This Project uses Python 3+, Selenium, and Beautiful Soup

DISCLAIMER: This project does not promote Keno gambling of any kind. It is merely a study project.

This Project is used to study MA Keno and see if it is worth playing and spending money on. Of course the odds should be against you, but I'd like to know some things:
1) what's the best number of picks to take below 5 (I assume anything beyond that takes pure luck or foresight to win)
2) is there a way to predict some numbers based on the ones already played

Future Materials:
1) Automate some data collection to collect a repo of data (Need more HD space)
2) Create a detailed analysis of the collected data and have it in a more presentable form (charts, tables, stuff like that)
3) Adjust terminal output to print out appropriate/desired output

=======================================================
Below is some very light analysis based on a single days worth of draws..............

For my sample testing, what I do is run my analysis at draw number 200 on 7-14-2016 data. So you'd theoretically have bought a ticket for draw 201 that day, up to 230 (30 draws per ticket is the maximum)

For counting wins, I consider it a win if the combination of numbers matches any prize for the pick.

Matches only to be included if it has a payout

For formatting, it'll look something like this:
NOTE: Below output is all manually typed in- future stuff is to modify output when I find the time

11/15, (6 m1, 4 m2) 6f, 18g, 2p

This reads:
-Won 11 out of 15 draws
-Match 1: 6 times
-Match 2: 4 times
-All numbers show up about 6/15 times (f=frequency)
-Won 18 dollars (g=gain)
-Profit of 2 dollars


******/Pick 1/******
Match 1: $2.50

Picking one number is just not good as even with the most optimal pick from a sample of a day's worth of draws, you only get back about roughly $5 for any number of draws

>win 4/5, make 10 (profit 5)
>win 6/10, make 15 (profit 5)
>win 8/15, make 20 (profit 5)
>win 10/20, make 25 (profit 5)
>win 11/25, make 27.5 (profit 2.5)
>win 13/30, make 32.5 (profit 2.5)


******/Pick 2/******
Match 1 (M1): $1
Match 2 (M2): $5

For this testing, I looked through the keno draws for the top two numbers that showed up. I used those 2 numbers as my most optimal pick to get me the most return. Here is a sample return:

>5/5,[(1, 4), (2, 1)],f1,g9,p4
>8/10,[(0, 2), (1, 4), (2, 4)],f2,g24,p14
>10/15,[(0, 5), (1, 6), (2, 4)],f3,g26,p11
>12/20,[(0, 8), (1, 7), (2, 5)],f5,g32,p12 (r=17)
>14/25,[(0, 11), (1, 10), (2, 4)],f6,g30,p5
>21/30,[(0, 9), (1, 19), (2, 2)],f7,29,-p1

From this sample, too few and too many draws are not that profitable. However, picking something in the middle yields pretty reasonable gains for risk. However, since the average number shows up only 5 or 6 times out of 20 and 25, respectively, there's still only ~20% chance of getting profit from a pick of any 2.

******/Pick 3/******
Match 2: $2.5
Match 3: $25

>1/5,[(1, 2), (2, 2), (3, 1)],f1,g30,p25
>6/10,[(0, 1), (1, 3), (2, 4), (3, 2)],f2,g60,p50
>6/15,[(0, 2), (1, 7), (2, 4), (3, 2)],f3,g60,p45
>8/20,[(0, 4), (1, 8), (2, 7), (3, 1)],f5,g42.5,p22.5 (r=10)
>9/25,[(0, 8), (1, 8), (2, 8), (3, 1)],f6,g45,p20
>8/30,[(0, 4), (1, 18), (2, 8)],f7,g20,-p10

With the reward for getting 3 matches, if you can get them all right, this is a pretty good bet, BUT the odds are heavily against you and not for any high drawings. You only have a 20% chance of getting your numbers if picked randomly and getting all three does not seem likely.


******/Pick 4/******
Match 2: $1
Match 3: $4
Match 4: $100

>3/5,[(1, 1), (2, 1), (3, 3)],f1,g13,p8
>5/10,[(1, 3), (2, 2), (3, 5)],f2,g21,p11
>10/15,[(0, 2), (1, 3), (2, 7), (3, 2), (4, 1)],f3,g115,p100
>12/20,[(0, 2), (1, 6), (2, 10), (3, 1), (4, 1)],f5,g114,p94
>12/25,[(0, 3), (1, 10), (2, 10), (3, 2)],f6,g18,-p7
>13/30,[(0, 1), (1, 16), (2, 11), (3, 2)],f7,g19,-p11

Getting 4 matches is very good, but unfortunately they do not show up often. You can get matches of of 2's and sometimes 3's, but getting 4's is very improbable. For a bet of 4, you need to get the 4 matches to make any reasonable profit.