After taking in some great criticism about the last set of rankings and thinking it over, I have completely reworked the system. I have thrown out the old "keep it simple" mantra and decided that simplicity should not outweigh a logical and functional methodology.  Although I didn't take every suggestion, I considered all the comments aside from those that suggested rankings in general have little value.  I also realized that I forgot a bunch of fighters.  Here are the changes…

1. Fighters can accumulate points by beating anyone in the BE top 25. This gave several fighters a bump up. Using the new scale, they earn 25 points for beating a #1 ranked opponent, 24 for #2, 23 for #3 and on down to 1 point for beating a #25 ranked opponent.

2. Another issue that was illuminated was that weak divisions were treated equally to strong ones. Originally I thought it was unfair to penalize the smaller guys for the state of their division. It’s not their fault they weren’t born heavier. However, it is true that fighters get ranked quicker in a inexperienced division and therefore can take a shorter path to P4P status. This made me look at the cumulative experience of each division. I tallied the total fights of the top 25 in each weight class:

As expected the lighter divisions had less experience. Since MW had the most, it was used as the measure for all other divisions. Dividing each cumulative fight total by the MW total, we get the following "Division Experience Factors":

Once the lighter divisions get more experience, this adjustment will become less drastic.

3. The old saying, "The bigger they are, the harder they fall." rings true in MMA.  BWs manage to keep it out of the judges hands 50% of the time while HWs finish their fights a whopping 71.9% of the time.  That makes it very dangerous and difficult to compete in the heavier divisions.  There is such disparity that it must be accounted for.  Here is how the percentages break down:

This time, we use HW as the measure for all other divisions.  Dividing each finishing percentage by 71.91 we get the following "Division Finish Factors":

Similar to the experience factor listed previously, this finish factor will have to be reevaluated after a while; annually or perhaps more often.  This factor also partially addresses the suggestion of awarding win streaks.  HW is the most difficult division to put a streak together in, which is one reason why Fedor is so special. This data was compiled using sherdog.com, ufc.com, and wikipedia.com and is comprised of UFC events from June '06 to March '10 for HW through LW and WEC events over the past four years for FW and BW.  This data was the most easily accessible and sortable and happens to be most relevant to the fighters on the P4P list.  It can be viewed here.

To get the total adjustment for each division we simply multiply the two factors together.  In general this factor is multiplied by a fighter’s total score in order to get their adjusted score. 

4. Another suggestion was that fighting up in weight should earn you more when you win and cost you less when you lose.  I couldn’t agree more.  After all, it is very difficult to do.  So, if a fighter wins in a weight class above theirs, then their score for that fight is multiplied by the sum of the divisional adjustments ranging from the lighter to heavier division.  For example, when Anderson Silva defeated #4 Forrest Griffin at LHW he earned 22 x (0.962+0.765) = 38 points.  For HWs this concept works in reverse.

The calculation is slightly different for a loss.  When Aoki lost to #12 ranked WW Hayato Sakurai he would have lost 10 points under the old system.  Now his loss looks like this -12 x 0.734 x 0.640 = -5.64.  This is Sakurai’s rank multiplied by the divisional adjustments for the LW and WW divisions. 

These adjustments have made for a more sensible and fair set of rankings.  Some people disagree with fighters like Mike Brown and Brian Bowles being ranked over the men that just defeated them.  Please remember that these rankings are not about who was better than who on a given night.  Instead, they represent a fighter’s body of work.  The fact remains that Brown and Bowles defeated a higher ranked set of opponents to get their P4P status.  Try to think of it this way; when Serra knocked out GSP was he the better P4P fighter or just better that night?  Also remember that this compilation is not based on retrospection.  Just give it time, and the cream will rise to the top.  So without further ado, here are the new rankings:

As you can see, GSP still reigns supreme, but Anderson Silva is close behind.  Those of you who saw the first version of these rankings should notice that Silva has moved up drastically.  The new division factors and reward for fighting up in weight have given him a more deserving rank.  Hopefully his antics in his fight against Maia haven't spoiled a potentially great fight between the #1 and #2 P4P fighters in the world.  Penn's loss has bumped him out of the top ten, while the man who defeated him makes his first appearance on this list at #34.  In recent fights Frankie has lost to an un-ranked (at the time) fighter then beaten numbers 6 and 1.  His score breaks down like this (-25+20+25)*(0.640)=12.80.  It remains to be seen if he can put together a string of title defenses against top competition as Penn did.  Up next is Strikeforce Nashville which should be fantastic because it features the #8, 14, 16, and 38 ranked P4P fighters.

As before, I am open to constructive criticism.  Obviously I don't care to hear from people that don't value rankings.  I would like to hear if anyone has a good argument in favor of awarding more points for a TKO, KO, or Sub versus a decision.  My general stance is that all methods of winning are equally valid and that to award one over another is playing favorites.

Rules (updated):
1. Factors based on divisional experience and finishing rates will be applied to all divisions to account for the difficulty of competing in a particular division.
1a. The experience factor is calculated by dividing the sum of all fights in a particular division by the sum of all fights in the division with the most total fights.
1b. The finishing factor is calculated by dividing the percentage of finishes in a particular division by the percentage of finishes in the division with the most finishes per fight.
1c. These two factors multiplied together equal the division adjustment (DA).

2. Beating a fighter ranked in the BE top 25 earns (26 – r)(DA) points, where r equals the losing fighters rank in the division the match was.  Beating a fighter outside the top 25 earns you nothing.

3. A fighter’s optimal weight class is defined as the division they have the highest BE rank in or the division they have the most wins in over the past two years.  This can change.

4. Beating a fighter outside your optimal weight earns (26 - r)(DA1+…+DAn) where DA1 equals your division adjustment and DAn equals your opponents division adjustment.

5. When losing to a fighter ranked in the top 25 in your division you lose r*DA.

6. When losing to a fighter outside your optimal weight you lose r*DA1*DAn

7. When losing to a fighter outside the top 25 you lose 25*DA.

8. Fighters do not carry their divisional rank from one weight class to another.  They have separate ranks in each division.

9. Between two fighters with the same score, the fighter with the more recent win is ranked ahead.

10. All methods of winning have equal worth.

11. Inactivity has no bearing on this list.

- - - - - - - - - - - - - - - -

Here are the rankings without the "Experience Factor" which has been debated in the comments below.

* Some additional research on the Finish factor:

There are only 5 HWs that have never been (T)KO’d before and all of them are relatively new to the sport.  Of the 20 HWs that have been (T)KO’d, 13 of those have been very recent and 7 of those have suffered multiple (T)KO’s in recent fights.  This is all top tier competition.

Now compare to BWs.  There are 9 top 25 BWs that have never been (T)KO’d 6 of which are in the top 10.  Of the remainder that have been (T)KO’d, 11 have been closer to the start of their career when it happened.

It seems pretty clear that the top tier HWs are able to (T)KO each other with more regularity than the top tier BWs.  As you go up the divisions the trend generally looks like this /