Your personal Lotto quick pick

Wanna improve your chances on winning a few million rupees?

The Mauritian Lotto was first introduced on the 07th of November 2009 by lottotech. The first jackpot was Rs 12 Million. The first winner of the lotto was Guillyano Zamir, who won the Rs 30 Million Jackpot. I am going to post the source code of a program i made that generate an arbitrary number of Lotto simulations. While writing this blog post, i researched a bit on Lotto and found some very interesting stuff.

Lotto lets you pick a certain number of choices from a card. In Mauritius you get to pick 6 numbers from a card having 40 numbers. Then, the Lotto managers pick 6 numbers at random. If your choice exactly matches theirs, you win a few million rupees or so. What is the probability of having a winning pick? Mathematics says the following about it: (40 × 39 × 38 × 37 × 36 × 35) / (6!) = 3,838,380. You have 1 chance in 3,838,380 of picking all 6 numbers correct! Actually it is not so huge. Some other countries or states have probability of over 20 million to 1 in winning their Lotto game. But this also equates to more frequent winners and lesser jackpot stakes.

There exists techniques and strategies for playing Lotto. The strategies are based on mathematics and logic. With techniques such as tracking, wheeling and pooling, you may be able to improve your odds. (Click on the hypertext link to read more about them).

How to Win the Lotto?
Basically, you can’t. Random draw lotteries are VERY carefully designed not to be predictable. And any ‘system’ is basically claiming to be able to predict results. Well best option is to gather a group of 10 people or so and play every combinations possible (3,838,380) then you split the gains. Lol, even if you do this, you are not sure you will be able to make a profit margin. I found some ‘systems’ on the internet claiming they improve your chances, significantly. Check out the following links: 1 | 2 | 3 | 4 | 5.

At Last: Your personal lotto quick pick
That was my primary aim for this post. A program that can generate your own-tweaked personal quick pick. I used Python to write it. It is not difficult to code. Yes, a function that successfully predicts the winning picks themselves would be more useful, but Python, although powerful, has yet to implement psychic faculties. Note that there exist a lot of software (some are shareware actually) that enable you create your own strategy and then generate numbers. They also enable you to analyze the numbers and derive conclusions of all sorts. Here are a few:

• Odd/Even Analysis, which is where you determine the frequency of odd or even numbers.
• Pairs/Doubles Analysis, which is determining the frequency of certain numbers appearing together.
• Pick 3 and Pick 4 positions, such as in which position–first, second, or third–the digit 5 most frequently hits.
• Sector Analysis, which is tracking how frequently numbers in the low sector hit, how many in the midrange sector hit, and how many in the higher range hit. For instance, if you’re playing a 6/40 Lotto and believe the numbers will be midrange, you might choose 14, 16, 17, 19, 22, and 28.

Here i made a random generator program and that counts the number of times there were 3 even numbers and 3 odd numbers in an arbitrary number of simulations. And it gives you its percentage. In my program i made 1000 simulations. You can change this of course to your own. Actually this was a practical test and had 2 hours to complete it! 2hrs may seemed a lot of time but actually it was in the first year, first semester. It was by far the coolest test i did so far in my course! Very few (actually i don’t know anyone) fininshed it completely without any flaws. I took my entire 2 hrs to do it. I did a mistake though, but a honest one, i wrongly used the random.randrange() library function, it just never generated the number 40 in the simulations.

You will need to generate random numbers for this. Generating true random numbers is a very complicated process and is an area of active researchin computer science. Python has pseudo-random number generator. It is called pseudo-random because it make use of an algorithm (MersenneTwister algorithm for that matter) to generate them. Well, you can make your own ;). You can use the random.randrange([start], stop[, step]) which return a randomly selected element from range(start, stop, step). Note that stop is non-inclusive. So we are ready to get our hands dirty.

Here is the Program listing

# roshans89

simulations = 1000 # Number Of simulations. Change only this variable.

from random import randrange

def simulate():

    # Declare all variables
    generated_number = []
    count = 0
    c_odd,c_even = 0,0

    while True:
        if count == 6: break

        a = randrange(1,41)

        if (a in generated_number):
            continue # Skip rest of loop and loop again from start

        generated_number.append(a)

        if a % 2 == 0:
            c_even += 1
        else:
            c_odd += 1

        if c_even == c_odd: #3 times
            num_times = 1
        else:
            num_times = 0
        
        count += 1


    return generated_number, c_even, c_odd, num_times

# Program start executing here

print "game\tWinning Numbers"
print "----\t---------------"

count = 1
times = 0

while count <= (simulations):
    r, c_even, c_odd, num_times = simulate()
    print count,"\t%d, %d, %d, %d, %d, %d" % (r[0],r[1],r[2],r[3],r[4],r[5])

    if num_times == 1:
        times += 1

    count += 1

print """
Number of times there were 3 odds and 3 evens: %d
Percentage Of time there were 3 odds and 3 evens: %.02f%s
""" % (times, (times/float(simulations) * 100), "%")

