Weight of Money

[Xeres]

My activity is based on placing back bets rather than trading so I hope this question is still ok in this section.

I was wondering if anyone had any insights into Weight of Money. I am trying to make sense of some of my results and whether WOM is worth the additional effort to try and incorporate it into my model.

Currently I use a combination of form and market data (price/volume/movements etc.) and make predictions/place bets a few minutes before the race. I started collecting more data so I could calculate the wom about 2 weeks ago, when I add this to my P&L this is what it looks like.


If I look at my bets 1 minute before the race, my overall strike rate is 34% with 25 points profit over 143 races

If I split this by wom and take the bets where it was weighted to the back price I get

strike rate - 38% over 103 races
profit - 25 points
average wom was 63% on the back side

So I get the same profit, meaning I should avoid any bets where the wom is weighted to the lay side

If I look at the bets 2 minutes before the race, the overall strike rate is 38% with 35 points profit over 126 races

Split by wom on the back side
strike rate - 38% over 79 races
profit - 24 points
average wom was 60% on the back side

The strike rate is the same whether the wom is weighted to the back or lay side and while I do make a profit on both, I make twice as much when its weighted to the back side.

Appreciate this is a very small sample, but since I don't have any data older than 2 weeks I would either need to purchase a lot of the historical data or wait a few months to build up my data set.

I noticed from my predictions that at 5 minutes before the race the signals struggle, and I was wondering if this is because of the noise in the markets, auto bots coming online etc.
Also the profit line is much more volatile 1 minute before the race, again I was wondering if this was slightly skewed by a lot of traders taking their positions out to creating more noise?

Any help/insights would be appreciated.