I have noticed over the five Saturday meetings this year at Ascot that the quinella dividends are more than reasonable and good value in most cases.
On the NSW tab the average quinella dividend over the 5 meetings (41 races) is $33. This does include a couple of high prices but also a couple of very low prices.
There are 8 main betting mediums that we all use in one for or other.
WIN.
PLACE
QUINELLA
EXACTA.
TRIFECTA
FIRST FOUR
SUPERFECTA
DOUBLE
If you were looking at a single selection in each race with combinations around it for the exotics then only the PLACE and QUINELLA could provide you with a result should the selection not win.
Taking a line through those two betting mediums then the banker quinella would provide a better value prospect while the place bet would provide a better strike rate.
It may be worth investigating a system around these two betting mediums.
Thoughts appreciated.

GG,
I've been keeping a casual eye on quinellas at Ascot also. The only comment I would offer is that it usually represents better value than the exacta. As two exactas equal one quinella you would expect a quinella to be about half the return of an exacta. In the majority of cases, the quinella pays much, much better than this, probably due to the smallness of the exacta pool compared to the quinella pool.

It is a common misconception that all exactas should pay double the quinella div - the reality is that this is true ON AVERAGE but the marjority of divs will be less than double (but some will be significantly more than double). The reason for this is that if you look at a given quinella combination (say horses 1 & 2) where horse 1 is more likely to win than horse 2 then the Exacta combination 1-2 will pay less than the combination 2-1. The 1-2 combination will also occur more often so you get the impression that Exactas are not paying double the Quinella as you would expect. However when the 2-1 combination does win the payout will be much higher and therefore the average will be maintained.

For example lets say the chances of 1-2 is 12% and 2-1 is 8%. The chance of the quinella 1-2 is therefore 20% (simply being the sum of the chances that the result will be 1-2 or 2-1). Without allowing for TAB take you would expect the divs to be:
Quinella 1-2 $5.00
Exacta 1-2 $8.33
Exacta 2-1 $12.50
Since the 1-2 combination happens 50% more often than the 2-1 combination you might think that the quinella is paying less than double the exacta but if you work it out the long run payout is double the quinella payout

I did a quick query on the Exacta and Quinella payouts for Ascot for the last 12 months and the average divs are as follows:
Quinella $49.32
Exacta $99.95

__________________
"Computers can do that????" - Homer Simpson

For the risk involved the illusion created is that the exacta dividend should be more than double the quinella but in reality it will be less.
The $49 average dividend is a nice one to be aiming at. I might give this more thought.

I agree completely with your analysis, BC. When I said one quinella equals two exactas,I was thinking more of the unit cost rather than the dividend. Irrespective of the odds of the chances of the first and second placegetters, the fact remains that you can cover both exacta combinations by doubling the cost of your bet.
I'm must admit surprise at the 12 month dividend result as I thought there was a definite bias, probably caused by the difference in pool sizes.
Good punting anyway.