I asked TheDoctor if he could explain us the details behind BallerBude stats, and he kindly sent me an extensive explanation, of which explanation he authorizes the use in this site. So here it is almost in its entirety, I only removed any reference to known players. It’s an honour for *LAME* Clan to give hospitality to such an authoritative contribution. Also thanks to @spi for his comment to previous post.

Obviously, BB stats try to measure how good players are. However, one should keep in mind, they represent just one of many possible metrics and no metric is ever perfect. What I mean here is, it is possible for a very successful player winning almost every game to be not in the top ten, while somebody who does not win every game has a higher score.

The following values are considered by BB’s xlrstats: skill, ratio and kills. Skill is the default ranking value.

Kills is the number of opponents a player has killed.

Ratio is calculated by dividing the kills of player by the number of deaths of the player (“kills/deaths”). A high kill ratio is an indicator for players who have been constantly performing well over their whole account life time. One can optimize ratio by playing very carefully and defensive, possibly by camping at good camping spots. The optimization strategy is obviously to survive as long as possible while killing other players. Ratio is a long-term value meaning, if you have a series of bad games, it will lower your ratio and it would take the same amount of very good games, to let ratio reach the previous value. In other words, it is possibly to ruin your ratio for a long time.

Skill, compared with ratio, is a much more dynamic value. Each player has a skill value (default 1000) and killing an opponent adds skill points while dying reduces skill points, both relative to the skill relationship between the player and it’s opponent. If you kill a more skilled player, you earn more points, while you lose more points if you are killed by a nobody. Similarly to the ratio, it helps if you don’t die often. But if you are really hell-bent on optimizing your skill points, you first have to check the skills of the other players before you would join a game. If you play against overrated (=”overskilled”) players, you will quickly earn points while you can lose many points if a good player decided to create a new account, starting with skill 1000. The skill value will react very quickly to your current level of playing, so one shouldn’t play if one is not a hundred percent focused.

Xlrstats support even more complicated metrics, e.g. one could define that a knife kill is more skilled than a sawed-off shotgun kill. Currently, these features are not used.

I think this explanation is quite informative. At least I finally understand the difference between skill and ratio, and why skill has so frequent fluctuations (at least for an average skilled player as me).

Let me conclude this post in rhyme with a lamerick on the topic 😉

Sir Camp-a-lot, sitting on top,
Waits for a prey (possibly a bot),
Proud of his skill,
Till he gets killed,
Falling from top, Sir Camp-a-lot.