Lottery with a rules flaw

One of the Lottery's available in my country has a flaw in deciding winning nubers which leads to a skewing of the distribution of winning numbers. Numbers are selected on the basis of score draws however if insufficent or too many numbers qualify, then the numbers are selected from the "highest number to score draw" downwards. This means that their is a bias in favour of high numbers being selected and was bourne out by an analysis of lifetime results of draws to date. The frequency is :
No Times Rank
1 136 38
2 139 37
3 143 35
4 150 34
5 140 36
6 164 33
7 167 32
8 179 29
9 169 31
10 185 27
11 174 30
12 192 25
13 181 28
14 191 26
15 201 23
16 213 20
17 226 17
18 212 21
19 208 22
20 239 14
21 196 24
22 240 13
23 227 16
24 225 18
25 221 19
26 256 9
27 238 15
28 261 5
29 259 8
30 287 1
31 264 4
32 254 10
33 260 6
34 252 11
35 276 2
36 260 6
37 269 3
38 252 11

If you graph the above you will get an even clearer picture of the bias and its solely down the rules they adopt to settle winning number selection when this is required to be invoked. My question is "How to profit from it". Now given the take from total ticket sales is quite high it maybe that even with this percieved advantage it might not be possible to exploit the bias sufficiently to overcome the -EV. Does anyone know how to calculate this or can recommend a strategy?

Quote: ajindarwin

Numbers are selected on the basis of score draws however if insufficent or too many numbers qualify, then the numbers are selected from the "highest number to score draw" downwards.

Can you help me understand this? Maybe an example would help, or a link to the official rules.

Quote: Wizard

Can you help me understand this? Maybe an example would help, or a link to the official rules.

The last line on the link is the relevant one.

(6) If, having been ranked in accordance with this rule, 2 or more matches are ranked equally, then those matches are to be ranked in descending order according to the numbers corresponding to them in the match list (ie. the higher the number on the match list, the higher the ranking).

This is a soccer pool where you are trying to pick games that end in a scored tie (or draw as they say in the UK/Austrailia). So based on the teams playing, I don't think you can find patterns in what numbers have won in the past....it will depend on which teams are matched with which numbers, correct?

I guess what I am trying to say is that this isn't a ping pong type winning lottery number system, this is a sports betting pool.

Quote: Wizard

Can you help me understand this? Maybe an example would help, or a link to the official rules.

I think he's talking about the Australian Pools system. The way I read it, it works like this: it's a 6/38 Lotto, but instead of the numbers being random, there are 38 soccer matches, which are ranked as follows:
1. Draws, ranked in decreasing order of goals scored (e.g. all 3-3 draws come before 2-2, which come before 1-1, which come before 0-0)
2. Visiting team wins, ranked in increasing order of margin of victory (so all 1-goal wins come before all 2-goal wins, and so on; note that the number of goals scored in this case appears to be irrelevant)
3. Home team wins, again ranked in increasing order of margin of victory.
The top six numbers are the "drawn" numbers, with the seventh number being the supplementary/bonus number.
If there are multiple matches with the same result (e.g. 1-1 draws, or 1-goal home wins), they appear to be ranked in decreasing order of match number, which is what the OP meant by higher numbers doing better.

This site appears to describe it.

https://docs.google.com/file/d/0B1RSxZ2plqHPcXpQRkk2b0hHd0k/edit?usp=sharing

Here's a chart of the win frequencies

@Paradigm : Although you are correct in that its pointless following numbers since the underlying fixture changes every week, the exploit I identified is what I am seeking to take advantage of, it really doesnt matter which games are allocated to which number unless you are exceptionally good at doing soccer form and can identify score draws.. The flaw in the rules DOES lead to a bias as the later posters frequency table shows....the big question is how to take advantage of it. I have looked so far at system entries and Pick 4 and Pick 5 which are available in some states. The results are interesting with the caveat that using past results (even the lifetime set of them) of course is no predictor of future outcomes. That said a Pick 4 with the numbers most often selected - 30,31, 35, 37 (and 2 given numbers that always win on a Pick 4 ticket (You are given two free winners)) resulted in winning first division 4 times in the life of the game. A system 10 did the same (selecting numbers 27 through to 38) and yes a profit. Realistically I guess you would have to argue that if you chose the pick 4 method you should have to take 35,36,37,38 since the rule looks like it has a bias towards them even though the real results havent yet backed that up. Maybe the take from ticket sales is too high to make this pursuing, but its too tempting to let an obvious exploit like this go without further detailed examination.