Hey there Vortech, I'm not at home at present, down in sydney visiting with the kids. The last few weeks before I left home I set up a very long drawn out spreadsheet where I could rate the last 3 runs of every runner.

Thinking this would give me an edge with races of say 1500m and beyond. But after a short while I tended to notice that the rating for the "last start" outperformed the rating for the "average last 3 starts".

I did of course for second uppers just use the "last start" and for third uppers use the "last 2 starts". First uppers are of course the problem. Races with to many of these just should be ignored.

Not having any bets yesterday (saturday), being driving in the car all day, my last betting day was last saturday. My program rated the last winner in melbourne on top in its "last race", paid around $30 for the win. On the average of the last 3 starts it was not on top. From memory it was a 1600m race at Flemington. It was the roughest horse in the field. So I am thinking why did it rate so well. It had its previous run at Mornington I think on a slow track over also 1600 and the time of the race was better than my par time for that track/distance/going.

CP's data base has allowed me to fill in a lot of blanks that I had for tracks here and there. On another thread somewhere recently which I think you also started, I commented that I am using the average of the best 10 times for track/distance/going. I think I now have well over 4,000 times in my vlookup columns. So for what its worth I tend to think anyone using ratings, that the last start rating outperforms what I originally intended to use, which was the average of the last 3 starts.

Paul