Seems odd?

[zman]

8/10/2005 00:33 160998 ---- NETeller - Automatic Update -- -- 50.00 50.70
8/10/2005 00:34 4851624 View European BlackJack 2.00 0.00 -2.00 48.70
8/10/2005 00:34 4851625 View European BlackJack 2.00 0.00 -2.00 46.70
8/10/2005 00:34 4851626 View European BlackJack 4.00 8.00 4.00 50.70
8/10/2005 00:34 5596552 View Table - Vegas Strip BlackJack 3.00 0.00 -3.00 47.70
8/10/2005 00:34 5596553 View Table - Vegas Strip BlackJack 3.00 0.00 -3.00 44.70
8/10/2005 00:34 5596554 View Table - Vegas Strip BlackJack 3.00 0.00 -3.00 41.70
8/10/2005 00:35 5596555 View Table - Vegas Strip BlackJack 5.00 0.00 -5.00 36.70
8/10/2005 00:35 5596556 View Table - Vegas Strip BlackJack 5.00 0.00 -5.00 31.70
8/10/2005 00:35 5596557 View Table - Vegas Strip BlackJack 5.00 0.00 -5.00 26.70
8/10/2005 00:35 5596558 View Table - Vegas Strip BlackJack 5.00 0.00 -5.00 21.70
8/10/2005 00:35 5596559 View Table - Vegas Strip BlackJack 5.00 0.00 -5.00 16.70
8/10/2005 00:35 5596560 View Table - Vegas Strip BlackJack 5.00 0.00 -5.00 11.70
8/10/2005 00:35 5596561 View Table - Vegas Strip BlackJack 5.00 10.00 5.00 16.70
8/10/2005 00:35 5596562 View Table - Vegas Strip BlackJack 5.00 12.50 7.50 24.20
8/10/2005 00:35 5596563 View Table - Vegas Strip BlackJack 5.00 10.00 5.00 29.20
8/10/2005 00:35 5596564 View Table - Vegas Strip BlackJack 5.00 0.00 -5.00 24.20
8/10/2005 00:35 5596565 View Table - Vegas Strip BlackJack 5.00 0.00 -5.00 19.20
8/10/2005 00:36 5596566 View Table - Vegas Strip BlackJack 7.00 0.00 -7.00 12.20
8/10/2005 00:36 5596567 View Table - Vegas Strip BlackJack 7.00 0.00 -7.00 5.20
8/10/2005 00:36 5596568 View Table - Vegas Strip BlackJack 5.00 0.00 -5.00 0.20

This is the 17th losing sessions i have had at this casino.

[zman]

BTW I am playing perfect BJ.....

Is this normal at MG casinos?

[ftg]

Originally Posted by zman

BTW I am playing perfect BJ.....

Is this normal at MG casinos?

Your result is very normal in general.

[jamiester]

I count 4 wins and 16 losses. I know it is a very small sample but that still is a pretty horrible result. What are the odds of hitting 4 in 20 even bets?

I get 9690 * .5 to the 20th power

err.. I think I get a probability of .00924 or just less then 1 in 100 for 4 wins in 20 tries at 50/50 odds. My stats is rusty so maybe someone will correct me, but a 1 in 100 result is not that amazingly bad actually.

[raol]

Originally Posted by jamiester

I count 4 wins and 16 losses. I know it is a very small sample but that still is a pretty horrible result. What are the odds of hitting 4 in 20 even bets?

I get 9690 * .5 to the 20th power

err.. I think I get a probability of .00924 or just less then 1 in 100 for 4 wins in 20 tries at 50/50 odds. My stats is rusty so maybe someone will correct me, but a 1 in 100 result is not that amazingly bad actually.

The probability for 4 wins or less on 20 bets with 50 % win probability is 0.59 %.

[jamiester]

Is my calculation correct for exactly 4 wins in 20 tries?

Also could you show me the caclulation you did to get .0059 for at most 4 wins? Sorry I guess this should've been a PM

[raol]

Originally Posted by jamiester

Is my calculation correct for exactly 4 wins in 20 tries?

Also could you show me the caclulation you did to get .0059 for at most 4 wins? Sorry I guess this should've been a PM

No, the probability for exactly 4 wins is 0.462 %. You seem to have missed a factor of 2.

The probability for exaktly k wins in n trials if the win probability is p is
n!/(k!*(n-k)!) * p^k * (1-p)^(n-k)

[vinylweatherman]

I am running an experiment on this game for 100,000 hands. I am not looking at number of wins, but the overall outcome of 100,000 hands of perfect BJ with unit stake per deal.
I have subdivided the trial into 20,000 hand blocks, and tabulated the change in balance after each 1000 hands (this will show the streaks, good or bad, and the 20,000 and 100,000 hand results will show the overall results).

The interesting calculation will come at the end though, how close to expectation is the end result, and is it statistically significant in terms of identifying a bias in favour of the dealer (or player!) in the long term.

Vegas Strip, played perfectly, is the closest to 100% payout found online, and MG swear by the random nature of the games, and Price Waterhouse have verified publically that specifically MG Blackjack is random over the long term, and also does not show bias over the medium term.

This may have changed though, as MG updated the games, and have updated the BJ tables strategy guides. My experiment is using this updated table, which is verified on wizardofodds and correct. The MG table has a few extras though, as the strategy can vary according to the number of cards present in the players hand, not just the value of the hand against the dealers up-card.
If the game is random, I should lose some of my bonus, possibly all of it!

Can anyone post the probability that I will end this 100,000 hand experiment at MG Vegas Strip in profit on my unit stake of £1 per initial deal?

[scrollock]

Originally Posted by zman

BTW I am playing perfect BJ.....

Is this normal at MG casinos?

from my recent results at MG's that is very normal.

i'm glad to know that its not just me that has noticed the difference in MG BJ recently, there has been some kind of tweaking going.

my last 10 sessions on MG BJ, all of which have been 500+ hands have shown a loss ratio of at least 5%.

i.e. my best session out of the 10 was a still loss of 5%