Rooburger, an interesting one this, there was a variation as well i.e. any dog/trap that wins twice usually wins a third time, (STRESS I HAVE NOT done the figures there as it was from the UK) but worth some investigation.

But getting back to your post, one thing I like about this is: it is a downright certainty that 1 trap has to repeat, followed by the fact as Ubetido pointed out that it could be 1-10.

But I still believe that if you're using a cover to win staking plan, it is possible to make a profit OVERALL, but there would still be losing days as it could not possibly always go your way (odds wise)

My idea here is to always bet to a minimum price of 2-1, i.e. if one of the "investments is below 2-1, back it as if it was 2-1, YES of course sometimes it WILL win and your return is less than your target, BUT on the flip side bet anything above 8-1 as if it was 8-1.

So when that longshot gets in at 25-1 you make the killing and hopefully offset the times when you bet the 4-6 winner with not enough invested.

Does that make sense? going for the average in other words.

An interesting variation, applying the idea to horses, is this, most winners come from TAB nos 1-6, so you only consider these runners - the top 6 TAB numbers (AFTER SCRATCHINGS)
so therfore with a usual of 8 races the fact is that one number will probably repeat.
Just did a quick check to see if it worked today, and it did EVERY meeting including Perth trots tonight

Make no mistake, it WILL happen that no numbers 1-6 will win in races 2-7, hardly ever BUT it will happen, so your staking has to take that substancial loss into consideration offset againts the times when you back a 50-1 winners as a 8-1 shot, I guess like all systems there is an element of luck, and an element of risk.