WrongWay 0 Posted October 12, 2006 Share Posted October 12, 2006 Where's Actuary when we need him?I'm not the guy to explain this, but I know someone here can prove, using math, about 95% confidence level and statistical relevency. 19 is too small a sample size we know, but you can actually prove what the correct sample size is for any given set of parameters. I just don't know how to to it.... someone help....Actually, there is no "set number" of samples that would be required. You decide what you're testing, how many groups of samples you're going to have, and how many samples in each. For example, we decide were going to measure 50% suck outs using 20 groups with 20 hands per. So, we we randomly select hands and assign them to the groups. We count the number of times the suck out happend for each group. Then we average.So, let's say or groups are 10, 12, 8, 10, 16, 3, 8, 8, 10, 14, 15, 6, 7, 7, 14, 6, 19, 10, 12, 9. We add them up, then divide by 20 and get... 10.4We're getting 10.4 suckouts when we should only be getting 10? So, is that statistically significant? We don't know. We have to figure out the variance. We take the amount that each number was off from the average.... and average that:10 is -.4 from 10.4, 12 is 1.6 from 10.4, repeat for each of the 20 samples. -.4 + 1.6 -2.4 -.4 + 15.6 -7.4 - 2.4 -2.4 -.4 + 3.6 + 14.6 - 4.4 -3.4 -3.4 + 3.6 - 4.4 + 8.6 - .4 + 1.6 - 1.4 = 16 16/20 = .8.Now, take the square root of that to get the standard deviation... which in this case rounds to .9Now, 1 standard deviation equates to 70% condifence.... So, when we should expect 10 suckouts, using this bogus sample data, I'm 70% confident thant the actual number is 10.4 +/- .9.So no, this is not statistically significant.Or, if we want 95% confidence, then we have to go to two standard deviations... using this bogus data, I'm 95% confident the answer is in the range of 10.4 +/- 1.8Now, if you're not happy with this range... well, than you need to make the samples larger so that the averages get closer together, and you need more groups so that the average varience gets smaller.Data points determine necessary sample size, not the other way around. Link to post Share on other sites
omahahilo 0 Posted October 12, 2006 Author Share Posted October 12, 2006 I tell you what. This has been fun. But I will give you a serious challange.If you can prove that AA can't be beaten 100,000 times in a row and that its not possible that you can get your money in with the best of it and lose more than 50% of the time over the course of 100,000 hands. And its solid mathmatical proof that can be tested and proven. I will remove my account from this message board and I will never post again.You have my word.Matt Link to post Share on other sites
LongLiveYorke 38 Posted October 12, 2006 Share Posted October 12, 2006 If you can prove that AA can't be beaten 100,000 times in a row and that its not possible that you can get your money in with the best of it and lose more than 50% of the time over the course of 100,000 hands. And its solid mathmatical proof that can be tested and proven. I will remove my account from this message board and I will never post again.No one will prove this because it's so obvious that you can. You can lose ten billion times in a row. You can lose 5 Ghram's number in a row. Obviously. The chances of this happening, however, are quite low. Obviously.You can lose more than 50% of the time in a given sample even if your overall percentage of winning is greater than 50%. Why? Because it's random.What was your point again? Link to post Share on other sites
Money022 0 Posted October 12, 2006 Share Posted October 12, 2006 Next time you want to experiment with $37.50, just transfer it to me. You forgot about the lovely 33% entry fee for these over at Pacific, he actually should send you $50.The lesson here, Pacific rapes their SNG players. Link to post Share on other sites
