Overall Best Sets

[peter]

This question is directed to Goswinus, as I consider him a mentor on best overall performing sets.For the Can 649.My question Goswinus is this, what is the best over all set, with the number 1, the number 2 and so forth with all of the numbers.
If for example I chose the19 as my DN, I would want to know what the BOS is. BOS="best overall set".
Thanking you in advance Goswinus, for taking the time to research this.
Kind Regards, Peter.

 

[Goswinus]

Peter,

No trouble at all to help you.
I assume you wanted a set size of 6.
The program is running as I type and the results will be posted shortly.

Regards,
Goswinus

P.S. LT, I really miss the : hattip : smiley...

 

[Goswinus]

Peter,

Here are the BOS results...

01: 01-20-31-43-44-49
02: 01-02-03-27-31-47
03: 03-18-21-31-34-47
04: 04-27-30-34-43-45
05: 05-20-31-34-39-43
06: 01-06-27-31-44-47
07: 07-08-18-31-34-43
08: 08-20-31-39-43-47
09: 09-20-27-42-43-45
10: 10-20-27-31-41-43
11: 11-20-27-30-31-43
12: 12-20-31-39-43-49
13: 13-18-27-31-34-47
14: 14-20-27-31-34-43
15: 08-15-20-31-39-43
16: 16-20-22-27-31-43
17: 17-20-34-38-39-43
18: 03-18-23-31-34-47
19: 19-20-27-32-34-43
20: 09-20-22-27-42-43
21: 20-21-31-34-43-47
22: 20-22-27-31-42-43
23: 01-03-23-31-34-47
24: 20-24-26-34-41-43
25: 20-25-27-34-42-43
26: 20-26-34-41-43-47
27: 04-20-27-30-43-44
28: 11-20-27-28-30-43
29: 18-27-29-31-34-47
30: 20-27-30-34-43-45
31: 08-20-31-39-43-49
32: 17-20-32-39-43-44
33: 20-30-31-33-43-49
34: 03-18-27-31-34-47
35: 20-26-34-35-43-47
36: 20-31-34-36-43-44
37: 20-27-30-37-40-43
38: 20-32-34-38-43-47
39: 17-20-30-39-43-44
40: 20-27-38-40-43-45
41: 20-27-30-34-41-43
42: 20-27-31-42-43-44
43: 12-20-31-39-43-49
44: 20-27-31-39-43-44
45: 20-27-34-43-45-47
46: 20-31-39-43-46-49
47: 20-31-32-34-43-47
48: 05-20-31-34-43-48
49: 20-30-31-39-43-49

I also checked the history to see what would have happened when all these combinations where played in every draw:

0+0: 35,594
0+1: 06,096
1+0: 34,829
1+1: 04,481
2+0: 13,286
2+1: 01,675
3+0: 02,400
3+1: 000201
4+0: 000155
4+1: 000013
5+0: 000005

Not to impressive as you can see.

I hope this helps.

Regards,
Goswinus

 

[peter]

A great big thankyou Goswinus.

 

[Maggie]

Anyone else out there with Sympatico High Speed that's slower than molasses today?? It's driving me nutty.

 

[Beaker]

Anyone else out there with Sympatico High Speed that's slower than molasses today?? It's driving me nutty.

Very slow indeed Maggie. I think it might just be this site

 

[Maggie]

I thought of that, but wasn't too sure if it could be because of the site or not.

Beaker, don't wanna give away any secrets or anything, but, noticed you removed your location.. Witness Protection Program??? Or just trying to be incognito??

 

[Beaker]

I thought of that, but wasn't too sure if it could be because of the site or not.

Beaker, don't wanna give away any secrets or anything, but, noticed you removed your location.. Witness Protection Program??? Or just trying to be incognito??

Preparing ... for the BIG win and I don't want my 'friends' to know where I am

 

[peter]

Anyone else out there with Sympatico High Speed that's slower than molasses today?? It's driving me nutty.

I have Telus High speed velocity, which was crawling slower than a snail.

 

[doc]

The sets provided here are interesting at first glance, but I have to ask a few questions. If I understand this correctly, the table is showing the <<ONE>> best performing line for each of the 49 numbers if they were used as a DN and wheeled with all of the remaining 48 numbers. If I am wrong on this premise, then everything that follows will also be wrong.

At the time of this writing, there have been 2015 draws for the CDN649 (with 2016 coming up fast!). I am assuming that some kind of test was run against the entire draw history in order to arrive at the results given.

Now, since each number used as a DN and wheeled with all remaining numbers produces 1,712,304 lines, I have to ask how you arrived at the conclusion that the lines you show have performed better than any of the other 1,712,303 lines that also contain the same DN?

Let's use the number 31 as an example. This is the number that has been drawn the most times at 293 (not incl. bonus). It should be immediately clear that since 31 has appeared in 293 draws, there are at least 293 lines that have performed better than the line you provide. When I test the line you provided, it has a best of only 2 match 4's and 56 match 3's over the entire draw history.

