Notable Information

• Code updated to 1.15b.  Re-worked technical decision outcome consideration. Fixed bug in draws which could occur in fights w/ huge gap in points.  Reduced extreme upset protection, relying more on past results then prediction.
• Effects from shows in past week, as well as old shows recently added.

Rankings have been updated on March 27, 2008

Notable Info

• Rating code updated from 1.13c to 1.14b.  Changes related to situations in which fighters come back from inactivity.  Especially those who have had good performances in their 1st and 2nd bouts back.  This has caused moderate reshuffling throughout the divisions, but leaves the top 25s relatively untouched.
• Shows from the past week (WEC, Pancrase), as well as old bouts have been added to the database.

Expect the next update to come on Sunday.

Information

• WEC 33 results, as well as recently added bouts.
• Updated basic code for properly placing fighters coming back from inactivity.  This will cause minor effects throughout the divisions, especially for those who had recent instances of inactivity, as well as their recent opponents.

Rankings posted.  Data from March 23, 2008.

Notable Changes

• Ratings code updated to 1.13c. GTR modification, support for technical decisions/draws (rare), code cleanup.
• Rankings expanded to Top 100 for Lightweight through Heavyweight.
• Rankings expanded to Top 50 for Bantamweight, Featherweight, and Women’s.
• Rankings expanded to Top 25 for Flyweight.
• Fighter names are now hyper-linked and directed to the fighter’s Sherdog profile.
• Approximately 250 bouts added since last update for shows in March 2008.
• Approximately 160 bouts added since last update for shows prior to March 2008.

Now that the weight classes are expanded so far in the thinnest divisions (Women, Bantam, Fly), it is almost a certainty that we are mistakenly excluding fighters toward the bottom of the lists due to lack of division assignment.  If you feel there is a fighter that should be in one of the lists, please feel free to ask us for their current division assignment for verification.

Its a slow week… but we will be releasing an update on Sunday.

Expected Changes

• Expanding Lightweight through Heavyweight to Top 100.  There should be expansions for the smaller divisions as well.
• Fighter names will now be hyper-linked to their Sherdog profiles.
• Code updated to v1.13.  Only change is an update to the GTR initiative mentioned in the previous post.  The effects will be minor and limited to the lowest weight divisions.
• Approximately 250 old bouts added to the database.
• A few minor recent shows added to the database.

As you may or may not know, the PRIDE and UFC titles have been acknowledged by our rating system and have been given precedence, or “double verification” that a fighter has been a top fighter in a top brand of MMA.  We hand selected those organizations due to the quality title defenses, quality of reigns, and overall quality of fighters in general.  Its really a combination of a half dozen things.

Those participating in these title contests get very small boosts proportional to their performances, and extended exemptions from quality performance decays.  The downside to this is when they are no longer participating on the title level, they are subject to falling down the ranks more quickly.

Through my experiences in computerized rating/ranking systems, I have noticed that when there are different brands with sometimes vast, yet unquantifiable differences in talent pools, there needs to be a method for planting an artificial seed or subjective variable to represent that “little extra” to show that the lineage, prestige, and other intangibles that exist in the dominant brands.  Furthermore, this also acts as a defense against the smaller organizations pooling points within their fighters, hence the start of the GTR initiative.

After looking at the global MMA scene, its really split into two distinct markets where fighters don’t crossover THAT much.  Those two markets are Asia, and Europe/North America.  Our goal is to have a quality title, for each weight division in each market, at any given time.  If none exists, it will simply be vacant.  Currently, the GTR provisions are acknowledged for the following titles:

Fly – Euro/N.America: Vacant / Asia: Shooto

Bantam – Euro/N.America: WEC / Asia: Shooto

Feather – Euro/N.America: WEC / Asia: Shooto

Light through Heavy – Euro/N.America: UFC / Asia: Vacant (was PRIDE)

The effects of the GTR will be seen in the next update, but will only cause mild changes in the lower weight divisions.

MMA rankings have been updated to include results from this weekend’s DREAM & Icon events.

We also decided to make the Division Dominance List a permanent part of FightMatrix.com!  This list will be updated with every rankings release.

This tournament will without a doubt, reshape the Lightweight rankings.  The rankings below are current, with the latest code, which shouldn’t be much different than the live rankings.