omahahilo 0 Posted October 12, 2006 Author Share Posted October 12, 2006 No one will prove this because it's so obvious that you can. You can lose ten billion times in a row. You can lose 5 Ghram's number in a row. Obviously. The chances of this happening, however, are quite low. Obviously.You can lose more than 50% of the time in a given sample even if your overall percentage of winning is greater than 50%. Why? Because it's random.What was your point again?That's right. There may be a 0.000000000000000000000000000000000001% chance or likely even less. Again ... all I was looking for was discussion. Since I showed it may be a possibility I think that it would be very interesting to talk about. However, this is FCP and all anyone wants to do is **** with people and tell them how ****ing stupid they are so I guess that's what happens. I give up. Even if my theory is so unlikely its almost uncountable it is a possability. I know that, everyone knows that. And they just wanna be dicks. So **** them and everyone who thinks they know everything. Just keep being assholes online. You'll **** with somone like this at a live table someday and that'll be the end of how tough you are.See ya,Matt Link to post Share on other sites
WrongWay 0 Posted October 12, 2006 Share Posted October 12, 2006 We take the amount that each number was off from the average.... and average that:10 is -.4 from 10.4, 12 is 1.6 from 10.4, repeat for each of the 20 samples. -.4 + 1.6 -2.4 -.4 + 15.6 -7.4 - 2.4 -2.4 -.4 + 3.6 + 14.6 - 4.4 -3.4 -3.4 + 3.6 - 4.4 + 8.6 - .4 + 1.6 - 1.4 = 16 16/20 = .8.Wait... I screwed up... It is absolute value of the variance... That's what I got for going from memory from the stats class I took 18 years ago...Sorry, standard devistion is 2, so we're 95% confident the answer is 10.4 +/- 4. Using my bogus data set. Link to post Share on other sites
navybuttons 15 Posted October 12, 2006 Share Posted October 12, 2006 I really don't have a conclusion yet. Just possibilities. But let me tell you something. Read my first post. I simply suggested an idea. And suddenly I had the whole FCP army screaming at me about how wrong I was. And I may be but I may also be partially correct. I had someone tell me I was 100% wrong and I have def proved that is not the case. I came here looking for a discussion about an idea and found myself the butt of every pre-teen jackass who knows everything about poker. How the hell was I supposed to act. I was much calmer than I should have been. If people would have tried to discuss this insted of dismiss it, they might learn something, and they might not but they would have had a good opportunity to learn. Most of the people on this site think that they have all the answers and they don't. If they did they would be down at Bellagio making millions. Not on this site. I know I have alot to learn and that's why I would never tell someone they are completely wrong, before alot of consideration on a particular topic. If you told me that AA was the worst starting hand in Hold 'Em I would consider it before dismissing it if that was my final decision. Maybe a little less dismissing and a little more consideration would help improve everyones game????And for the reccord ... I have def proved there is a question. If you can't see that I'm sorry. Maybe you should read through my arguement about it ... but at this point even if you saw my point you wouldn't agree to it anyway. You do not know everything if you did you'd be making money. Not fighting with me online. Mind you're own G'damn business.By the way if there is a .00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000001% chance that AA could lose 10,000 hands in a row heads up then your "answer" is incorrect. Have a nice day.MattI do not know everything. i never claimed to. telling me to mind my own business in a thread on a public forum is laughable. you're right however, i don't make money. i am a losing player. i'm a gambling addict and have probably lost $50,000 online. I ran a little expirament.I played 25 $1.50 SIT N GOS on Pacific Poker.I chose this game because I knew the players would assist me.I pushed all in pre-flop with the first pair in the pocket I had 10-10 or higher and folded all the rest.Since I got called every time I found this very interesting.Out of 25 times I pushed all in I was the pre-flop favorite 19 times. I lost 16 of them to draw-outs.Discuss.Oh and I know I'm an idiot but look at these results.Matti am having problems discerning the question (maybe some question marks would help) but i don't need you to feel sorry for me, thanks anyway. AA could loose 10000 times in a row. i don't recall anyone saying it couldn't. people however were saying that 19 is too small a sample size to determine anything and you seemed to be under the impression that it was adequate.perfect sample: run it infinity times. adequate sample: run it 100,000 times.or just read /\ /\ /\ and \/ \/ \/ Link to post Share on other sites