Here's another line: the numbers 5-16-17-31-38-45 produced 1 match 5, 5 match 4's and 39 match 3's before finally hitting the JP at draw 1950 on 9/28/02. There are other examples for each of the 49 numbers, but this post is going to be long enough without going into all that.

Based on that, I'm sorry but I have to be reasonably skeptical about the validity of the rest of your data. As a matter of fact, I ran the 49 lines you provided against the 2015 draw history and I come up with very different results as shown in the report below. The only count where our figures match is for the match 5. The software I use only counts a bonus hit in the one place that it belongs. You have to have at least 5 numbers correct in order for a bonus match to mean anything at all. The rest are meaningless, so why count them? But that's another story in itself.

*************************************************************
Lottery: CA 649
Wheel File: Bos.txt
Testing 49 lines against 2015 draws

Match-5 for Draw 193 11/27/1985 At Line 24
Match-5 for Draw 245 5/28/1986 At Line 3
Match-5 for Draw 887 7/22/1992 At Line 16
Match-5 for Draw 1540 10/24/1998 At Line 19
Match-5 for Draw 1725 8/2/2000 At Line 2

The following line numbers had 5 or more matches:
2
3
16
19
24

Match 0 - 41,690 Expected is 43,045.000
Match 1 - 39,310 Expected is 40,779.480
Match 2 - 14,961 Expected is 13,070.340

Match 3 - 2,601 Expected is 1,742.713 Score is 147,362.100
Match 4 - 168 Expected is 95.637 Score is 173,442.700
Match 5 - 5 Expected is 1.779 Score is 277,456.700
Match 5+ - 0 Expected is 0.042 Score is 0.000
Match 6 - 0 Expected is 0.007 Score is 0
Total Plays - 98,735
Total Losses - 95,961
Total Wins - 2,774

Expected Total Score Is 301,084.900
Actual Total Score Is 598,261.400
This Wheel Scored 198.702% Of Random Expectation
Total Draws With At Least 1 Match3 - 737
*************************************************************

I don't know how such major discrepancies can be reconciled, but I would be willing to bet the farm that the above counts are accurate. I have tested this software umpteen different ways to verify its accuracy and it has been bang on every time.

In conclusion, for Goswinus, Peter and everyone else who reads this thread, please understand that I am not trying to step on anyone's toes, or come across like Mr. Know-it-all or cause trouble with this. I know what I know and I make no apologies for that. I just hate to see bad or unsubstantiated data being passed around.

Goswinus---perhaps a bit of an explanation of how you arrived at your data would help everyone to understand this better. Please don't take offense to any of my remarks or questions.

 

[Beaker]

Don't worry doc, Goswinus will clarify for you. Anyone that posts data can expect to have it challenged - It is not meant as a slight but as you say we want to be accurate here for the benefit of everyone.

But, I'll make a comment on the inclusion of the bonus number.

It is irrelavant to me that the bonus only applies to one of prize tiers.

7 balls come out of that machine and I use all of them for analysis - for me this is a 7 number lotto.

I think looking at both 6 and 7 number analysis is beneficial. To disregard that bonus may remove some crucial pieces of the puzzle.

Just a thought.

I didn't see your picks but good luck if you are playing tonite.

 

[Goswinus]

Hi Doc,

First of all, I'd like to thank you for your valuable comments.

The idea of BOS is nothing new.
There are two issues regarding 'best' set.
1) Testing all combinations takes a LONG time
2) How do you define BEST

As to issue 1, I have solved this by selecting random sets.
In this case, for number 1, I tested 124865 random sets.
For number 2, I tested another 124865 random sets.
And so on.

As to issue 2, I took the liberty to determine BEST as follows:
I give points for each hit, starting with 1 for a 3 hit, 2 for a 4 hit, 3 for a 5 hit and 4 for a 6 hit. So there's no distinction between 3+0 and 3+1 hits, because not everyone likes the bonus.
You may think that 4 points for a jackpot is far too low, but if I award 100 points, the best sets would only show the 2015 draws already drawn.
I posted a question for better suggestions a long time ago (something like a year ago).

Actually, there is another issue: It's all testing after the fact.
How a set has performed in the last 2015 draws has got little to do with how well it's going to perform in the next draws.
But let's forget about this for now.

As for the number-31-example:
You state "It should be immediately clear that since 31 has appeared in 293 draws, there are at least 293 lines that have performed better than the line you provide".
You would be right if 31 had occurred with (at least two of) the same numbers every time it comes out.
But since it comes out with other numbers each time, you're not.

With the current scoring system, "your" line has a 'score' of 58, whereas "mine" has 60.
Again, this all depends on the scoring system.

Now about your farm:
You can keep it, 'cause your program is absolutely great!
In fact, so is mine.
The results you show provide are exactly the same, so we have cross-checked our software.

I say: 3+0=2,400 and 3+1=201.
You say: Match 3=2,601.

It may look like a coincidence to you that 2,400+201=2,601 but I ensure you that it isn't. I'll tell you why...

If you hit 3 and the bonus, but you don't care about the bonus, you still would have hit the 3, right?
I don't think that there's a special rule in 49/6 that if you hit the bonus number all other hits don't count?

