Subscribe
Posted on December 25, 2008 by jcs
Our FightMatrix database shows the following rematch-related statistics:
- 2,307 bouts have taken place where the fighters have fought at least once previous.
- 196 bouts where fighters have fought at least twice previous…. and 24 bouts, at least three times previous.
- Two pairs of fighters have fought each other six times! (Travis Fulton & Rory Prazak, Brian Robinson & Nate Kopel)
In instances where fighters meet in their second encounter, the same winner has prevailed only 64.5% of the time. If the first fight ended in a knockout or submission, this number increases to 67.7%. These statistics prove that MMA is a very volatile sport compared to boxing, where the percentage of repeat winners after a knockout in the first bout since 1985 is almost 90% (I researched this myself).
Not many rematches have taken place after the first bout ended in disqualification. In fact, its very rare, as we only have record of 19 said rematches. The previous winner won 9 out of these 19.
How about when a fighter wins the first two matches by KO or TKO, and then rematches the loser for a third time? In other words, does Quinton stand a chance on Saturday? The dataset is extremely small, but the statistics show that the KO/TKO winner of the first two bouts, will win the third bout 89% of the time.
Posted in MMA Statistics | Comments (1)
December 29th, 2008 at 6:04 am
JCS, these are great statistics. I’ve always thought that statistics like these could be the key to creating an alternate type of rating system. This type of data describes, in statistical terms, what a “decision win” means and what a “KO win” means.
Its also interesting that the data so far is pretty consistant with probability theory. For example, if fighter A loses to fighter B by KO in their first fight, then the data suggests that fighter A only has a 32.3% chance of winning the rematch. If fighter A loses by KO to fighter B in their second fight, then probability theory would predict that fighter A has a 0.323 x 0.323 = .104 = 10.4% chance of winning the third fight. In your data, you found that fighters who lost by KO in the first 2 fight won the third fight 11% of the time. This is very close to the 10.4% that probability theory predicts.
If you plan to do further research into this, one suggestion I have would be to also look at stats involving 3 fighters. If fighter A beats fighter B and fighter B beats fighter C, then what’s the probability of fighter A beating fighter C. These types of statistics might give further insight on how “styles make fights”.
Great research. Thanks!