rolling averages

Can't you just replace the oldest time with the new one and keep a note, as to which is the current last time?

If you want to keep all the old times though, you can try this: =AVERAGE(OFFSET(C1,COUNT(C:C)-10,0,10,1)), assuming the times are listed in C column starting from cell 'C1'

In cell D10, write '=AVERAGE(C1:C10)', and then click in the bottom right corner of that cell (so that you see a cross), and drag downwards. The cells below will fill up with =AVERAGE(C2:C11), =AVERAGE(C3:C12), etc. which is what you need.

Thanks for your response.
It's not quite what I need.
With greyhound data the most recent data is the most useful.
What I do at present is to delete the entry for C1 (The oldest), drag (C2:C9) up and drop it in the C1.
That then leaves C10 free for my latest entry.
What I was hoping for was a way to automatically do the above without deleting and drag and dropping!
Thanks for your efforts so far.

So you need a weighted average where recent data is more important than older data, as opposed to arithmetic average where each data has the same importance?

I understand you already have a formula to calculate this, but you need your times arranged automatically?

Try this:
COLUMN D
in D10, put the following:
=IF(AND(C10"",C11=""),AVERAGE(C1:C10),"")

Copy this down column D

As you fill in D11, it will automatically blank D10 and D11 will bring up the average.

This also allows you to keep all the data and if you want, create another column with a different average (i.e. 15 races for comparison).

CORRECTION, it didn't post right:

=IF(AND(C10"",C11=""),AVERAGE(C1:C10),"")

Copy the cell and bring it down through D

so how are you using this? your pick is the one with the best average? is your system working?

Sunset,
I make allowances for "Bumped -.10", "Crowded -.20", "Baulked -40"
Its in its early days but so far so good.

cheeky....you might have incorporated this already but you'll find that results from certain grades will have less 'credibility'.  Some really random A7, A8, etc.

Also, check your returns against certain tracks.  I had a filter that was useless except at Nott and Pbar where it had exceptional strike rates and returns.

I thought Excel already had formulaes that do ARIMA modelling ... perhaps you also need to consider the weighting of the recent clocks versus older clocks ....

I run a greyhound system and find that I need to put in a stop-profit flag in with the software I am using.

Good luck with your venture.

As feix mentions above, a weighted average would be better, including all results but taking more notice of the most recent.

In Excel, this is fairly straight forward, you just need a ratio in a cell (say $E$1) to show the weight you want attributed to the most recent piece of data, if you were thinking of taking a level average of 10 latest results, an appropriate ratio might be 10-15%.

Put a second column (D1, D2 ...) next to your full set of data (C1, C2 ... as above)

In D1 just reference C1, i.e. formula of "=C1".

In D2 enter the formula "=((1-$E$1)*D1)+($E$1*C2)" and copy this formula down through the rest of your second column.

This column will then show your updated weighted rating as your results increase.