Markov chains

[luisgmira]

Hi.

I'm interested in Markov chains theory and I wonder if GAT uses an algorithm based in it.
Just for curiosity.

Or, perhaps somebody has an idea about the effectiveness of Markov chains prediction in lotto.

Thanks and regards.
Luis

[lottoarchitect]

The simple answer is no. Markov chains are basically predetermined probabilistic rules and given a specific state, you have a% to do A, b% to do B, c% to do C etc. Not sure how this can be applied in lotteries and make some sense. So for example, we could say current state (draw) has number 5 in there, so 2% for next draw to have number 1 drawn, 4% for next draw to have number 2 drawn etc where the % part may be computed from overall analysis of the available data. This approach sounds like a typical "announcer" filter really. Unless if someone has a more intelligent use of Markov chains on lotto, I don't see how this can be any better than traditional statistics.

[luisgmira]

Ok. Thanks lottoarchitect. Indeed it is a very poor approach to pattern discovering, just another "skip and hit" method that has proven to be not very useful in lotto.