Baccarat 5 column statistics

Hello to anyone still interested:

Brief recap.

In this thread, OP Violater777 described a method whereby he groups all shoe hands into columns of 5. Then he ran some stats looking for betting triggers which gave an advantage over random betting. He reported those triggers at the start of this thread. I was curious so I wrote a program which duplicated his approach and subsequently analyzed whether some of his triggers, in fact, gave an advantage using the first 600 Zumma shoes. I have been posting the results piecemeal because it take time to input the results and record the outcomes.
Nevertheless, I have now completed inputting the 600 shoes (which represent roughly 42,000 hands). Here are the results for his best reported triggers over the entire 600 Zumma shoes (my program's win % finding in brackets)

Start of a column with B and betting B as the next result: 2099W and 2026L (50.9% win)
Start of a column with BB and betting P as the next result: 1076W and 1008L (51.6% win)
Start of a column with BP and betting B next result: 1051W and 958L (52.3% win)
Start column with BBB and bet B next: 533W and 465L (53.4% win)
Start column with BBBP and bet P next: 227W and 232L (49.5% win)
Start column with BPBP and bet P next: 243W and 241L (50.2% win)
B Lapper for rows 3 and 4 (3 and 4 hole): 1869W and 1750L (51.6% win)
Total of the above: W (7098) and L (6680) for a 51.5% win rate (expected 50% win rate).

Some of the above bets may be worth pursuing as advantage bets once you group the hands into 5 handed columns as V777 has suggested.

Thank you. The study is complete from my end. Perhaps these numbers would weaken or strengthen given even more shoe input but I only had 600 shoes. For those with better simulation programs please group the shoe hands in columns of 5 and then do your analysis to be fair to how V777 described his method.

 

can anyone share their experiences with it and how it has been going

 

Mansi:

I was at the casino last night standing behind a bac table just quietly recording the results from a shoe and using V777's approach. It was one of "those" shoes where the entire table was losing hand after hand. I wasn't doing too bad on paper in comparison. A younger guy clearly new to baccarat had lost most of the last 7-8 hands. He turned to me and asked what I was writing. I explained briefly. He stopped betting and asked me to tell him when to bet. I waited until one of the triggers came up (a bank lapper in row 3) and let him know. The younger guy and his friend both bet banker. Banker won. He asked for me to continue to advise him. So I waited for the next trigger etc. Long story short, he won 5 out of the next 7 bets. The rest of the table? Not so successful. Take that for what it's worth. Anecdotal.
I have written another program where I combine the best 4 bets of V777 and added one more trigger from another source which a different program showed an advantage. I'm going to run 1000 shoes through this omnibus program and see how it does. May take a while but it seems V777 may have something here. I emphasize may.

 

That`s impressive that so many triggers in 1 shoe considering it takes lot of amount time to write it all down. But why with that kind of system testing in casino were you can save life years of time shuffling yourself at website. Casino is more like playing around emotions and entertainment, but thats another topic already...

 

Wondering if you did (or could) show the W/L registry in terms of W and L IAR. That would be helpful showing interested parties what they're up against even with a positive EV. Assuming 51.5% holds up I would expect a pretty severe variance requiring a betting strategy to overcome same.

Thanks

 

I am finding that in a shoe of some 70 hands; I would estimate that using the 4 best bets of V777, you will see between 12-15 triggers come up every shoe. So, you're mostly waiting on average about 6 hands to make one bet. But so far, I've been seeing an advantage with those bets. As I mentioned, I have added one trigger that V777 didn't mention (it is one of Tabone's triggers which shows a slight advantage of 50.6%). When you add that one in, you're getting 15 or more solid bets a shoe.

 

Jimske:

Yup, you and me both. I originally wrote the program just to see if V777's triggers worked. Four of them did at least over 42,000 hands. I've spent the last few weeks rewriting the program to include one of the Tabone bets and then to track the sequence of W and L. It hasn't been easy. You'd think coding for a new trigger would be quick but it introduces some difficulty which took me 3 days to overcome. I have also added a filter to the results which records the results separately to compare the filter and non-filtered results. It's just about ready. I'm working on getting the last few coding bugs out but it is almost ready to do those calculations. Initial results show LIAR so far have been 4. Let me get this thing up and running and I'll report back with the results. I'm a university prof and it's finals time so give me a week or two. Work calls.

 

Chew,
I'm not following the bets in this thread. But I like to see stuff programmed and analyzed. 42,000 hands? Or 42,000 bets? Max of 4LIAR would be remarkable. Another stat which would be important is average bets per shoe (ABS). No rush.

Also, I have thousands of shoes. If you want another 600 for testing let me know.