WrongWay 0 Posted October 12, 2006 Share Posted October 12, 2006 There is plenty of silliness on both sides in this thread."Out of 25 times I pushed all in I was the pre-flop favorite 19 times. I lost 16 of them to draw-outs."While 16 out of 19 does seem extreme.. Let's toss out 1(to get a number divisable by 3) and divided the remaining into 3 groups of 6 and got 4/6, 5/6 6/6. Your average is 5, but your vairance is 1+0+1/3 = .666 Standard deviation is .8.So, given this limited sample size, we're 95% confident the answer is in the range 5 +/- 1.6. So there will be a suckout 3.4 to 6.6 out of every 6 hands.... technically significant, but I don't think many will accept that answer.As for pocket aces losing to any 2 random cards 50 times in a row.... Well, Aces are, what? 80% favorite over a random hand. So you'll lose 1 in 5. You'd lose twice in a row 1 in 25. 3 in a row, 1 in 125. 10 in a row approaches Lotto statistics of 1 in 10 million-ish.50 in a row rounds to something close to 1 in 89,000,000,000,000,000,000,000,000,000,000,000 (8.9x10^34)If the universe is 14 billion years old, that is only 441,504,000,000,000,000 seconds.....So, unless there have been thousands of quadrillions of poker hands dealt every second that the universe has existed, then it is very unlikely to have happened yet. Link to post Share on other sites
LongLiveYorke 38 Posted October 12, 2006 Share Posted October 12, 2006 50 in a row rounds to something close to 1 in 89,000,000,000,000,000,000,000,000,000,000,000 (8.9x10^34)If the universe is 14 billion years old, that is only 441,504,000,000,000,000 seconds.....So, unless there have been thousands of quadrillions of poker hands dealt every second that the universe has existed, then it is very unlikely to have happened yet.Tell that to Lee Jones. Link to post Share on other sites
antistuff 0 Posted October 12, 2006 Share Posted October 12, 2006 So, unless there have been thousands of quadrillions of poker hands dealt every second that the universe has existed, then it is very unlikely to have happened yet.now just for kicks, can you tell us how unlikely? lets say we were to run the univerise up to now a million times..... Link to post Share on other sites
LeighWdWTravel 0 Posted October 12, 2006 Share Posted October 12, 2006 Haven't any of you heard of the theoretical situation where if you have an infinite amount of monkeys typing randomly at a keyboard for an infinite amount of time that eventually Shakespeare's Hamlet would be typed out in its entirety.Vaguely recall it, but if you have those monkey's numbers, I'd be willing to hire themHourly labor though Link to post Share on other sites
WrongWay 0 Posted October 12, 2006 Share Posted October 12, 2006 now just for kicks, can you tell us how unlikely? lets say we were to run the univerise up to now a million times.....The answer is 42. Now I just need to write a computer program that can calculate the proper question for that answer. Link to post Share on other sites
brvheart 1,752 Posted October 12, 2006 Share Posted October 12, 2006 No one will prove this because it's so obvious that you can. You can lose ten billion times in a row. You can lose 5 Ghram's number in a row. Obviously. The chances of this happening, however, are quite low. Obviously.You can lose more than 50% of the time in a given sample even if your overall percentage of winning is greater than 50%. Why? Because it's random.What was your point again? exactlyThat's right. There may be a 0.000000000000000000000000000000000001% chance or likely even less. Again ... all I was looking for was discussion. Since I showed it may be a possibility I think that it would be very interesting to talk about. However, this is FCP and all anyone wants to do is **** with people and tell them how ****ing stupid they are so I guess that's what happens. I give up. Even if my theory is so unlikely its almost uncountable it is a possability. I know that, everyone knows that. And they just wanna be dicks. So **** them and everyone who thinks they know everything. Just keep being assholes online. You'll **** with somone like this at a live table someday and that'll be the end of how tough you are.See ya,Mattwrong, you were originally looking to prove that somehow online poker wasn't completely random and then after you realized you were totally wrong you changed your mind to make up something else you were doing. ban.Oh and I'm not saying that online poker is rigged. Its not. But it definately isan't as random as you may think.And obviously this was a NL HE tournament test.And also, I am not discussing player skill levels here. Just the randomness of the shuffle and results.Mattyes it is. Link to post Share on other sites
LeighWdWTravel 0 Posted October 12, 2006 Share Posted October 12, 2006 I was waiting for someone with a brain to come in here and help me, thanks Yorke.How old are you? Stop acting like a little kid. You came in here for answers and you got them.Read Yorke for a long time....before I decided to be a member..a true geniusAh, but the answers didnt match his...thus time to take his toys and go play in some other sand box Link to post Share on other sites
yergan 0 Posted October 13, 2006 Share Posted October 13, 2006 19=perfect sample sizeIts probably a bit excessive. I would aim for 5 - 6 for a sufficient sample size. Link to post Share on other sites
omahahilo 0 Posted October 13, 2006 Author Share Posted October 13, 2006 Its probably a bit excessive. I would aim for 5 - 6 for a sufficient sample size.I said many times during this thread that I completely understand that this def isan't enough to get solid results. I said discuss ... not prove I'm an idiot at all costs. Its called reading. Try it.Mattexactlywrong, you were originally looking to prove that somehow online poker wasn't completely random and then after you realized you were totally wrong you changed your mind to make up something else you were doing. ban.yes it is.Once again incorrect. I simply wanted to discuss if you thought this would be a result of online poker not completely random or variance or anything else ...I said discuss ... not online poker is rigged. What the **** is the matter with you people. Again I get it ... I know nothing ... you know everything ... you're a poker god and that's why you don't play and spend your time here. How nice it must be to be you. Matt Link to post Share on other sites
brvheart 1,752 Posted October 13, 2006 Share Posted October 13, 2006 It was about a real discussion as to if I might be onto something but then you all had to prove me wrong so then it became a fight. IE: "100% Incorrect"I don't have to prove its happend ... just that its possible. It is. Can you prove it hasn't happend. I would say you are on the side of false information Sir.That fact that you can't consider that as a possiblity if funny and shows your limited ability to think. It is possible. I am correct. I may have a point and I'm done with this thread you can be as ignorant as you want.ByeMatt MODERATOR PLEASE LOCK UP THIS THREAD. I AM TIRED OF THIS ARGUEMNT. I HAVE LEARNED MY LESSON. THEY WILL NEVER CONSIDER MY THOUGHS ON THIS AND I HAVE NO RESPECT FOR PEOPLE THAT "KNOW EVERYTHING" I WILL NOT POST IN THIS MANNOR AGAIN.LOCK IT LOCK ITLOCK ITLOCK ITMatt's number of posts in this thread since saying this = 10,000 +/-Once again incorrect. I simply wanted to discuss if you thought this would be a result of online poker not completely random or variance or anything else ...I said discuss ... not online poker is rigged. What the **** is the matter with you people. Again I get it ... I know nothing ... you know everything ... you're a poker god and that's why you don't play and spend your time here. How nice it must be to be you. MattHow can what I said be incorrect when I quoted you? Are you saying your quote is wrong? Like all of your 'points' in this thread, I don't understand. Link to post Share on other sites
omahahilo 0 Posted October 13, 2006 Author Share Posted October 13, 2006 I ran a little expirament.I played 25 $1.50 SIT N GOS on Pacific Poker.I chose this game because I knew the players would assist me.I pushed all in pre-flop with the first pair in the pocket I had 10-10 or higher and folded all the rest.Since I got called every time I found this very interesting.Out of 25 times I pushed all in I was the pre-flop favorite 19 times. I lost 16 of them to draw-outs.Discuss.Oh and I know I'm an idiot but look at these results.MattThis is my origional post ... tell me exactly where is says online poker is rigged? After this I fielded 50 remarks from people telling me I was an idiot. I got upset, which is definately not what I should/wanted to do and I said what I had to in order to defend myself. Again this is my first post.Matt Link to post Share on other sites
