Skill is not a linear measurement. A player can do well one game and terribly the next. It would be difficult to create an accurate ranking system to compensate for these variations between games, which is why some system choose to go for a median skill range, such as TrueSkill.

The problem with TrueSkill is the very thing that Sze likes about it so much; it doesn't expire.

A player who hasn't been on the ladder for 3-4 months will have the same rating as before, but might not be at the same skill level. Naturally, after playing a few games, his rating would drop, but at the time he played his rating was higher and he would have been paired with players of near-equal rating, which would have been inaccurate.

Also, this isn't Chess and WarLight isn't as popular or well known as Chess, so we can't be so choosy about players who enter the ladder or block higher rated players from competing outside their rates for easy points. What would perhaps be more interesting is a feature that shows longest time held in their top ranked position, among many others (ideas).

In any case, there should be a number of updates and fine-tuning to all the ladders to make them more interesting and more accurate, but whining about it instead of providing some sirius research and actually aiding Fizzer instead of berating him, only plays the holier-than-thou route and doesn't change anything.

“Gardens are not made by singing, “Oh, how beautiful,” and sitting in the shade."
-Rudyard Kipling

Stop singing people, my ears are bleeding.