 

Jimske:

Well finals are over and I have a bit of time. Yes please I'd like another 600 shoes as long as they don't record ties. My program isn't built for that.
I also have the results of the first 300 Zumma shoes as a PDF. I will explain in a later reply.

 

OK I have a bit of time now. I've run the first 300 Zumma shoes through the new program (now called Crazy Ivan) which uses V777's 4 best triggers and one of Tabone's triggers. The program also analyzed FTL (follow the last) and another nonsense approach out there called Base. I have no confidence in either of them so they will be my controls. Since neither of them should work, I am expecting no advantage. I included them just to see if the first 300 shoes could produce a 50:50 result with approaches that shouldn't work.

The results. Indeed, neither FTL nor the base showed any advantage (see PDF. %W in green, %L in red in the middle and right sections). So we essentially got 50:50 win : loss for both of them. No surprise.

More interestingly, Crazy Ivan showed a 51.9% win rate over 5168 bets. That's not bad. The average number of times you will bet in a 70 hand shoe turns out to be 17 bets. So, on average, we will be waiting 3 hands making no bets and then make a bet on the fourth. Acceptable. You will notice that the win loss sequence is also recorded for each shoe. Notice that multiple losses cluster in large continuous sets. We could take advantage of that with a rule to stop betting after 2 losses in a row and resume after a win. Something like that.

Is there anyone out there who would like to take a stab at making a money management system for these results? One that makes a profit would be nice.

I will input the remaining 300 Zumma shoes and report on the results. Give it a few days or so. I am encouraged by these early results and hope for roughly similar stats for the next 300 shoes. The program has changed from the initial one so the results may vary from the initial report. We'll see.

 

To anyone still listening:

I have completed the analysis of the Zumma 600 shoes using Crazy Ivan and the 2 control approaches FTL and Base. See attached PDF entitled Crazy Ivan Combo Complete showing the three methods in different colours.

The FTL strategy (brownish colour) turned out to be as expected. 40,016 bets made. 19,930 wins and 20,086 losses. 50:50. No advantage.
The Base strategy (salmon colour) turned out to be as expected. 8333 bets. 4174 wins and 4159 losses. 50:50. No advantage at all. I am happy to see both useless approaches generating random noise (50:50). This gives us confidence that the 600 Zumma shoes can reveal nonsense approaches for what they are. And... should give us confidence that approaches which do not generate random noise may be worth looking into. With that segue....

Crazy Ivan strategy. 10,525 bets. 5,456 wins and 5069 losses. 51.8 : 48.2. Yes. A 1.8% advantage over the expected 50% win rate which is more than blackjack counters "enjoy". The average number of bets per 70 hand shoe worked out to 17.5 (so 1 bet for every 5 hands). The attached PDF also shows the Win-Loss sequence for every shoe. Something of interest is that the losses cluster into sets. We should be able to lever this into even more of an advantage by following a rule which says to stop betting after 2 losses and resume upon the next win. (perhaps an even better rule could be formulated?) In this manner, groups of 5,6 or even 7 losses in a row will pass us by without consequence as we are not betting. I would welcome the opinions of other more experienced Baccarat players concerning these results and how to use them to generate profit. I would especially encourage others to come up with and share a money management system on this forum to generate said profits.

Thank you to anyone who has followed this analysis. I hope this can be developed further by others to either show it as an advantageous approach to winning at Baccarat or a deceptive blip of positive results over 600 shoes. But having said that, the FTL and Base approach were revealed as useless approaches. Crazy Ivan? Somehow gave a positive result.

 

Is Ivan posting at one or two topics here at the baccarat forums ? $ 500 chips and WG $ 1,500 - $ 2,500 per shoe , 3 shoes per fay ?


ND

 

Let me know if you want more shoes. I can send you a few thousand shufflemaster shoes in easy to work format.

 

This may be greatest post in this forum ever. Great results with clear explanation.
Progression seems very risky, but waiting 2 losses and then jump into train.?= it may took some time to let eye get over the list and see does it bring more profits.
Thanks

 

more the better

 

is this the your betting method with 1 3, 5 progression sequence which you win 53% of the time?

 

It's best to take any formation and evaluate it with the whole baccarat shoe and most recent play in mind.

Start column with BBB and bet B next: 533W and 465L (53.4% win) will work very well on a neutral to strong Bank shoe. If there is a strong Player shoe then BBB is more likely to get rolled. CONTEXT.

 

Live shoes? I would love to have a copy for testing purposes. I will take all you have. Let me know.

 

I am posting these results as Chewbaccarat. I've entitled the program as Crazy Ivan after the Hunt for Red October. No idea who the poster Ivan is.