Ok a lot of talk about running doubles etc. I ask this question as I haven't done a lot of personal research but I do have my suspicions.

Lets take race to race doubles as some suggest, if we have a given SR and a given average price, wouldn't it therefore transpire that the outcome would be exactly the same as betting level stakes??? I suspect so.....BUT

Also though, lets take a given selection method with say 30% winners (usually the fav.) and then expand this into taking the top 3 in the market, so we have race to race doubles 3x3 = 9 units, now we know that 1 in 3 will win the first race and that 1 in3 will win the 2nd race, so we should average 1 winning double every 9 times with an average return of 2-1 (hopefully) = $3x$3 so we break even, BUT we can also expect that sometimes we have 1 winning Fav. + 1 winning 2nd Fav., and also sometimes 1 winnimng fav + 1 winning 3rd fav, etc etc.to the point where we have 1 winning 3rd fav x 1 winning 3rd fav

I would be interested if anyone has explored this idea????