Oddguy2017
Joined: Feb 2, 2017
• Posts: 7
February 2nd, 2017 at 11:40:05 AM permalink
Update on venetian. There are two large screen \$5 min blackjack 5-player machines. They are 3-2, 4 deck, early surrender.

I am suspicious these machines are fair. I spent my day reading on here about basic strategy. I printed the strategy table vle and referenced it every hand. I played \$10 bets on a \$100 roll.

I lost all \$100 in 40 minutes with 5 people on game the whole time. I was dealed several bad cards, getting 12-14 several times.

With that game and my bet limits, fair odds and house edge calculator would have said I could last hours. Of course there is variation, but \$100 in 40 minutes is quite a large deviation and highly unlikely.

Since this was my first longer run with the machine, I am suspicious this machine isn't exactly fair. I could have just gotten very unlucky but that is very unlikely from probability. I wanted to play for a few hours and modified my betting to give me at least a couple hour expected play.

On paper, that machine offers the best blackjack in venetian, and good overall. However, if something isn't right with it then it's very misleading.

Any thoughts?
Joined: Nov 3, 2014
• Posts: 108
February 2nd, 2017 at 11:43:33 AM permalink
extremely small sample size. i've lost \$100 playing \$5 blackjack in 20 mins or so. Tell me what it's like after 1000 or 5000 hands....
SM777
Joined: Apr 8, 2016
• Posts: 762
February 2nd, 2017 at 11:50:54 AM permalink
The machines are fair.
Romes
Joined: Jul 22, 2014
• Posts: 5518
Thanks for this post from:
February 2nd, 2017 at 12:00:16 PM permalink
I'd be rather willing to bet these machines are fair.

Assuming they have a generic .5% HE, you played 40 min so let's say something like 60 hands, and you bet \$10 the whole time.

OriginalSD = 1.15 * AvgBet = 11.5 ...(1.15 is the variance of blackjack)

EV(x hands) = (AvgBet*NumHands)*(HouseEdge)
SD(x hands) = Sqrt(x) * OriginalSD

EV(60 hands) = (10*60)*(-.005) = -\$3
SD(60 hands) = Sqrt(60) * 11.5 = ~\$90

to be 95% confident we need to use 2SD... which 2SD = \$180

So with 95% confidence when you played 60 hands you could expect to lose \$3 +/- \$180... which means losing \$100 is well within that range. Nothing out of the ordinary here at all since you're looking at a VERY SMALL sampling size so the variance can (and will) be wild. Would you think the machines were gaffed for the player had you won \$100 in 40 min? After all, each and every time you play 40 min on that machine for the rest of your life you should expect to lose \$3.
Playing it correctly means you've already won.
mcallister3200
Joined: Dec 29, 2013
• Posts: 2877
Thanks for this post from:
February 2nd, 2017 at 12:09:52 PM permalink
I think they are late surrender not early surrender.
Oddguy2017
Joined: Feb 2, 2017
• Posts: 7
February 2nd, 2017 at 12:11:44 PM permalink
The number of hands was definitely under 60. No more than 40 for sure with the 5 people on and waiting time for bets on the machine. How does that affect your assessment? It should be a smaller range right?

Since it was my first time really on the machine, getting - \$100 on a -\$183 - +177 variation I knew was unlucky (and I think your calculations will have a tighter range with smaller hands) . I didnt count the number of hands but it was definitely under 60, try using 30 & 40.

I am just suspicious at this point, and getting a very bad result on the first shot adds to that. It is possible, sure, just unlikely.

People say machines are fair because the law requires it, but humans have to code the machines and there are always bugs. Ones that help the house don't make a priority list unless someone else discovers it (good luck proving that). Ones that hurt the house will be fixed pronto.
Oddguy2017
Joined: Feb 2, 2017
• Posts: 7
February 2nd, 2017 at 12:13:21 PM permalink
Sorry I might have said that wrong. It only lets surrender for your first option, not after a hit. Also, no surrender allowed when dealer has an ace.

I took photos of the machine rules.
odiousgambit
Joined: Nov 9, 2009
• Posts: 8657
February 2nd, 2017 at 12:16:53 PM permalink
I am ashamed to say I went a long time without realizing the machines often pay even money for a natural. In fact I flatly refuse to reveal how recently it was when I wised up! Thankfully, I almost never played them.

Since then I have been unable to find a machine on the East coast that paid correctly; not that I have any way of being thorough.

So I guess you can tell I am saying I hope you have made sure naturals don't pay 2 for 1 - which is how they usually put it.
the next time Dame Fortune toys with your heart, your soul and your wallet, raise your glass and praise her thus: “Thanks for nothing, you cold-hearted, evil, damnable, nefarious, low-life, malicious monster from Hell!” She is, after all, stone deaf. ... Arnold Snyder
Romes
Joined: Jul 22, 2014
• Posts: 5518
February 2nd, 2017 at 12:17:43 PM permalink
Quote: Oddguy2017

The number of hands was definitely under 60. No more than 40 for sure with the 5 people on and waiting time for bets on the machine. How does that affect your assessment? It should be a smaller range right?

