In his piece about the growing concern over MMA refereeing, Dave Meltzer slipped this paragraph in:
Fighters often feel pressure trying to please the live crowd, wanting to be perceived as having quality fights and become more marketable in a community that is controlled by a zealous Internet fan base.
“Look at [Jorge] Gurgel,” referee Yves Lavigne said of a recently released UFC vet. “He’s a Brazilian jiu-jitsu black belt, and he’s very good on the ground. He never even shoots to try a takedown.”
During the period in question, UFC held 60 matches, 32 of which were stopped by a referee call after a knockout or a TKO compared with nine submission finishes. That’s a 42 percent drop in submissions and a 36 percent increase in knockouts and TKOs over the same 12-week period one year ago. This means a larger number of split-second decisions are being made as to whether fighters can continue, and those decisions have a higher percentage of likelihood of affecting outcomes and careers.
For starters, I'm interested in why Dave chose the same time frame from the year before, instead of the previous 60 fights, 60 fights from 2003, 60 fights from Dream, or the past 60 WWE matches. The point being, what use is there in comparing a random cluster of 60 fights to another random cluster a year prior? This isn't retail, there aren't patterns based on the day, week, month, or season.
I want to point something out before going further. In my two articles about this topic, I shouldn't say that we definitely aren't seeing a trend of increased knockouts with decreased submissions. We might be, we might not be. I'm trying to make two points. 1) Gross and now Meltzer have used faulty logic getting to their conclusions and 2) as of right now there is not enough evidence to start making claims that the UFC is pushing for more knockout finishes or that submissions are becoming a less viable means of finishing a fight..
With that out of the way, jump over the break to see some fancy graphs.
What you see here is the KO:Sub ratio over the past 3 years, broken up into six month overlapping chunks. This graph trends slightly upward. However, look at these next two graphs:
These two graphs show the KO:Sub ratio year-to-year and over periods of two overlapping years respectively. Notice the trend downward until an uptick at the end.
I still don't think we have enough evidence to support the notion that the increase in KOs:Subs we're seeing are anything more than random fluctuation at this point. If this number continues upwards and ranges around 1.75-2.00 KO:Subs, we have something to talk about. We also just might be seeing a regression to the mean after a period of increased submissions.