Repeats

[niceguy]

Hi stan
Just wondering if it is possible to include in the statistics part, statistics to show the "average " number of repeats for a given amount of past draws. This would be a big help in using the repeats filter to decide how many repeat numbers we should include in our next lot of numbers, from the last "x" amount of weeks results.

[stan]

what is average number of repeats?

[niceguy]

If you were to look at how many repeat numbers there are from one draw to the next, and work out how many numbers are repeated on average, this would give you a good idea of how many to include in a wheel, or filter when working out what the next games results are likely to include. For example:- if 5 games ago there was 3 repeats in the next game, 4 games ago there was 3 repeats in the next game, 3 games ago there was 4 repeats in the next game, 2 games ago there was 2 repeats in the next game, then the average number of repeats would be 3+3+4+2=12 divided by 4 (games) gives us an average of 3 repeats for the last 4 weeks. Therefore, when using filters for example, you know to include UP TO 3 numbers in each game from the last game. I think it would be a very helpful innovation to have, & would give a more accurate result over a large number of games? Be interested in others comments as well.

[Jerzy]

niceguy,

The simple statistic of the numbers repeated from the last draw is available in Statistics --> Winning Numbers --> Repeated [1]. You can see how many numbers from the last draw is repeated most often. If you want to know actual occurrence of repeats in the past draws in percents you need to calculate it outside of EL.

In my 6/45 game,
none of the numbers from the last draw was repeated in 40.3% of the past draws,
one number from the last draw was repeated in 42.0% of past draws,
two numbers were repeated in 3.0% of past draws,
four numbers were repeated in 0.1% of past draws (on one occasion only).

Most commonly 1 number is repeated. Using Match Numbers Filter I can filter the package leaving in it only the tickets with 1 repeated number, or only the tickets with no repeats and tickets with one repeated number. I can even decide to leave only the tickets with the count of repeated numbers which is "due".

Unfortunately, the "everage number of repeats" does not offer any advantage over the most common count of repeats.

Calculating the average count of repeats usaually gives a fractional value which needs to be rounded to the nearest integer and this integer is equal to the most common count of repeats.

For example:

Repeats in the last 10 draws:
1, 1, 2, 2, 1, 1, 0, 1, 2, 0 Total count of Repeats = 11 Average count = 1.1
Repeats in the last 20 draws:
1, 1, 2, 2, 1, 1, 0, 1, 2, 0, 1, 1, 1, 2, 2, 1, 2, 1, 0, 1 Total count of repeats = 23 Average count = 1.15

In both cases the average counts of 1.1 and 1.15 must be rounded to the nearest integer which is 1. And the most common count of repeated numbers is also 1.

Regards

Jerzy

[niceguy]

Hi Jerzy
That's interesting. Your draws seem to be a lot more consistent than ours. In NZ (6/40) in the last 50 draws for example, we have had anything from 0 to 4 repeat numbers from last weeks draw. My thinking here was that if you could find some sort of "average" figure to go by, (rounded up or down to the nearest whole number), it would make it easier to decide how many numbers to include from the previous week in your next selection, when using filtering. Also, if we had this statistic available you could experiment over a very large amount of draws to get a more accurate figure instead of trying to work it out manually yourself. ( also use it to get the "average" number of repeats for the repeat columns as shown at present in statistics?)

[Jerzy]

niceguy,

I saw the data for NZ 6/40. The repeats from the last draw are similar to those in AU 6/45. 0 repeated numbers and 1 repeated numbers are most common.

The choice for setting the Match Numbers filter is rather limited and quite obvious. You select something which is most common or something which seems to be "due". Why would you select something which is just average?

The numbers drawn are full numbers not fractals. And, for example, how would you round the average of 1.53 to make it a full number? Would you round it to 1 or 2?

And such average would be always higher than 1, "discriminating against" 0 which is very common count of repeats. If you want to filter out the tickets with no repeats, do you need to see the average count of repeats?

I can't see any role to play for the average repeats parameter. The best approach is to keep things simple.

[niceguy]

Hi Jerzy
Thanks for the advise. It is really appreciated. I will now follow your thinking when it comes to repeats!!. :cheesy: