MMAth: Predicting Fight Outcomes With Math

What is MMAth? MMAth is the process of using a sequence of past mixed martial arts fight results that provide a logical and objective argument for predicting the result of a future mixed martial arts bout.

A basic MMAth equation:
Tito Ortiz beat Forrest Griffin who beat Shogun,
...therefore Tito Ortiz should beat Shogun (theoretically).

Typically, MMAth uses the greater-than (>) and less-than (<) signs in them, the > meaning "beat" and the < meaning "lost to."

So our previous example could be written:
Tito > Forrest > Shogun,
 ...therefore Tito > Shogun


The Four Basic Rules Of MMAth:

(1) The longer the MMAth equation, the less credibility should be given to the outcome. (2) The greater the time span between the bouts in the MMAth equation, the less credibility should be given to the outcome. (3) If fighters have faced each other before, greater credibility has to be given to past performances versus each other rather than other opponents. (4) When multiple MMAth equations can be formulated that lead to the same outcome, greater credibility must be given to their collective outcomes.

Advanced MMAth Examples:

MMAth as a predictor of cross-promotion fights:
Hughes > Frank Trigg (twice),
Frank Trigg > Ray Cooper,
Ray Cooper > Jake Shields,
Jake Shields > Carlos Condit,
...therefore Hughes should beat Condit.

MMAth as a predictor of cross-weight fights:
Matt Hughes lost to Dennis Hallman,
Dennis Hallman lost to Drew Fickett,
Drew Fickett lost to Edwin Dewees,
Edwin Dewees lost to Joe "Daddy" Stevenson,
...therefore Joe "Daddy" should beat Matt Hughes (theoretically)

MMAth as a predictor of cross-promo and cross-weight combined:
Fedor lost to Tsuyoshi Kohsaka,
Kohsaka lost to Babalu,
Babalu lost to Kevin Randleman,
Kevin Randleman lost to Bas Rutten,
Bas Rutten lost to Ken Shamrock,
Ken Shamrock lost to Royce Gracie,
Royce Gracie lost to Matt Hughes,
Matt Hughes lost to Dennis Hallman,
Dennis Hallman lost to Drew Fickett,
Drew Fickett lost to Edwin Dewees,
Edwin Dewees lost to Joe "Daddy" Stevenson,
Joe "Daddy" Stevenson lost to Josh Neer,
Josh Neer lost to Spencer Fisher,
Spencer Fisher lost to Hermes Franca,
Hermes Franca lost to Yves Edwards,
Yves Edwards lost to Mark Hominick,
Mark Hominick lost to Rani Yahya,
Rani Yahya lost to Chase Beebe,
Chase Beebe lost to Miguel Torres...

...therefore WEC Bantamweight Champ, Miguel Torres should beat former PRIDE Heavyweight champ, Fedor Emelianenko (theoretically)

** Obviously, the above MMAth equation is impossible to believe (Torres could never beat Fedor) and violates the first two rules of MMAth.

Fighters of different eras:
Hong Man Choi lost to Fedor,
Fedor lost to Tsuyoshi Kohsaka,
Kohsaka lost to Babalu,
Babalu lost to Kevin Randleman,
Kevin Randleman lost to Bas Rutten,
Bas Rutten lost to Ken Shamrock,
Ken Shamrock went to a draw with Royce Gracie,
...therefore Royce = Hong Man Choi (draw)

** Obviously, the above MMAth equation, while arriving at a not-far-fetched conclusion, does drastically violates the second rule of MMAth, thus little credibility should be given to this argument.


MMAth and Top 10 Rankings:

Though so many of the MMA elitists claim that MMAth does not work, the truth is that even they use MMAth to draw conclusions on Top 10 rankings, and on fight picks for that matter.

In fact, I would argue you CAN NOT produce a top 10 ranking of any worth that does not useMMAth to a large extent. But you have to use MMAth with some subjectivity and with some bias (obviously people fighting in the big shows get better rankings than guys in the smaller shows) to create a credible Top 10 list.

Let's take a look at one popular MMA website's Top 10 Heavyweights list:

1. Fedor Emelianenko (beat Nog and Sylvia)
2. Minotauro Nogueira (beat Sylvia)
3. Andrei Arlovski (beat Werdum and Rothwell)
4. Josh Barnett
5. Tim Sylvia (beat Arlovski twice and Rothwell)
6. Ben Rothwell
7. Alistair Overeem
8. Junior Dos Santos (beat Werdum)
9. Fabricio Werdum (beat Gonzaga twice)
10.Gabriel Gonzaga

The people that put together the above rankings may claim that MMAth was not used to compile this list, but clearly this list reeks of MMAth.

Using MMAth To Predict TUF Winners:

You can use MMAth as a good predictor of TUF reality show outcomes. For example, let's look at the TUF Four season (The Comeback). The eight fighters in the WW division of TUF 4 were: Mikey Burnett, Shonie Carter, Matt Serra, Rich Clementi, Pete Spratt, Chris Lytle, Din Thomas, Jeremy Jackson.

One could logically have made the following argument before the TUF Four tourney:

Din Thomas> Jens Pulver > Joe Stevenson > Jeremy Jackson > Shonie Carter > Chris Lytle > Pete Spratt > Rich Clementi > Pete Spratt > Robbie Lawler > Felaniko Vitale > Dave Menne > Carlos Newton > Pat Miletich > Mikey Burnett

The bolded names in the above equation were TUF Four participants. This equation shows how Din Thomas should beat six of the seven men in the TUF Four welterweight tourney. The seventh man (not included in the equation) was Matt Serra, but Din already held a win over Matt Serra before the show.

So a logical argument could have been made for Din Thomas to win the TUF Four show. Obviously, in hindsight, he did not win the show. But MMAth can never be 100% accurate.

The MMAth Paradoxes:

The MMAth Loop:
An MMAth loop is a sequence of fighters who have all lost to one of the other fighters in the sequence, and also beat one of the other fighters in the sequence, thus giving you a loop and no clear answer.

Example:
Chuck < Rampage < Wanderlei < Chuck < Rampage < Wanderlei < Chuck < Rampage < Wanderlei... (goes on infinitely)

Simpler way to write an MMAth loop:
Chuck < Rampage < Wanderlei < Chuck > Wanderlei > Rampage > Chuck

Can the MMAth loop ever be used as a predictor?
Lindland > Schembri > Minowa > Baroni > Dave Menne > Vitale > Lindland.
...therefore, Lindland (theoretically) should beat himself (in fact, Lindland did just that when he KOed himself against Vitale).

Variables Unaccounted For In MMAth:

MMAth is powerful predictor of fights when used properly; however, MMAth can never be an absolute 100% predictor in fight outcomes. Their are too many variables, such as:

(1) Luck—everyone gets lucky sometime, (ex. Matt Serra over Georges St-Pierre).
(2) Style clashes—MMAth demanded that Wandy should have beat Chuck, but most observers realized that Chuck had a style advantage.
(3) Injury—fighters often come into a fight injured, or injure themselves during a fight.
(4) Officiating/judging—MMAth can't account for poor or crooked officiating and judging.

So, if you have a MMAth equation that you think is absolutely 100% correct (like everyone thought  St-Pierre > Serra was a lock), then how many percentage points should you shave off that 100% to account for the above variables? I would say at least 20% for every fight. So never assume a fight is more than 80% in favor of one fighter over another.

If you assume that a fighter is never more than an 80% favorite, then anyMMA bet where a guy is a four to one dog or worse is worth placing a bet (I'm not advocating gambling...if you or someone you know have a gambling problem, call Gamblers Anonymous at 888-424-3577).