the point of computerised random generators if they are not random? Why are they not random??

imlach

dnb - don't!! You're just teasing him

SaxoBoy will just get all excited trying to find the seed for the pesudo randon number generator - he should be out sowing seed instead at his age (24)!

I would hope Ladbroke etc would not be silly enough to use pseudo random....

imlach

SaxoBoy - do you want my computer science lecture notes

dnb

You'd be surprised who use psuedo random numbers, believeing them to be truly random.

It's not that they're that bad - until you use extreme measures to crack them.

BTW I'm only 25, so less of the age thing

imlach

Do a lot of work using PRBS (Pseudo Random Binary Sequences) at work, so know a bit about them

Used for testing telecomms networks....

imlach

SB - just for you, the science bit