Short term strategies, part deuce

[oej719]

Got to find a restroom after reading all these loooonnnnggg post.
wizzzzzzzzzzz.
Ok keep em coming.

[New2vp]

I was going to suggest a way of doing it, but I saw that New2vp beat me to the punch with not one but several suggestions.  It might take longer than the couple of hours he says () but still much less time than actually playing out sessions with VPFW. Decided to try the challenge.  It took 75 minutes to put together the spreadsheet, do 208 simulations, and summarize the results.  It would probably now take less than 5 minutes to do 208 more.I know you're all dying to know the results.  Parameters: Bankroll $1000; Win Goal of $100; 7/5 BP, 8/5 TBP; steps of $0.1,$0.25,$0.5,$1 (no playing extra after whatever big wins would be, though that could be figured in as extra-session play if necessary; of course, a basic tenet of short-run play on negative machines would be to take the money and run, since extra play generally leads to losses).  It was difficult to decide what "AP" to play against the ARTT but since these are both negative games, the decision would be to play $0.10 Triple Bonus Plus the same number of hands as it took for each session (highest EV of the 2 games available, lowest stakes).Summary numbers are 146 wins and 62 losses.  Total net $ lost:  $2756.75 out of $542,935.25 through machine (about 0.51% lost)112,702 hands (average bet $4.82)Fewest hands in a session 16.  Most hands in a session 2377.  Average hands in a session 542.  Most dollars won $1236.  Most dollars lost $999.25.44 of the sessions (21%) ended with losses of more than $970.  Recall rd's max loss was in the $960s.  Could rd have been that lucky and me that unlucky for me to have that many of my losses nearer to $1000?  I'll leave that to you to judge.Comparison to constant play with TBP at dime stakes:  143-65 Total lost: $330.00 out of $56,351 (loss of 0.59%).  To be fair, if the results ended in net wins, ARTT would generally win because of the higher stakes being played.I thought about varying the bet size of the AP game, but then that wouldn't be a fair comparison since those who do not play ARTT are generally not varying their bet size.  And of course if I were to always play at stakes of $1, then ARTT would have the advantage since its average bet is less than what you see on a $1 machine.  There are no 96-cent machines, so a perfectly-balanced comparison is not realistic.Disclaimer:  I should note that in my haste to beat the challenge time, I may have made some underlying error, but cursory checks have not uncovered these thus far.

[New2vp]

Ran everything again.  Looks like it really takes 10 minutes to do another 208 simulations and prepare a Forum post together.  I had issues with pasting from Word which is why the previous post is blank.WEBMAN, IF YOU WANT YOU CAN REMOVE MY PREVIOUS BLANK POST.2nd run:  Summary numbers are 125 wins and 83 losses.  Total net $ lost: $26,528.25 out of $542,935.24 (about 4.10% lost).133,461 hands (average bet size $4.85)Fewest hands in session: 4.  Most hands in a session: 2914 (Good thing it was not too many more; my current spreadsheet maxes out at 2984 hands).  Average hands in a session 642.  Most dollars won $3384.25.  Most dollars lost $999.25.  (If the player does not have enough left from his bankroll to make the next play, he stops in my simulation.  $999.25 is a relatively common figure as it represents the case that someone loses the first 4 bets and never gets out the red.  About 1 time in 11, the player will whiff on the first 4 plays, though even when he exhausts his bankroll, he usually banks some money along the way.  Also, there will be some more times that the first few plays never yield a win if you consider that hitting a pair of jacks through aces do not win anything for the player.)54 of the sessions (26%) ended with losses greater than $970.  Looks like my previous run wasn't that unrepresentative.  Either I have a major bug, which is possible, or there is considerable evidence that rd's worst session result was made up. Comparison to constant play with TBP at dime stakes.  124-84.  Total lost: $2556.00 out of $66,730.50 (about 3.83%)Looks like an unlucky run for both ARTT and "AP"

[cddenver]

Decided to try the challenge.  It took 75 minutes to put together the spreadsheet, do 208 simulations, and summarize the results.  It would probably now take less than 5 minutes to do 208 more.

 
Eek.    I wasn't making a challenge,  I was thinking of how long it would take me to do it.   That's nice work!

[New2vp]

Thanks,  You do learn things that in retrospect are obvious if you actually do the simulations.My guess is that neither RD (nor his mentor) has a good enough statistical background to actually understand the properties of what reasonable results from these playing strategies should look like.At the risk of helping both of them to fudge the data better in the future, it should be obvious that a significant portion of the losses would be quite close to the bankroll.  And with a bankroll of $1000, in a sample as large as 208 sessions with the playing strategy announced, it is beyond a reasonable doubt that the data is fudged if the worst session reported is a loss of less than $970.And this really isn't dependent on which games were played.

[shadowman]

When I did my analysis of single line play one of the first things I discovered is it proved the math is correct in all areas. When I tried games with poor paytables then the results got worse unlike the claims of the inventor.
Good work, new2vp, when dealing with people making ridiculous claims the use of real data is always a nice touch.
 

[Tedlark]

Real data, gee guys is that like using math?

[rolanddude]

Thanks New2VP...So since my comparisons were keeping the AP bet constant at $1 instead of .10 then I now know that in your first sim  ARTT lost $2,756.75 and Advantage play would have lost $3,300.  In your second, ($26,528.25) ARTT......  ($25,560.00) AP.......We now have 3 "208" session sims with ARTT fairing better than AP 2 out of 3 times so far...(it is a nice touch to use real data Shadowman).What if the sims were run on better paying games?Can you email me your sim spreadsheet w/instructions?  I'd be glad to run many sims as I sit here.  Since 624 total sessions is a far too few samples to go by....My worst losing session was accurate...one in a million I guess..???  You need to blame the rng in Dancer's  VPFW that I purchased for producing that result.  Not "beyond a reasonable doubt" say I fudged anything.  Just one example that actually playing the hands will produce random results far skewed from EV sims sometimes. (most of the time not this wide of a gap though)FYI...great article...www.ggbmagazine.com/articles/Evolving_Slot_Scamswritten by a casino security supervisorA little part of the article deals with rng tampering by employees, techs, etc.....VERY Interesting read.Again New...I don't know how to program excel to do what you've done...If you could help by sending your spreadsheet program to me you would be a saint.Thanks,JoePS...Am I finally getting some help trying to accomplish what I came here for?   I only want to trial this system ...  not sell, endorse, or preach here to get converts for it.  Simply which is better...

[royal flush]

play here you cant lose better off for you

[shadowman]

FYI...great article...www.ggbmagazine.com/articles/Evolving_Slot_Scams

written by a casino security supervisor

A little part of the article deals with rng tampering by employees, techs, etc.....VERY Interesting read.




 
Most of this has been covered on this forum previously.
 
Ron Harris worked for gaming control. He had inside access.
 
The American Coin situation was 20+ years ago and in the control of the company making the machines. They lost their license.
 
The one where the cheaters timed the machines was in the early 90s and is why the manufacturers went to running the RNG between deal and draw.
 
I believe these are the only ones related to VP. Note that the only one where the return was impacted for regular players was American Coin which was a mnor player in the market and they payed for it. This is pretty good evidence that the liklihood you are being cheated is very, very small.