A new truly 'divisional' ranking system

I started a numerical ranking system a while ago... and now finally have the results to unveil how it works! The three fundamental principles of the system are: 1) Only fights at the weight division can contribute to a ranking in that division. Weight divisions exist because size matters and, moreover, each weight division is a different competitive field, so allowing fights at other weights to contribute to a divisional ranking isn't fair to fighters proven at the weight. 2) Level of opposition is paramount. You must beat (or draw with) a ranked fighter to get in the rankings. It's unacceptable to hand a fighter a high ranking without beating a ranked fighter. Unique to my system, a loss to a ranked fighter doesn't move a fighter down the rankings: a fighter shouldn't be able to sit on a ranking by taking on lower opposition, so losses to top opposition aren't punished. 3) Only results in the last three years are included. Three years is a fair period because boxers are generally at their current level for about that time. Older fighters shouldn't be given credit for their prime years, while unproven younger fighters shouldn't be overcredited for their most recent fight or two. Ultimately this is trying to show that boxing rankings, just like rankings or standings in all other sports, should only reflect actual accomplishments. Ability, potential, or impressiveness of performances is far too subjective, so allowing that they should be reflected even somewhat in rankings will never allow consensus rankings to be established. That members of the TBRB publish opposing rankings based largely on differing opinion of ability shows how pointless and ineffective this method becomes. The 'subjectivity' of Boxrec's ranking system is also the clue to the failure of their system. The statistical details they include are problematic because they're included in an attempt to thoroughly 'evaluate' the quality of wins. They give less credit for split decision wins, but don't seem to get that three judges are only used to arrive at a single decision. They give more credit for knockouts, but boxing isn't about knockouts. They adjust ranking based on past or future results of opponents, but it's pretty obvious a fighter should only be credited in rankings based on the level of an opponent when he fought him. I could go on, their system was made by stats nerds who miss the forest for the trees. The only indisputably objective facts are: 1) the date, 2) the weights, and 3) whose hand was raised, and those are the simple, neutral criteria used for this system. Please check out the detailed rules, and the welterweight rankings here... http://worldboxingrankings.proboards.com/ The one aspect of this system I suspect will be controversial is that the winner doesn't necessarily move above the ranked fighter he beat. Manipulating the system so the winner is always higher ends up devaluing the three year period and proven success over more than one fight, which should matter more--as one fight can be just a matter of styles, or a bit of a fluke. As it turns out, an unranked fighter beating a fighter ranked but below top 10 usually puts him higher, but an unranked fighter beating a top-10 opponent usually doesn't move above him.