Percentrank complexity

[jcartwri]

I’m utilizing the PERCENTRANK.inc function when converting a long list of data into percents / I’m looking to have an unequal distribution of outcomes i.e. if the largest value is 2,000 that would be the 100% in my percent rank list, if the second number is 800 then instead of a high 90% result in a list of over 100 entries the result of 800 being the second highest entry (with 2,000 being the highest) would return a percentage of 40%. The percentrank function does an even distribution of results and I’m looking for a truer representation of the results versus the total spread of the entries. Maybe there’s a better formula to use than percentrank? Thanks for your help!