The reason I gave the generic formulas is so that anyone ( ;-) ) could now run them for whatever numbers they desired...

EV(40 hands) = (10*40)*(-.005) = -\$2
SD(40 hands) = Sqrt(40) * 11.5 = 72... 2SD = \$144

You're still well within the regular range... to imply cheating would have to be outside of 3SD AND you'd have to have a decent sampling size to go with it.

The smaller the number of hands the LESS of a case you have. That's like losing 1 single hand and saying "The game clearly has a 0% payout programmed in!" Well no, you just lost one hand...

Quote: Oddguy2017

Since it was my first time really on the machine, getting - \$100 on a -\$183 - +177 variation I knew was unlucky (and I think your calculations will have a tighter range with smaller hands) . I didnt count the number of hands but it was definitely under 60, try using 30 & 40.

I am just suspicious at this point, and getting a very bad result on the first shot adds to that. It is possible, sure, just unlikely.

Why would getting bad results add to anything other than your non-math based suspicion? Sure, it can add to your suspicion all day, but your suspicion doesn't mean anything towards the machine being fair or not. Math is what can determine that. I've seen the exact machines you're referring to, and I highly doubt in a prominent 'mid-strip' casino such as the Venetian that they would put in gaffed \$5 video blackjack machines =p. So much so that I would be willing to bet you money these are fair games =).

Quote: Oddguy2017

People say machines are fair because the law requires it, but humans have to code the machines and there are always bugs. Ones that help the house don't make a priority list unless someone else discovers it (good luck proving that). Ones that hurt the house will be fixed pronto.

I'm also a programmer, and you're right they could, but why would they? They have a game where they are GUARANTEED TO GET MONEY FROM because it has a House Advantage built in. There's literally no need to program it to make "more" money while putting their gaming licence at risk and adding the possibility of fines. Past that, they have companies that review the code searching for said bugs and certifying the games randomness. Yes, there have been cases of cheating casinos before (mostly indian casinos but yes non) and while it does happen it's very few and far between (and usually not in Vegas).
Playing it correctly means you've already won.
Oddguy2017
Joined: Feb 2, 2017
• Posts: 7
February 2nd, 2017 at 9:37:28 PM permalink
Quote: Romes

The reason I gave the generic formulas is so that anyone ( ;-) ) could now run them for whatever numbers they desired...

EV(40 hands) = (10*40)*(-.005) = -\$2
SD(40 hands) = Sqrt(40) * 11.5 = 72... 2SD = \$144

You're still well within the regular range... to imply cheating would have to be outside of 3SD AND you'd have to have a decent sampling size to go with it.

The smaller the number of hands the LESS of a case you have. That's like losing 1 single hand and saying "The game clearly has a 0% payout programmed in!" Well no, you just lost one hand...

Thanks for updating on that, that confirms what I suspected that I was even unluckier on my first try at this game, being near the bottom of the range (and as you know, the tail ends of the range are less likely than near the mean)

Quote: Romes

I'm also a programmer, and you're right they could, but why would they? They have a game where they are GUARANTEED TO GET MONEY FROM because it has a House Advantage built in. There's literally no need to program it to make "more" money while putting their gaming licence at risk and adding the possibility of fines. Past that, they have companies that review the code searching for said bugs and certifying the games randomness. Yes, there have been cases of cheating casinos before (mostly indian casinos but yes non) and while it does happen it's very few and far between (and usually not in Vegas).

I'm not implying that they intentionally are rigging the machines, but I am implying that any bugs favoring the casino have a lot of reasons why they wouldn't be high priority fixes (unless it was discovered through a legal regulator or through months of collecting data on it to show an issue). However, a bug that was against the casino would only need the casino to discover it and you can bet that they would make that a very high priority fix.
There are bugs in every piece of non-trivial software - I would bet money that bugs exist in every single video/machine game in a casino.

... and btw, I was surprised you would think that a casino has no reason to want to add an advantage above the mathematical values? That's like asking why any entity would want more than a given amount of money - greed, self-interest, profit. Of course they would be motivated and tempted to add in additional advantages that they can get away with it. That's why they are very self-interested in discovering and fixing any bug that hurts them, but not as interested in discovering bugs that help them.

Getting a statistically very unlucky streak on your first seat at a game raises suspicion. I also doubt that there is much of a fault with the machine, but it sure made me wonder. I took photos of the rules again to see if there was anything I was missing in them - but they are fairly liberal rules for vegas (and I prefer the video machine where I can sit non-embarrassed with the basic strategy tables in front of me).

Most likely, I was just very unlucky - but it sure gave me pause and disappointment after I just spent hours learning how to play basic strategy and finding one of the lowest house-edge blackjack plays in Vegas.

btw - I went across the street to TI based on web reviews. They did have a live 2-deck table, \$10 minimum, 3-2, h17 during a weekday evening, which would be better odds if other rules were the same as the video. I didn't have time to ask about the other rules though like surrenders and doubles, but they were probably good since the main ones were. I would give up a little edge though to play at a machine personally though.