Spot the difference between the educated man and the chap who read half of a book on statistics once.
But I have few observations:
You have taken out the 9 fallers but not the 40 runners from that race.
No I haven't...I have just not gone as far back as that...that was 11 years ago.
If you take out the 40 runners the odds become 15/2 (or 6.2/1 with a 20% reduction)
No need, see above.
If you use the entire sample (assuming 38 runners on average), the odds come in at a shade under 14.9/1 (or 11.9/1 with the 20% reduction).
There have been so few non-runners in the last 10 years, that I just went with 400 runners. Won't make a bid difference.
Which is still a lot better than I had it at but not a value bet.