Microgaming Blackjack

ugh...:mad: never again...

Has anyone ever managed to beat this blackjack?

The results seem about as random as an Italian football match.

It's VERY streaky. I did beat it though, 1 flat betting with the expert mode. 100,000 hands and 800 up. Highly suspicious if you ask me, that is one hell of a performance for 100,000 hands, and if it was 800 DOWN I would have been less surprised.

Because of this, I did wonder if there was some kind of bias in the way the RNG produced the results, and I looked at tweaking the strategy to match this quirkiness, rather than using perfect strategy. It didn't get me very far, but it turned out MG had made an error with the defaults, and these have been corrected.
 
Since November of last year, I've played at Lucky Nugget, RiverBelle, Wild Jack, 777 Dragon, Casino Share, 7 Sultans, King Neptune, and Red Flush. I've done very well in MG Blackjack (Vegas Strip version.) At one point last month (June), my total withdrawals were almost $14,000 on total deposits of around $4,000.

Then in mid-June, things went to hell. I got a lot of blackjack sessions of one-and-dones which are like 5 losses in a row, then one win, 4 losses in a row, 1 win, 1 loss, 1 win, 3 losses, etc. In other words, streaks of losses separated by one win. I usually up my bets on a win but obviously one-and-dones are bad for that type of betting strategy. The long-run winning percentage of Vegas Strip blackjack is around 43% but I was/am getting a number of sessions in a row with below 40% winning percentage to boot.

I've deposited over $4,600 since things went bad. One thing I've noticed is that when I go to a new MG casino, it appears my player history from other MG casinos follows me. After making a deposit, the sessions immediately are always filled with one-and-dones. Many times I'll flat bet the one-and-dones then a double or triple win streak happens, which gets me out of my flat-betting mode. Then the one-and-dones immediately kick in again making me lose a lot. Ah ha! They are now employing psychological tactics now, for this has happened often. This would only make sense if you believe, as I do, that the online casinos use Artificial Intelligence in analyzing your betting strategy. Random?

Also, since mid-June, I've noticed a lot more losing double-downs/splits after a win in which I usually up my bets. Random?

It's harder for me to get past $500 in my balance. It's as if they know I do well when I get past $1000.:D
 
It does definately feel like it has an A.I. cause the games get a lot harder the longer you play, and the difficulty continues onto future sessions, to the point where it becomes near impossible to win. You can often just start a new game and lose 5 times in a row... even though you haven't played for several days.

I get the feeling it has a payout meter which is constantly trying to balance out the amount you can win, a bit like the slots do, designed only to payout a certain amount, dictated by a fixed percentage of win/lose - which if this is true means it can't be random.

I am trying to confuse it at the moment by betting random, constantly changing my strategy, this seemed to work ok for a while but soon the dealer was constantly getting 20's, 21's and bjs :( Think maybe I must have pissed the A.I. off... time to play a different game for a while I think...
 
It does definately feel like it has an A.I. cause the games get a lot harder the longer you play, and the difficulty continues onto future sessions, to the point where it becomes near impossible to win. You can often just start a new game and lose 5 times in a row... even though you haven't played for several days.

When I first started playing online blackjack, I often got games where I would have blocks of wins: win 3, lose 4, win 5, lose 3, win 2, etc. Not so hard to win with the way I play. That was then. Nowdays, I get this type of session only when the overall win percentage is "adjusted" back up to reach the normal 43% overall win percentage for Vegas Strip BJ. It definitely has gotten harder and harder to win a session when it appears to be adjusting to your style of betting.

I do not use the martingale strategy but I did use it several times the past month. Results? Each time I got loooong losing streaks and a win right when I stopped the progression. To me, this means the A.I. or whatever is immediately counter-acting ANY successive progression betting I use. It used to be that it took the software from 10 sessions or longer to "figure you out" but now it is almost immediate that it notices that you changed or tweaked your strategy. Random?

I get the feeling it has a payout meter which is constantly trying to balance out the amount you can win, a bit like the slots do, designed only to payout a certain amount, dictated by a fixed percentage of win/lose - which if this is true means it can't be random.

Right, random it ain't. I certainly would be frickin' mad if MG (or any other software casino platform) start to force losing hands anytime I bet my big bets. Lasseters Casino did this on their 2 cents minimum bet Blackjack game when I played their several years ago. It was obviously NOT random there as there were an abundance of back-to-back-to-back looong lossing streaks (record for me: 22 straight losing streak followed by another 10 :eek: ) Everytime I bet bigger, BAM! I lost! That was absolutely rigged! If Lasseters does/did it, I'm sure the others do it too though to a lesser extent (but still has same effect on your bankroll!)

I am trying to confuse it at the moment by betting random, constantly changing my strategy, this seemed to work ok for a while but soon the dealer was constantly getting 20's, 21's and bjs :( Think maybe I must have pissed the A.I. off... time to play a different game for a while I think...

