Page 1 of 1

Clarify the ranking algorithm?

Posted: Fri Apr 24, 2009 4:24 pm
by Zeta
I did an analysis of the output of http://www.hacker.org/challenge/map/get.php and it didn't match up with the ranking shown in http://www.hacker.org/challenge/top.php.

I simply used the "sum of the squares of the difficulty" method to compute my score and that resulted in a considerable smaller score compared to the official ranking.

How do I have to understand the Sentence "The top solver gets 200 points, the next (200 * 0.9), the next (200 * 0.9 * 0.9) and so on"? Who is the top solver? The first one who solved a challenge?

Posted: Fri Apr 24, 2009 7:45 pm
by bsguedes
I've been realizing a lot of mistakes in Challenges' Ranking. As Zeta has found, my score in the ranking does not match with my score calculated considering the squares of levels dificulties. I've calculated 116096 points, however, I have 166415 points on the ranking.

About the 200*0.9^(n-1), where n is your position on the ranking of Challenges, a friend of mine, portal, should have 15 points due to his position on rank (he is actually 25th). However, his score on global rank seems to not include 15 points relative to challenges.

Also, there are some inconsistencies on the general rank. How Kellpete has 255 points? This is only explainable if Puzzles have different weights on the total sum, but I suppose that each puzzle counts 100*0.9^(n-1), where n is your position in that puzzle again.

Posted: Sun Apr 26, 2009 1:51 pm
by jeetee
Don't know about their weight, but yeah, puzzles and bot-rankings are taken into account for the global ranking

Posted: Sun Apr 26, 2009 3:00 pm
by bsguedes
Ah, now I understand. Kellpete is leader of BitBath. I thought Bot wars aren't taken into account for the ranking.