Unfortunately for this blog, the bad news probably outweighs the good, so I should probably start with that. My schedule has gotten a lot tighter recently, and as a result, I’ll have far less time to work on this blog. Since I already put a lot of time into working on and maintaining SILVA, I’m going to have to figure out how to find time to keep this up as well as I would like. The absolute worst case scenario would mean I abandon this project entirely; that’s a very dire scenario that will only happen if I feel I am no longer making any sort of positive contribution to the MMA blogosphere.
The much more likely scenario is one that sees me simply reduce the amount of content I produce. I’ll probably have to restrict previews to UFC main card fights, and there will only be one preview post for each UFC event that takes place, with shorter previews. A standard five-fight main card might now mean a single 1500-word post to break it down. I’ll do my best to put as much good stuff into those 1500 words as possible.
Another scenario would have me abandon SILVA, which still takes a lot of time to update and maintain. While this would allow me to continue writing a good amount of actual blog content, SILVA is the only thing that really separates me from everybody else. Maybe I’ll find a way to make updating SILVA more efficient. We’ll see.
What you can expect in the near future is that I’ll continue to post SILVA predictions for UFC, Strikeforce, and Bellator fights, and I’ll preview UFC main card fights and hopefully write the occasional commentary piece as well. Ultimately, what I want to accomplish with this blog is provide a unique perspective on the sport of mixed martial arts, one which hopefully advances our understanding of the sport. For now, I’m still around and committed to accomplishing that goal.
Now, the good news: I’m in the process of implementing SILVA 1.2. For those who are unfamiliar with SILVA, SILVA is my mathematical rating system that estimates how good MMA fighters are using ONLY wins and losses as data. SILVA does this by looking at a fighter’s last ten fights, and estimating both the fighter’s quality of wins (Fight Level) and ability to win on a consistent basis (rAP).
In SILVA 1.1, fights against opponents with less than five fights on their record were thrown out entirely. I have to thank Magnus Cedenblad for illustrating the flaw in this approach. Cedenblad entered his UFC debut against Francis Carmont with a 10-3 overall record. However, because six of Cedenblad’s 13 fights were against opponents with fewer than five fights on their record, only seven of Cedenblad’s fights were rated by SILVA. Unusually, all three of Cedenblad’s losses were to inexperienced opponents, meaning they were all thrown out by SILVA and not rated at all. That means that Cedenblad was rated as if he was undefeated! His resulting SILVA score of 61.65, which put him alongside fighters like Michael Bisping, was clearly wrong.
SILVA 1.2 addresses this problem by applying a penalty for each loss thrown out by SILVA in a fighter’s last ten fights. For Cedenblad, this means a “new” SILVA score of 42.85, which I believe is much closer to a proper estimate of his abilities as a fighter. Sorry Magnus.
The other change that’s been made, and I’ll admit to doing this purely in anticipation of this Saturday’s fight, is to throw out all fights ending in disqualification. I don’t have large quantities of hard data justifying this move, just logic. I don’t think disqualifications are any sort of decent predictor or descriptor of how good a fighter is. Whether it’s Jon Jones being disqualified against Matt Hamill or Anderson Silva being disqualified against Yushin Okami, all that counting such fights does is distort an otherwise good estimate of a fighter’s abilities. Jon Jones’s SILVA score increases from 93.88 to 100.48 as a result.
So stay tuned, this blog is still alive for now, it’ll just have reduced content.