Questions regarding the creation of an "unbiased" database

Questions regarding the creation of an "unbiased" database

Postby fiwip » Tue Feb 04, 2025 5:56 pm

Hello,

I would like to have your opinion and possibly your advice on the following matter. My objective is to build a database without bias.

The lottery in question is the French Lotto: 5 balls drawn from 49 + 1 additional number drawn from 10. These are two separate machines. For the "classic" Lotto, there are three draws per week. As of today, there have been approximately 2,560 draws.

Additionally, without a fixed schedule, extra draws are occasionally conducted using the same machines for special events (Christmas, the Olympics, etc.). These special draws are called "Super Lotto" or "Grand Lotto." Since their introduction in 2008, they have accounted for a little over a hundred draws.

Here are the technical details I was able to gather:

- Regarding the classic Lotto, the Super Lotto, and the Grand Lotto for major occasions, the machines used are supposedly the same. They are "Ipomée" machines manufactured in France (https://www.ryo-catteau.com/).

- The balls appear to be identical in design (same material, weight, and diameter), except for their colors, which vary depending on the type of Lotto. Each type of Lotto has its own set of balls, which then are not used with the same frequency.

- The original machines from 2008 were replaced in 2017 by the "Ipomée" machines mentioned above.

While I am aware that the influence of 100 or 200 additional or fewer draws may not have a significant impact on over 2,500 draws, statistics, even in a completely random world, are what they are (for example, in terms of number gap distributions). It would be best not to introduce any additional bias.
⇒ In your opinion, should these three Lotto types be treated separately in statistical analyses? Or can all draws be combined? Should the classic Lotto be analyzed separately from the other two?

Due to the machine change in 2017 (this may seem like an obvious question), I would still like to get your opinion : should the initial series from 2008 be disregarded, and a new series be started from 2017?

Thank you in advance for your insights and for any additional information you may provide.

Best regards,
Philippe
fiwip
 
Posts: 11
Joined: Mon Oct 22, 2007 3:10 pm

Re: Questions regarding the creation of an "unbiased" databa

Postby tdnl46w » Thu Feb 06, 2025 1:22 am

fiwip wrote:... Additionally, without a fixed schedule, extra draws are occasionally conducted using the same machines for special events ...

If the extra draws are semi-random or don't occur often, then I would not include them in the regular draws. But you should still be able to play the special events draws using the regular draws statistics.

fiwip wrote:... should the initial series from 2008 be disregarded, and a new series be started from 2017? ...

I would analyze the 2008 and 2017 draws(compare in charts etc.) to see if the new machines changed something. If there is any clear differences in statistics between them, then I would only use the newer 2017 draws.
tdnl46w
 
Posts: 425
Joined: Thu Jan 29, 2015 10:56 am
Location: USA


Return to Comments, suggestions, feature requests

Who is online

Users browsing this forum: No registered users and 1 guest