Dr_Shakes 0 Posted October 13, 2006 Share Posted October 13, 2006 This thread should go in the general poker forum hall of fame.We do have one of those don't we? oh and "Not as random as you think" = rigged Link to post Share on other sites
omahahilo 0 Posted October 13, 2006 Author Share Posted October 13, 2006 This is my origional post ... tell me exactly where is says online poker is rigged? After this I fielded 50 remarks from people telling me I was an idiot. I got upset, which is definately not what I should/wanted to do and I said what I had to in order to defend myself. Again this is my first post.Matt This is my origional post ... tell me exactly where is says online poker is rigged? After this I fielded 50 remarks from people telling me I was an idiot. I got upset, which is definately not what I should/wanted to do and I said what I had to in order to defend myself. Again this is my first post.Matt Link to post Share on other sites
omahahilo 0 Posted October 13, 2006 Author Share Posted October 13, 2006 Look ... I want to apologise.I am not sorry for defending my point because I believe I have something here ... at least a little bit. And I am trying to improve my game.But I apologise that I had to keep going. I know I'm gonna get flamed. We have too many people here with nothing better to do than fight regardless of validity. But I should have just dropped it.My next post will be better or just non exsistant. So, I have sent the Mods a PM and hopefully they will lock this up now.If you guys wanna fight that's awsome but I will controll myself from indulgeing you from now on. But I am sorry that this has carried on so long.Matt Link to post Share on other sites
Actuary 3 Posted October 13, 2006 Share Posted October 13, 2006 DAMNITmy god!I was reading thru this wondeing if any fuckin sense would show upThanks LLY!Probably the dumbest thing on here was IAN suggesting that losing 19 in a row or 16 was just like taking 16 of the losses from a 1000 sample and just bunched together. And then BrvHeart quoted it. IAN, stick to finding avitars with bouncing chineses tittiesLets asume by Favorite, we on ave were 55%.Losing 16, 17, 18, or 19 out of 19 should happen: 0.052790% of the timeThe stdDev for this experiment is 2.168524844So we are about 2.559343516 StdDev away from the EV of 10.45 wins.So while this as a low probability, it certanly is not crazy nor proof of non-randomness.And Matt, are you suggesting Pocket Pairs get out drawn, or every hand that goes in as a favorite, more that the odds would say?So, then, the non-randomness favors the underdog? I'm not following you on that part.gang, take it easy with this gang mentality against new posters. He was civil early you all jumped him. Unfortunatly you smothered him with ignorance.We have enough data in PT databases to easily confirm sites are randomps. Call my freakin hottline next time a post like this hits. I was bowling tonight Link to post Share on other sites
Actuary 3 Posted October 13, 2006 Share Posted October 13, 2006 ************************** Where's Actuary when we need him?I should be on call!************************19=perfect sample sizewell, 12 is sufficient for a Nrmal approximation, if we are 55/45 each time.Is that what you meant?yeah, I know something being <2% is of no use to prove rigged, it is still something to tie a number too.*************************************now to what you said about "same as 16 from a sample of 1000".Well, ok.What is a better way to say it, is to say: If we run this 19 hand thing 53 times, we'd expect to lose at least 16 out of 19 at least 1 time of the 53 trials, 2.741% of the time.(my numbers were wrong a few min ago... I was calculating winning 16+. which would be less rare) Link to post Share on other sites
Actuary 3 Posted October 13, 2006 Share Posted October 13, 2006 Furthermore, 25 is way too small of a sample size. Statistically you could've easily lost all 25.lolno. Link to post Share on other sites
Dogpatch 2 Posted October 13, 2006 Share Posted October 13, 2006 Finally, a voice of reason. Link to post Share on other sites
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now