I am using the following formula to determine the individual rating of a
group of players running a race.

F3=CEILING((COUNT($A$3:$A$23)+1-RANK(E3,$E$3:$E$23,1))/(COUNT($A$3:$A$23)/5),1)

A B C D E F G H
2 ID. No. Time #1 Time #2 Time #3 Average Time Rating Gender Level
3 6 01:01.1 02:59.6 01:59.9 02:00.2 5 Male U14
4 5 03:00.0 01:01.1 02:24.7 02:08.6 5 Male U14
5 16 03:02.8 02:47.7 01:01.1 02:17.2 5 Male U14
6 4 02:24.7 01:59.9 02:32.6 02:19.1 5 Male U14
7 17 02:25.2 02:11.3 02:47.4 02:28.0 4 Male U14
8 19 02:15.6 02:21.2 02:47.7 02:28.2 4 Male U14
9 11 01:59.9 02:25.8 03:01.9 02:29.2 4 Male U14
10 12 03:01.9 02:24.7 02:12.6 02:33.1 4 Male U14
11 2 02:12.6 03:01.1 02:36.9 3 Male U14
12 7 02:25.8 02:55.5 02:42.2 02:41.2 3 Male U14
13 1 02:21.2 02:42.2 03:02.8 02:42.1 3 Male U14
14 13 02:47.7 02:12.6 03:12.2 02:44.2 3 Male U14
15 20 02:59.6 03:01.1 02:25.8 02:48.8 2 Male U14
16 9 02:59.9 03:03.5 02:23.4 02:48.9 2 Male U14
17 10 02:42.2 03:01.9 02:47.7 02:50.6 2 Male U14
18 3 03:18.5 02:56.5 02:45.6 03:00.2 2 Male U14
19 8 03:01.1 03:12.2 02:55.5 03:02.9 1 Male U14
20 18 03:12.2 03:02.8 02:59.6 03:04.9 1 Male U14
21 15 03:13.2 03:05.2 02:58.6 03:05.7 1 Male U14
22 14 03:15.2 03:46.6 03:30.9 1 Male U14

My problem is that this group of 20 players is only a fraction of the
players in the total database. I need the formula to be able to pick these 20
players out of a list of 400 or more players before rating their times.

The filtering criteria would be from columns €˜G and €˜H. The example above
is for all €˜U14 males so I need the formula to rate each €˜U14 male compared
to all other €˜U14 males in the table. Other possibilities for column €˜H
would be U6, U8, U10, U12, U16 and U18. Note that the €œU€� in this situation
is short for €œunder€�; it does not refer to column €˜U.

In addition there may be 30 €˜U14 males to choose from but 10 of them may
not have actually participated in the race. So the formula has to weed out
the €˜U14 players that do not have an average time (the cell would read
#DIV/O!) and only rate the players that have a time. An option might be to
enter zeros in the time slots for the players that didnt participate thus
giving the player an average time of zero.

When I posted the table above the forum engine kept mixing up some of the
columns, sorry! I tried making the columns thinner but that didn't help.
The columns headers a
A= ID. No.
B= Time #1
C= Time #2
D= Time #3
E= Average Time
F= Rating
G= Gender
H= Level (whicn is represented by U14)

Thanks,
Ksean