In the spirit of early mixed martial arts, the MMA Top 100 is not separated into weight classes. As it stands, the rankings can be considered a "pound-for-pound" ranking, or simply as an overall ranking. So far it seems that heavyweights tend to move toward the top, but that may change depending on how few mixed-weight matches there are in the future.
These rankings are broken out into Weight Class Top 25 listings as well; there is no difference between the systems used and when a fighter moves up or down a weight class his score remains the same. As of this writing there are seven weight classes tracked, with Bantamweight encompassing Flyweight so that all 135-lb-and-under fighters are lumped together.
The ratings are calculated in a manner similar to the heavyweight boxing top 100. If you beat someone, you are guaranteed to be ranked above him in the rankings, and you have to beat highly-ranked fighters to achieve a high rank. Both of these ranking systems use only wins and losses (and draws) to calculate a fighter's score; the "degree" of win (KO, split decision, etc.) isn't taken into account, only the level of opposition defeated. As in boxing, after a year of inactivity a fighter's rating begin to decline, falling to zero after five years.
The main difference between these rankings and the boxing rankings concerns the list of fights used for calculating the rankings. Boxing's data goes back to 1881; the MMA data doesn't start until over 100 years later. In some ways the MMA data is more complete, as boxing records from the 1890s aren't nearly as thorough as today's MMA data. But the boxing data has had decades of fact-checking while early MMA data is still going through that process. Thus, for certain MMA fights, even in higher-profile organizations, there can be conflicting reports as to who won, or whether two fighters with similar names are indeed the same person.
Our MMA data starts with the Shooto organization, and the early 90s data includes all of Pancrase, the UFC of course, and pretty much any event that was on TV anywhere in the U.S. or Japan. If it features mixed martial arts (striking, kicking, and grappling all allowed, regardless of specific rules) and wasn't pre-determined (like most early Rings bouts), then it was included in my list of these events that I kept throughout the 90s. Later in the 90s as smaller promotions became more common, I have included as many of these as I can; basically, if one fight featured an already known quantity (e.g. Dan Severn fighting on a small-town card in Illinois), the results from the event are included.
Every fighter from the 1990s who competed is automatically a "registered" fighter. Moving into the 2000's, I track only the fighters who were registered in the 90s and anyone who beats them or draws with them. This is why the "Rated Record" for each fighter rarely corresponds with their "official" record or their record listed on various MMA sites. Many of their wins are over fighters who are "unregistered" in our system. Well over 5,000 fighters are in the database, with less than 1,000 actively competing today.
Tracking MMA results is as much of an art as a science, so when I find errors or discrepancies I fix them; this may change the current results, but usually in a very minor way. Much of the "revision" I do concerns worked fights, which is again a judgement call. As mentioned before, most early Rings (before 1999) is not included, except for a few known "shoots" per event. As the amount of MMA content on the web grows, more information about specific fights can be gleaned and the data should become even more "clean."