It would be difficult to illustrate without giving an concrete example, suppose you define the higher the sum of score of three dice as an indication of luckiness. At one day, you are not feeling lucky, and you draw the lowest sum of all: one, one, one, that has 1 out of 216 chance to happen. i.e. If we can rank what is most lucky and what is most unlucky, you are in the 216th of 216 ranks. Then you brought a quick pick of 6 numbers: 1,2,3,4,5,6; since you know that is not your lucky day, the chance of this number to appear in the lottery is 1/216. You can thus advise your friend not to pick any number from 1 to 6, which increase their chance of winning slightly. You can repeat the process to eliminate other numbers like 7,8,9,10,11,12; 13,14,15,16,17,18; 19,20,21,22,23,24; 25,26,27,28,29,30… until all but 6 number remained. That is something unusual given the computer picked number are most likely to repeated in each ticket. To fit the definition of being unlucky, it should reduce your chance of winning the lottery regardless of which strategy you devised to defeat it. So you should be expected to see a lot of overlapping numbers from each of the ticket your brought, because that would realistically defeat the scheme I devise here, otherwise the idea of a luckiness index is invalid. Say you have the worst luck of all, you have a repeat rate of five out of six(i.e., Given the first quick pick is 1,2,3,4,5,6; next quick pick is 2,3,4,5,6,7; and the 3th quick pick is 3,4,5,6,7,8), and it take you 37 more picks to eliminate all but 6 number out of 49. Nevertheless, you can pretty assured that the remaining number has a much higher chance (215/216) of appearing in the lottery.
On the other end, if you have the best of luck but not enough to win a lottery, this method could increase your chance of winning the lottery. How? Because which number doesn't appear in the quick pick must have a much lower chance to appear in lottery. You can apply this method in opposition direction. Since you are luck, it follows that the number of repeated number should be less for you to eliminate the one which has lower chance of winning the lotter.
Suppose we now have 216 people has luck ranked from 1 to 216th. If each of them buy a quick pick lottery ticket, since there is only one combination 6 out of 49 number that can win a lottery regardless of the luck of each buyer. We could easily use a computer program to guess which six number better fitted with 216 hypothesis that the chance of winning the lottery is reflected in luckiness index by throwing the dice three times. Of course, to further increase the number of quick pick that each person brought. For instance, each of them can get 6 quick pick, what the computer software has to do now is first evaluate 1296 hypothesis of different level of luckiness to get a coherent picture of the chance of each number appears in the lottery number; then evaluate the 36 hypothesis for fixed level of luckiness. So the computer can arrive at a coherent picture of the probability distribution of each number, and advise the best number to pick from.
Who would like to write such a computer software? It is just a lot of Mathematics. It can even apply this method in opposite, advising the buyer how many more quick picks to buy to maximize his/her chance of winning the lottery.
The opposite of unlucky is lucky. Shouldn't that lucky is the opposite of opposite of lucky?
沒有留言:
張貼留言