My program shows the bonus number because in the Dutch lottery we get different prizes when we hit the bonus number too.

Kind regards,
Goswinus

P.S. My toes are just fine!

 

[doc]

Thanks for your reply Goswinus. I will respond at greater length later. Pressed for time right now and gotta go.

 

[doc]

Hi Goswinus,

Thank you for your explanation and being gracious in your reply.

You are quite right on a number of points. It would indeed take an unacceptable amount of time to test ALL combinations. I estimate that a leading edge system with a 2-3ghz processor and lots of fast memory would take anywhere from 48-100 hours of continuous operation. Some kind of ingenious testing algorithm could reduce that considerably, but short of that, some kind of work-around is required.

The solution you describe is certainly one way to approach the problem. I find though, that using random sequences can produce very skewed results unless you run the same test several times to get some kind of median values. Otherwise, the results can be dangerously misleading or even downright unreliable. But doing that leads right back to the time issue and the results would still be open to question.

I really think that there is a simple answer that quickly solves the whole problem and it is this: there is no such thing as a best set. I know a lot of people have difficulty in accepting that because I used to be one of them for many years.

You hit the nail on the head yourself by pointing out that past performance has little to do with future performance. This is because the past data is always changing. As soon as analysis can identify some set of numbers or some other trend, its usually too late because some new and sadly unknown trend is already beginning to emerge, and the old one is already beginning to vanish. This is especially true when trying to pick specific individual numbers.

The only reasonably predictable aspect of the lottery that I have found so far is what someone here recently started calling "fingerprinting" ---I believe it was Peter. I have seen it called a "footprint" or "profile" elsewhere. Each set of winning numbers has certain characteristics such as the sum of the line, odd-even split, hi-lo split, decades, consecutives and so on. It is relatively easy to identify long-term trends with this method. Most decent lottery software or even an Excel spreadsheet will allow the player to filter their set of chosen numbers and only produce combinations that have a certain profile. The profile that I commonly use reduces the total number of combinations to 6,902,082 or about 49.4% of the whole total. Bear in mind that this is with all 49 numbers in play. There have been 40 draws so far this year. My profile has been correct 24 times or 60% of the time. Not a bad tradeoff, but of course this has nothing to do with the actual numbers that I selected for one reason or another. What it really means is that I have effectively been playing at less than half of the overall jackpot odds for 6 out of every 10 draws over an extended period of time.

This leaves little doubt in my mind that chasing short-term trends for individual numbers is mostly a fruitless and frustrating pursuit, while playing steady (and easily identified) long-term trends will inevitably lead to success. Of course it would be easy to argue this from the opposite viewpoint and I'm sure that someone will. In the end, it all comes down to personal preference.

Regards, Doc

 

[Dennis Bassboss]

Great to see people learning here as Doc is...But I think that Doc has much more to learn so I'll let him work a while before answering his toughts....While I was guiding a group of people for fishing yesterday I had this reflection....In the lake you know that all the fish are in there...Staying at one spot..At one point one fish might pass by and you might catch it...That's the long term strategy...But since the people were there only for a short while it was better to reject that strategy...
Another strategy was to go after the fish where they were located (short term strategy)....I did this found the fish and there was so many trouts out there that you could choose to keep only the big ones...caught two 8 pounders in the process and we only kept the ones over 5 pounds...Now that is good fishing... And a lot of people were ecstatics...
It is the same with lotto...you can fish for numbers that will come up one day...you know that all the numbers are there like the fish in the lake...Or you can go after the numbers where they are at a certain time.... tomorrow they most likely won't be there anymore as they move just like fish to different locations in time ....But next year when the conditions will ressemble the actual conditions...You might very well see some of the same fish or numbers in the very same area again... It is a cycle that repeat
itself over and over again like it or not....But if you don't look at every details even how insignificant it might seem (bonus number or wheather conditions) you might miss target and fish or play in a wrong area..
Saying that short term playing is not good shows to me that the person saying this as much more to learn....

 

[doc]

Yes, all of us will always have a lot to learn DB. I hope you haven't stopped including yourself among the eligible.

Just as a friendly reminder, I may be relatively new on the scene here, but I have been at this game for over 20 years now. I started visiting the LT site shortly after it first appeared on the web. I even registered in this forum about a week after it first became available. However, several months went by and there was hardly ever any kind of activity---or even other members. So I stopped visiting and eventually totally forgot about it. I only came across it again quite by accident a few months ago and was surprised and delighted to see how big and successful it has become. That being said, I don't expect anyone to bow from the ankles in my presence.

So, I would appreciate it if you don't talk down to me from whatever mountain you think you're on, and I will be only too glad to extend the same courtesy and respect to you.

On a lighter note, from your description above, I would love to go fishing with you someday. Maybe when we're both wealthy and retired we can do just that. Gotta fish out those numbers first though!

My Best Regards, Doc

 

[Dennis Bassboss]

I'm glad to hear that you will use the bonus number from now on...Way to go!!!

 

[peter]

The comparison of the fish and the lotto numbers was very good.
One thing you can say about native peoples, is they study and make observations about their surroundings.