ranking by weighted results

I am trying to obtain an overall ranking for a series of locations using several different categories, weighting each of those categories. I'm not sure if this makes sense, but here is an example and what I have attempted to do so far...

locations Result 1 Result 2 Result 3
location1 515 89% 4.5%
location2 431 72% 2.3%
location3 535 n/a 5.0%

Weighting...
Result 1 = 30%
Result 2 = 50%
Result 3 = 20%
Total - 100%

What I am trying to do is determine which location is ranked the highest
overall.

I inititally thought since the results varied in their number format that I
would first rank each result by location. ex: Location1 > Result 1 > Ranked = 2

I then took the rank and * by the weight. ex: Location1 > Result 1 > Ranked
> 2*30%

I did the above formula for each result per location. I then sumed all results together per location and divided that number by 100.

Then used that final number to obtain each locations rank.

I'm honestly not sure if I was correct in the mathmatics of the formula. I
also need to know how to not have a result count if the location did not have a result. ex: location3 > result 2 = n/a. And lastly, If
there is an easier way to obtain this result, I would love to know. I feel
I'm complicating this process when there is more than likely a more simple way to obtain.

Originally Posted by casdaq

I am trying to obtain an overall ranking for a series of locations using several different categories, weighting each of those categories. I'm not sure if this makes sense, but here is an example and what I have attempted to do so far...

locations Result 1 Result 2 Result 3
location1 515 89% 4.5%
location2 431 72% 2.3%
location3 535 n/a 5.0%

Weighting...
Result 1 = 30%
Result 2 = 50%
Result 3 = 20%
Total - 100%

What I am trying to do is determine which location is ranked the highest
overall.

I inititally thought since the results varied in their number format that I
would first rank each result by location. ex: Location1 > Result 1 > Ranked = 2

I then took the rank and * by the weight. ex: Location1 > Result 1 > Ranked
> 2*30%

I did the above formula for each result per location. I then sumed all results together per location and divided that number by 100.

Then used that final number to obtain each locations rank.

I'm honestly not sure if I was correct in the mathmatics of the formula. I
also need to know how to not have a result count if the location did not have a result. ex: location3 > result 2 = n/a. And lastly, If
there is an easier way to obtain this result, I would love to know. I feel
I'm complicating this process when there is more than likely a more simple way to obtain.

Hi.

a shake of a set of dice starts to look good huh . . .


Your values seem difficult to relate to each other, asin the values for category 3 are low %, does this mean that they are already expressed as a maximum of 20% and are thereby 'pre-weighted'?

What is the marks of 515, 431 and 535 expressed in relation to, marks from 1000 ? from 600 ? How would these relate to the % shown for category two and three?

Category two would seem easy to weight, as one-half of the category two % would equal it's weighted value, but only if the weighting is on an even basis with the marks as allocated to the other two categories.

any further clues?
---

my suggestion would be to Rank each category, then weight each ranking for an overall position.

---