Full Tilt Poker Report: What's The Deal with the Randomizor?
It's been talked about for a couple years now in regards to the hand randomizer at Full Tilt Poker. It stinks! More players complain about bad beats than any other site, and thorugh my network of subscribers poker websites, poker forums and blogs, along with personal experience it's hard to refute. JP and Sun 137 boted this in the forum below.
JP: It's been confirmed with e-mails to stars and full tilt.
Stars: once cards are shuffled and deal starts, it's fixed and dealt like live cards would be dealt.
Tilt: Shuffles and deals, then continuosly runs the numbers/cards until betting is done, deals the flop, and again spins the wheel non-stop until betting is done, deals the turn, then yet again starts the numbers running non-stop until betting is done and final card hits the table.
By changing when you place your bet alters the cards that will hit the table. Each player taking different lengths of time to bet also alters which card will hit the table. If you had bet quicker or slower, the board would have been altered.
Technically, this doesn't change the outs and % of cards remaining, but it DOES alter the outcome of the hand compared to a normal live or fixed deal. IF you hold AA and an A was 2 cards down, you would get that A on stars, but you won't get it on FT unless it rolls around again by the time the last bet is made.
The guy at Tilt also said he would research further for me in the next few days and give me more information on how the numbers are generated as much as he can without breaching any security precautions they have. He further went on to say that they are looking into 3rd party audits and putting more information probably published on their site in the future. No firm dates or commitments other than it is currently being looked into.
He sent me an e-mail about 2 pages in length and said more would follow. I'm just putting the breif key info I received above FYI only.
I'm going to look more at my bet timing and hands played on FT. If I find I take x time often to bet and it's fixed and my cards are getting trashed, I'll shorten or legthen my time to bet. While the card hitting next is always going to change based on what others do, my change may have little or no effect on the outcome. On the other hand, it just might. So I will be testing it myself over the weeks ahead.
Personally, I think it's a rotten way to deal cards.
Sun137: We must be telepathic JP. I noticed your short comment in another post about FT card dealing a few minutes ago, so I went and made a cup of coffee and was going to make a post describing how the deal works.
Then I come back and see you have just done it.
Maybe we are twins............
Anyway, just for the record, here is my 2p worth.
Last year, when watching Poker Night Live, one of the presenters was playing online in a MTT. They were showing a hand she had just finished, where she had a flush draw on the turn, villain bets and she folds, someone else calls and the flsuh card hit the river.
She said, I should have called his bet, and the other presenter said 'ah, but you don't know how that would have affected the RNG, the flush card might not have hit then'.
As soon as I heard that, a light bulb came on in my head (I know......), as I thought this is crap - you deal the cards, and that is it. So I checked with all the sites (6) I played at - you guessed it, they all, except one, use a combination of mouse movements, clicking (therefore betting/calling/folding) from all the tables, broadband/cable 'noise', random number patterns and other random things EACH TIME THEY DEAL A CARD AT EVERY TABLE.
They advised me it is their belief this produces a more random shuffle, but my view is simple. You shuffle 52 cards, and deal 1 hand from that 1 shuffle. That is poker, pure and simple. You do not lift the pack each time you deal a card, reshuffle it and deal another card. That is NOT poker.
The more critical thing is, however, a poker site should NEVER allow the random betting patterns of the players influence the outcome of the cards - even if they are not directly related and do not/cannot be linked in any malicious way. I consider it scandalous. It is even, judging by your comments, going to change how you play poker - and it shouldn't.
The only site with a proper poker deal - you guessed it - was Poker Stars. They advised me a year ago the deal for each game at each table was entirely random (and provided me with the detailed maths as how they do it, most of whic is on their site anyway) and once the deal is made, IT IS NOT CHANGED, irrespective of the betting patterns - all they do is deal the cards.
In fact the staff at Stars were very serious about pointing out to me the player actions are NEVER included at any stage, not even in the random conditions they use for a shuffle.
I have, on many occasions since I joined the forum, advised Stars is my home site where I allocate the majority of my monthly time budget. I have rakeback at Fulltilt, so low level cash games are all I will play there for steady rakeback income, and I will play other sites on occasion depending upon opportunities available.
My approach will not change until the other sites sort themselves out - I am not saying you can influence the cards by changing your betting patterns, not at all, just that I think their system sucks. Big time.
JP: Well it depends on if it's random or sequenced with the rng and how they use it. If I space my time a certain way and cards fall a certain way, then I change and it will change how they fall, if the change is consistent and how they fall is changed but consistent within that change then no, it's not random.
When I wrote to Tilt it was quite lengthy and stating my opinions, beliefs, what I had tracked, what my perception was and what I had found in research regarding RNG's and how they work. I also stated that it appears Tilt is skipping a step that others use and the continuos shuffle meant to difuse this skipped step is not difusing it and still having an affect on the game that is not "normal" in the terms of what poker is and should be.
You have to look at the fact that ALL the EV calcs are based on millions of hands and caclulations but based on a single deck, shuffled once, and dealt with burn cards. Because you change the shuffle and the deal, this is going to affect the EV. Because it is NOT dealt in the same manner following the identical parameters as the ones that for years were used to develop all the charts and ev results patterns.
If they use a sequencer or two mixing them together into a string, then keep the string running, they are in the same sequence when they come out. What point the sequence stops running to deploy is predicated on the time used to place all bets up to and including the final bet. MOST people are fairly consistent in the time to place their bets. They already know what they are going to do in most given situations and they time their bet no matter what to be close in time so as not to give a table read unless they are trying to give a false read making a fast or slow bet. Sometimes you have to really think and weigh the situation to make your wager. But for the most part the better the player, the more consistent their bet timing is. So if I find that my bet timing is yield xyz and I change my bet timing and end up with abc, then I have hit on something that I have to change in my game to deal with this bogus deal.
Yeah, theoretically it would still be random. But if that were the case, we wouldn't consistently be seeing patterns like we do. It would really only be random if the cards were run through a final filter rather than off the rng or sequencer. I think FT uses sequencers and no filter or real rng, it's a cheap way to do it it. In order to break the sequence, they keep the sequence string running non stop until the final bet is made, then it starts running again. This way you can't learn that the 5h comes after the Js in the sequence they are running. Because based on the time taken, the next card might be 3 or 4 spaces away on a 4th or 5th revolution of the numbers in that sequence when it stops running and is shoved out.
Say KTo has ev +.05 and AKo has ev +1.14 (not that it does, just numbers pulled as an example with one higher than the other) EV is because of known cards and hands played over time yielding certain results. If you change the deal then that ev is no longer the same. KT and AK may end up with the same EV instead of different ev's because the deal is constantly changing. I want cards to be random, not EV. What this type deal does is make EV random. Cards might still be random to a degree and the odds and outs are still the same. What's changed is the deal and how those outs come out. I don't see how long term, the EV can possibly be the same as what everything is based on since it's based on original true poker dealing premise and the continuous shuffle is something you would shoot someone in the casino for doing in the old west. Nowadays you'd problaby hop the table and attack the dealer or just call the pit boss over, they look at the video of the deal and then fire and ban the dealer.