With online casinos, you can have low Chi factors indicating strong randomness in the cards drawn AND still manipulate the win or lose outcomes. This is a problem with virtual, invisible playing cards that are handled by the casino from beginning to end as opposed to seeing the cards in a real casino shoe.

A real casino can get into trouble with government authorities for running rigged games. Is it the online casinos turn now?
 
It's VERY streaky.

If this is true then the game cannot be random or fair as it should follow the same laws or probability as a real deck of cards.

Westland Bowl said:
With online casinos, you can have low Chi factors indicating strong randomness in the cards drawn AND still manipulate the win or lose outcomes. This is a problem with virtual, invisible playing cards that are handled by the casino from beginning to end as opposed to seeing the cards in a real casino shoe.

I could run some statistical analysis in case someone is willing to provide a sufficient play data sample (1000+ hands). I would calculate the histogram of the lengths of consecutive runs and compare that to the expected frequency distribution with Chi-Square test, to determine whether the "streakiness" is outside of limits of normal variance.
 
I could run some statistical analysis in case someone is willing to provide a sufficient play data sample (1000+ hands). I would calculate the histogram of the lengths of consecutive runs and compare that to the expected frequency distribution with Chi-Square test, to determine whether the "streakiness" is outside of limits of normal variance.

Well, let me know the details of how to get it to you. The software I use to track the blackjack play can only export in .txt or html format and in one session at a time. So I guess you have to merge them together. I have about 6 casinos in which I suspect funny play. Just send me a private message through this forum. Thanks.
 
Due to my postings here, I gotten five negative reputation points.:what: I find that MiniVegasGroup representative visited my public profile so I suspect that is where I got it from but I have no proof.:rolleyes: Whoever did it should have emailed me as to what I written that offended them/him/her so.
 
Due to my postings here, I gotten five negative reputation points.:what: I find that MiniVegasGroup representative visited my public profile so I suspect that is where I got it from but I have no proof.:rolleyes: Whoever did it should have emailed me as to what I written that offended them/him/her so.

That's out of order.

Whoever did this should explain their reasons for doing so. I saw nothing offensive in Westland Bowls posts...
 
Due to my postings here, I gotten five negative reputation points.:what: I find that MiniVegasGroup representative visited my public profile so I suspect that is where I got it from but I have no proof.:rolleyes: Whoever did it should have emailed me as to what I written that offended them/him/her so.
Did you check your User Control Panel??...hell I have a troll who negative reps. and baits me for sport and five sounds about right...guess the troll thinks I will turn into a gentleman, no way:D
 
Perhaps it is a good idea to post some results in this thread. Maybe there is a pattern to be found?

Yesterday I played MG Spanish Blackjack and had 82% overall return over 470 hands played flat-betting. A simple analysis showed 1 in 5525 chance to lose this much over the whole session. As you can see the bankroll was constantly around purple -3 SD curve when it should fluctuate around the top black curve.

Statistically the bankroll should be above blue curve 97.7% of the time and above purple curve 99.9% of the time.
 
^Wow! That is pretty incriminating evidence there! I bet that the Chi-factor for the randomness of the cards shows "fair" randomness at the same time? A fair random set of drawn cards can be "rearranged" to favor the dealer (or player.)

You know, that chart can be very useful to know when to bet big. Where the bankroll line fell way below the purple line (representing the 3rd standard deviation) at around the 190th hand, you know that it has to get back to the purple line at least (meaning you start winning more than you lose), so you can start betting $100 or more until it reaches back to the purple line, which seems to be around the 220th hand. You could even martingale during that phase. So you could theoretically win $3,000 or more during that period and go back to flat betting until you start another up-move chart-wise.

I used to be a commodities trader! : )
 
Updated chart

^Wow! That is pretty incriminating evidence there! I bet that the Chi-factor for the randomness of the cards shows "fair" randomness at the same time? A fair random set of drawn cards can be "rearranged" to favor the dealer (or player.)

I didn't yet evaluate the dealer cards but I will do that when I have the time.

Actually the chart I posted before was UNDERstatement. I had assigned too high standard deviation to Spanish Blackjack. I now confirmed that the correct standard deviation is 1.20 per hand. This makes my result even more biased. I am not able to edit the previous post so here is the correct graph of my bankroll and Standard deviations.

The bankroll is BELOW 3 SD most of the time and reaches -4.2 SD at hand #189. There is only 1 in 107 000 chance to lose as much as I did by hand #189.
 
Spanish BJ results continued

I continued my Spanish Blackjack experiment with 32Red Monthly bonus. Starting with 64 credits, I reached 66 credits and then busted from there. Adding this result to my previous sessions I am now more than 4 SD below expectation. Statistical calculation shows 1 in 25000 chance for my loss, if the game is fair.

Here is the updated bankroll chart:
 
32Red Chat response

I also asked some overall statistics from 32Red staff and this is what they replied:

Juho: Having looked at these statistics my results look to be generally south from expected. For that reason can I request a full list of bets and results I made?

