UFC 153 / Odds vs. Ratings

If you’ve been following recent updates, most notably those associated with the Upcoming Events page, you already know that we now display a meta-odds feature — average odds from about a dozen sports books.

I thought it would be interesting to compare results of UFC 152 and how we stacked up versus the gamblers.  We ended up doing pretty good , though this is not an attempt to suggest the rating system is more accurate than gambling odds.  Gamblers can analyze each fight individually, while the rating system uses a variety of mathematical routines to supply a primary rating that is primarily focused on ranking recent achievement, with a secondary priority of gauging future expectation.  Either way, both systems have their place in determining who to expect to win the fight.

Back again, to see if we could repeat the success with UFC 153… and here we go.

Note: Moraes/Forte and Marcello/Madadi are excluded as at least one fighter in each was unranked/inactive.

Ratings vs. Gamblers Scoreboard

^ – Pre-fight divisional difference was not factored into rating difference.

First things first, in the fights and ratings included above, the rating system had ALL of the winners pegged.  However, in the event of Camozzi/Cane, Cane was due a provisional rating increase from dropping to 185 that would’ve put Cane slightly ahead, though this rating increase technically did not transpire as a unique wrinkle of the system requires he have at least one win within a third of his rating in the past 1080 days in that division, which he still does not.

Technical jargon aside, I had to split hairs in a few of these matches to declare a winner.  One of these was Nogueira/Herman.  The rating advantage over Nogueira is comparable to him being a much higher favorite than -241 and a submission victory in my opinion, is closer to the advantage that the ratings gave him versus the gambling odds.  The same thing goes for Davis/Prado.  Although Prado was undefeated with all dominant wins (one of the few situations in which I would not trust the rating system entirely), the rating of Davis was more than 10 times his.  In this fight, it sure played out that way.

I gave only the Bezerra/Sicilia and Brandao/Gambino advantage to the gamblers because I felt the scoring of the contests was more indicative of where the odds ended up, but in all honesty, I did not watch either fight.  As for ties, you could lean toward giving the Tibau/Trinaldo and Teixeira/Maldonado ties to the ratings, but I felt doing so may show bias.  Silva/Bonnar was accurately represented by both.