Ranking method wins Vs attempts

[RobOrBob]

Hi all,

I need to create a relative ranking system but am stumped!

It needs to assess the relative ranking of a set of competitor results but also consider the frequency. For example:

Competitor	Wins	Attempts
1	1	20
2	1	1
3	10	10
4	4	5

So number of Wins doesn't tell the story and neither does a % of Wins/Attempts.

Logically the competitors would probably be ranked in the order 3,4,2,1 but again this would not give the full picture (e.g. each Competitor relative to the others).

I think what I am looking for is a points value.

Any thoughts gratefully received!

 

[Joe4]

Welcome to the Board!

You could divide WINS by ATTEMPTS to get Win Percentage, and rank by that, and sort by Win Percentage first, and Wins second (so the order would be 3, 2, 4, 1).

However, if you want to put more of a weight on number of wins too, so 4 out of 5 is ranked higher than 1 out of 1, you will need to come up with some formula to determine how to weight them.
There is not one right/wrong answer here, it is up to personal preference, and it can get a little tricky.

For example, should 1 out of 1 be ranked higher than 2 out of 3?
Or is 2 out of 2 higher than 4 out of 5?

You could do two rankings, one based on Wins and one based on Win percentage, and then apply some weighting to each, and come up with a final ranking based on those two weighted rankings.
It really is up to you. There is no wrong or right answer here.

 

[Joe4]

I came up with one method that seems to return the order that you want.
1. Calculate the win percentage
2. Multiply the number of wins by the win percentage
3. Rank the number calculated in step 2

It would look like this:

 

[RobOrBob]

Hi Joe4,

Thanks for the reply.

Your column E is the key methinks. Perhaps if I then take each number then divide it by the highest value in the data (MAX) and then multiply the result by 100, this gives me a relative value. Maybe even use a deviation score.

Whatever, you got me over the brain stall!

 

[Joe4]

You are welcome.
Glad I was able to help!