Dominic: We expect you to get between 92 and 96% payout percentage that mean for every 1 unit stake you get .95 back
Dominic: Now if i give you the figure
Dominic: 67209.50 wagered for 64287.33 payout
Dominic: thats exactly 95.6% payout on tables game
Dominic: Totally between what we expect you to get
Dominic: Of course i will request a full audit for uyou if you wish but looking at your game play there is nothing abnormal to your figures
Dominic: Anything higher the 90% is totally normal and what is espected

Juho: All the table games I have played have low variance and a return percentage between 98.0% to 99.9%. Therefore 95.6% is way below expected.
Juho: 95.6% is expected if you play slots. It is not expected if you play blackjack.


Dominic: 98% is only for blackjack mainly
Dominic: the other one are listed anywhere between 92 and 96 %

Juho: Blackjack has a return of 99.6%
Juho: For example Spanish blackjack 99.62%

Dominic: we dont advetise those number Juho
Dominic: that you have gave me

Juho: It doesn't matter what you advertise but those are the numbers if the game is fair and random

Dominic: thats may
You do not have permission to view link Log in or register now.

Dominic: if you check all of them you will see that we are no way near your 99% on tables games
...

So basically the response I got was that the payout at 32Red is fixed between 92% and 96% and that my long term payout of 95.6% is normal according to them. The problem is that the true return on blackjack is between 99.5% - 99.9% and 95.6% long term payout in more than 10 000 hands played is way below expectation.

I hope that 32Red representative comes here and answers these matters that the customer support told me.
 
Were you using Microgaming autoplay for the results above? MG autoplay seems to have some errors for Spanish BJ, particularly with post double strategy. Perhaps post double strategy is not supported? There are also some other close calls hands which may have issues.

Note that the 99.62% payout assumes optimal strategy. Hardly anyone uses optimal strategy for this game, so the 32 Red payout data is expected to average well below the 99.62% figure.
 
Last edited:
Were you using Microgaming autoplay for the results above? MG autoplay seems to have some errors for Spanish BJ, particularly with post double strategy. Perhaps post double strategy is not supported? There are also some other close calls hands which may have issues.

Note that the 99.62% payout assumes optimal strategy. Hardly anyone uses optimal strategy for this game, so the 32 Red payout data is expected to average well below the 99.62% figure.

Yes, I tested the autoplay functionality in free play mode and observed its flaws when it played Spanish BJ. The autoplay cannot do double down rescue either. I played manually with complete perfect strategy using Wizard of Odds chart and the chart on your site.

I understand that due to playing mistakes/choices general payout ratios are usually lower. But it doesn't explain the results in my case in any way.
 
I ran a trial of 500 hands in practice mode with the same bet size as Jufo. My result was about 1.8 SDs below the mean, which is low, but falls under normal variance. Combining this with Jufo's data, results in a lower chance of the combined loss in a random distribution -- ~1 in 40,000

EDIT -- Added more hands & more data
 
Last edited:
I just want to clarify that the results I posted earlier were not just from 32Red but they were consecutive playing sessions at three different Microgaming casinos.
 
I have played probably a million + hands @ MG casinos over the last few hands, anywhere from $2 to $3k a hand.

It would be interesting to find out my numbers. How are you guys getting the data to figure out your return %?
 
I have played probably a million + hands @ MG casinos over the last few hands, anywhere from $2 to $3k a hand.

It would be interesting to find out my numbers. How are you guys getting the data to figure out your return %?

You can copy-paste play histories from Playcheck and add up the numbers OR contact casino support and ask them to send you your play history. When you get your play history, you just add up wagers and returns (this can be done automatically in Excel for example).
 
Exactly the same for me.

Since sommer 08, MG BJ sucks. If you ask me, its rogue !!

It's nearly impossible to win if you have some bonus money in there.

Nearly impossoible results.

Same for All Aces. If you play a bonus, you get 10 times less 4oak.

This changed 4 to 6 months ago. Rogue.
 
I do not use the martingale strategy but I did use it several times the past month. Results? Each time I got loooong losing streaks and a win right when I stopped the progression. To me, this means the A.I. or whatever is immediately counter-acting ANY successive progression betting I use.

... It's a little like upping bets on slot's really, only that you see it more obvious in slot's I guess since compared to BJ which is only double or nothing all the way, slot's are supposedly able to give you large hits!

Just like that counter-acting, you have the same on lots other way around. The best way to get a 100x win, just loose 150x.

Great to see the BJ experiences looses up a little now, if you are looking for streaky BJ for some calculations you should videotape some 2Dice sessions with real money play! do some up and downs there are tape away, I would love to see the results!

Also - it's better to do it now than later, since the systems most likely are tweaked to the better along the way making them more and more impossible to detect comparing to "true RNG" - except the fact we loose, LOL!

Had I only had some spending money on my Visa account I would have done the experiment on BJ for you, however I'm already out for this month in 75% payout already, maby next month I'll put some more money in their pockets!
 

Users who are viewing this thread

Meister Ratings

Back
Top