I'm often asked about sample size of a system for confidence.

Some say 100, some say 1,000, some say 10,000, while of course it depends on the odds, if you're targetting 50/1 shots, 100 certainly won't be enough, and if you're targetting even money shots, 10,000 is way too much research.

The best rule is 100 winners.

__________________
RaceCensus - powerful system testing software.
Now with over 415,000 Metropolitan, Provincial and Country races!
http://www.propun.com.au/horse_raci...ng_systems.html
*RaceCensus now updated to 28/02/2025
Video overview of RaceCensus here:
http://www.youtube.com/watch?v=W821YP_b0Pg

So what your saying is size doesn't matter!

My home help doesn't agree.

Another thing she said was it's how big it gets from when you just started.

My method to beat premature celebration is this: Once you think you've found a system that shows promise begin with a $1 wager. Follow the system until you have a ten unit profit then double the wager until the next ten unit profit is made and so on. If it is a worthwhile system you are soon into serious money. If it's a dud you haven't lost your house. This method has the advantage of keeping a record in real situations of what went down and it also keeps you honest to yourself of what you can and can't do.

Two cents! By the way CP thanks for your input on this forum.

__________________
"Not winning on a horse that came first is one thing.....Losing on a horse that didn't come first is something else entirely!!!"

It's not how big the data is, it's how you use it

__________________
RaceCensus - powerful system testing software.
Now with over 415,000 Metropolitan, Provincial and Country races!
http://www.propun.com.au/horse_raci...ng_systems.html
*RaceCensus now updated to 28/02/2025
Video overview of RaceCensus here:
http://www.youtube.com/watch?v=W821YP_b0Pg

IMHO I don't think it's really a numbers answer.

If you're backing only horses in the red you're 100 winners will come much quicker than the guy who backs horses at 20/1 or more. Or if you've backing 4 horses per race per day compared to 2 bets on a Saturday you're number of bets required to assess are over a completely different time [in theory].

It's really time - and in that respect it's a full racing season needed to accurately assess.

Even if you're basing your system purely on numbers with little regard to form - which very few are I'd suspect - you still need the full racing season which exposes you to all facets of form.

Group 1 races, blacktype races, standard races, lower quality races. High quality jockeys...lower quality jockeys...apprentices. Large fields, small fields, dry tracks, wet tracks. First up horses, horses that are well into a campaign. Sprints, middle distances, staying races.

Even if thoughts, ideas, systems choose to drill down on a theory tied to one facet of form you're still likely to be influenced by other factors. eg you may choose to only bet on sprint races in Victoria. But backing horses in May/June is completely different to backing them in say August when horses are resuming for spring campaigns and you start weighing up first up horses with a class edge over race fit types with less class.

Just my thoughts.