Here’s a sample run

game	Winning Numbers
----	---------------
1 	 11, 28, 39, 17, 4, 20
2 	 39, 6, 15, 34, 8, 31
3 	 38, 2, 12, 27, 30, 15
4 	 8, 16, 6, 18, 34, 40
5 	 21, 19, 14, 2, 11, 34
6 	 8, 21, 25, 5, 39, 23
7 	 5, 30, 16, 27, 10, 26
8 	 20, 35, 7, 31, 27, 29
9 	 4, 10, 2, 28, 3, 37
10 	 34, 3, 19, 22, 8, 11
.
.
.
997 	37, 20, 10, 35, 21, 15
998 	29, 27, 9, 7, 13, 20
999 	1, 13, 19, 5, 21, 32
1000 	25, 3, 7, 14, 34, 33

Number of times there were 3 odds and 3 evens: 345
Percentage Of time there were 3 odds and 3 evens: 34.50%

Program Notes:
You can change the simulations variable to your liking of course. Note that it took about 8 seconds to generate the output in my Intel 2.76 core 2 duo. I am sure there are faster algorithms to generate this. But performance is not an issue here 🙂 . The tricky part of the program is [obviously] you can’t generate the same number more that one time in a simulation. To resolve this, i generated a number and store it in an array (actually in Python it is list not array). Then i generate another number and check if it is in the list. If it not already in the list i can safely append in to the list else i ignore it and generate another number and then append it if it is not already in the list. This is what the following does

a = randrange(1,41)

if (a in generated_number):
continue # Skip rest of loop and loop again from start

generated_number.append(a)

Then i check all the numbers that i appended if it is even or odd and keep a count. If even == odd, i.e., got 3 even number and 3 odd number then i add it to the count of a variable which keeps the number of evens = odds (that variable is called times in my program). Finally i printed the % Of time there were 3 odds and 3 evens. It is just times/simulations * 100.

Notice that it will always display the percentage between 30% and 36%, the most frequent one being 33%. There could be some outliers, but i never actually encountered one. This percentage figure is questionable. Because as i said above Python uses the MersenneTwister algorithm to generate a random number. It is not a True random number.

By the way here are the previous lottotech results if want to make your own statistics and predictions:

Num	Draw Date	Winning Numbers	Order Drawn
034	26-Jun-2010	21-25-28-33-35-40	25-33-21-28-40-35
033	19-Jun-2010	07-14-27-28-38-40	14-07-38-28-27-40
032	12-Jun-2010	01-09-23-35-36-39	01-35-39-09-36-23
031	05-Jun-2010	01-07-15-21-23-27	23-27-21-07-01-15
030	29-May-2010	06-21-25-27-38-39	21-06-27-25-39-38
029	22-May-2010	11-14-21-23-31-35	31-11-14-21-35-23
028	15-May-2010	01-06-13-16-20-26	16-01-26-13-20-06
027	08-May-2010	06-08-23-30-32-33	06-33-32-23-08-30
026	01-May-2010	03-04-14-23-25-28	23-04-25-14-28-03
025	24-Apr-2010	02-05-26-31-36-38	31-05-02-38-26-36
024	17-Apr-2010	08-14-21-26-34-36	34-21-14-08-26-36
023	10-Apr-2010	01-11-16-17-18-31	11-16-18-17-31-01
022	03-Apr-2010	03-11-19-20-30-32	19-30-20-03-11-32
021	27-Mar-2010	06-11-12-27-39-40	39-12-11-06-40-27
020	20-Mar-2010	01-09-11-26-34-35	34-01-11-09-26-35
019	13-Mar-2010	02-04-06-09-19-29
018	06-Mar-2010	23-29-31-32-35-40	35-32-23-40-31-29
017	27-Feb-2010	05-08-13-23-26-36	05-08-23-36-26-13
016	20-Feb-2010	07-11-24-25-31-39	39-25-07-24-11-31
015	13-Feb-2010	01-06-12-14-18-40	14-18-06-01-12-40
014	06-Feb-2010	08-15-16-17-20-26	26-08-20-16-17-15
013	30-Jan-2010	08-13-17-23-25-38	38-25-17-23-13-08
012	23-Jan-2010	04-09-12-19-30-40	40-19-04-12-30-09
011	16-Jan-2010	12-15-28-30-34-39	39-12-34-28-30-15
010	09-Jan-2010	08-10-24-26-27-36
009	02-Jan-2010	07-17-27-29-34-37
008	26-Dec-2009	07-08-09-15-18-30
007	19-Dec-2009	01-07-25-31-38-40
006	12-Dec-2009	26-27-29-30-36-39
005	05-Dec-2009	05-10-14-26-37-39
004	28-Nov-2009	03-11-12-26-32-38
003	21-Nov-2009	08-15-21-22-28-31
002	14-Nov-2009	02-03-04-11-16-34
001	07-Nov-2009	07-10-20-21-28-34

Data obtained from this link

Note that you are free to modify my program or distribute it, but please give me due credit.

Hope you enjoyed this post, replies as always are most welcomed. 🙂

–roshans89