Some of the fighters are matched much harder than others!  So much for seeding…

Lightweight GP

[#2] Gesias Calvancante vs. [#6] Shinya Aoki

[#10] Tatsuya Kawajiri vs. [#65] Kultar Gill

[#4] Mitsuhiro Ishida vs. [#225] Bu Kyung Jung

[#77] Kazuyuki Miyata vs. [NR] Luiz Firmino

[#31] Andre Amade vs. [#39] Eddie Alvarez – Projected LW ranking for Alvarez

[#50] Artur Oumakhanov vs. [#151] Katsuhiko Nagata

[#17] Joachim Hansen vs. [#34] Koutetsu Boku

Anyone know why there aren’t 8 bouts??

Ranking Note: Gray font represents an approximate ranking. An asterisk represents a fighter who is ranked, but too low for an approximation. These two scenarios will clear up slightly as we move forward. “NR” means the fighter is not ranked, due to inactivity, or due to a rating that is at or below the default starter rating.

Ok, so that quality performance thing I said I fixed… its not fixed 100%. Its way better than it used to be, but still not perfect. The update to 1.12 will have very minor effects and will be launched with the new update on Sunday. This update will also include results from the shows on Saturday which should result in huge changes!

There have been a few more updates than usual lately, but it is necessary as the rating system matures and evolves. These minor updates do not affect the rankings of those which are decided by more than a few points, only those who whose positions are “toss ups” and those who are affected by particular situations.

Furthermore, in the near future, we will be looking to expand the division rankings once more. We will be expanding Lightweight through Heavyweight out to 100! Featherweight and Women’s should be OK to span to 50. Flyweight will likely stay at 10 for now, and Bantamweight is still up in the air.

As an experiment, I decided to create a “division dominance” list with the current rating set.  At its core, it is a hybrid of a pound-for-pound list and a current snapshot of the divisional rankings.  It basically says, which fighters are the most dominant over their own respective divisions.

Division Dominance List

Rank	Fighter	Record	Division Dominance Points
1	Anderson Silva	21-4-0	258
2	Fedor Emelianenko	27-1-0	255
3	Quinton Jackson	28-6-0	226
4	Georges St. Pierre	15-2-0	215
5	Randy Couture	16-8-0	198
6	Norifumi Yamamoto	17-1-0	197
7	Antonio Rodrigo Nogueira	31-4-1	186
8	Yasuhiro Urushitani	14-3-5	185
9	Urijah Faber	20-1-0	176
10	Miguel Torres	17-1-0	172

If you like this list, come in the forum and let us know!

Since the quality performance decay (QPD) is now working as intended, I figured I would do a quick write-up with regards to its purpose.

In short, the QPD is meant to gradually decrease the rating of a fighter who has not necessarily been a victim of losing, or inactivity, but one who has not recently had a performance that’s indicative of a fighter of that rating. It is important to note that a quality performance is relative to the fighter’s rating.

Without digging too deep into the decay, in order to register a quality performance, a fighter must beat an opponent at least 1/3* of their pre-bout rating, draw against an opponent at least 1/2* of their pre-bout rating, or lose a split/majority decision against an opponent at least 2/3* of their rating. *Values used are approximate.  Note: UFC/PRIDE title bouts have/had special considerations.

Please remember, that even if a fighter draws or loses a close decision to a lower rated fighter and registers a quality performance, they will still lose rating points as a result of the bout’s outcome.

If a fighter has went 360 days without a quality performance, they go into a mode of QPD. This decay is not nearly as strong as the inactivity decay, but is a bit less forgiving than a fighter coming back from inactivity as an inactive fighter can (in some scenarios) redeem a bit of their lost rating by performing well against a quality opponent in their comeback bout. The reason for that forgiveness is that we can gauge a fighter in QPD, but not one who is inactive.

Once a fighter in this QPD registers a quality performance, they are taken off of the decay mode. It is also important to note that since a quality performance is relative, as their rating slips, the requirements on breaking free from the decay are lessened.

The QPD is one of the main advantages of the rating system. It is gradual, and no fighter is immune. Many ranking sites may forget about certain fighters who have been bottom-feeding, or simply leave them at their position due to prestige. They may also over- or under-correct these fighters. This is not the case here. The decay treats everyone equally, as it is gradual, and proportionate.

Notable fighters currently in the QPD are: Fedor Emelianenko, Josh Barnett, Mirko Filipovic, Wanderlei Silva, Matt Lindland, Ricardo Arona, Antonio Rogerio Nogueira, Frank Trigg, Takanori Gomi, and quite a few others.

Ratings as of March 11, 2008

Notable Changes

• Rating code updated to 1.11 which includes the correction of quality performance penalty application as well as the strengthening of the penalty rate.
• Approximately 100 old bouts added to the database.

The overall changes to the rankings are very minor.