Proof of concept / Testing time span?

Hey,

i'm a newb to bot/trigger/automatic trading, and i'm trying to develop a simple in-play strategy.
I'm working on the horse racing markets so i'm testing around 30 to 40 races a day with a ROI 1% daily, for about a week.

How long do you guys (the ones who have long term profits) test your strategy?

Thanks
TFAP

hi
sorry but for horse racing it should be a minimum of 1 year , this is because of the different seasons , each season has different grades etc and different courses , what you win in march/april will not win in may/june and so on.

Quote from: mcbee on 30 May 2019, 22:10

hi
sorry but for horse racing it should be a minimum of 1 year , this is because of the different seasons , each season has different grades etc and different courses , what you win in march/april will not win in may/june and so on.

Wow.. had no idea of that.
In your case, do you develop different strategies for each season or do you tweak one main strategy for all the seasons?
Thanks
TFAP

Sorry, just thought of this..

If the strategy needs to change from season to season, wouldnt it also need to change every year or couple of years?

Thanks
TFAP

hi
you need a strategy for all weather courses , turf , jumps and flat .
then there is the age , 2 yr old are unpredictable as they are new to racing also races at the start of a new season horses racing after a winter rest etc.

mcbee

When I test a strategy, I usually use Time Machine first, to get a rough idea of what works and what does not.

This usually involves running my triggers over a period of one month. I have a set of race events from different months which I know are diverse enough to cover 90% of my needs.

If your trading does not matter with horse class, season etc.. just price and live feed if trading in play, if odds are at 2 inplay could be liquidity not the position of horse .

Proof of concept / Testing time span?

Hello Guest