How did you feel when Daniel Cormier successfully defended his UFC light heavyweight title against Alexander Gustafsson a few weeks ago? Were you happy or sad? Buying the next round of drinks or cursing at the TV? Were you one of the people who scored it close or at least had some rounds hard to call?
The official, finalized data for UFC 192 came in last week, giving MMA's quantitative official, RoboJudge (RJ), the opportunity to see just how close a fight we had based on typical judging tendencies. For more on RJ, see this piece analyzing all 248 UFC decisions last year. While the goal of doing monthly updates hasn't happened, there should be another annual review coming in January.
Most people seem to agree that DC clearly took the first and fifth rounds so there's no need to waste time on their stats. The real judging action is in the middle rounds, so here are their statistical summaries. Remember, RJ looks at more information than this. These charts are just a quick summary of the most important statistics.
In round 2, DC won the battle to the head with Gus taking the body and legs. DC had a few nice moments of hard uppercuts while Gus earned the largest crowd roar with what looked like a takedown but didn't seem to make the official cut from FightMetric (he had a more complete takedown later in the round). It wasn't a runaway round by any stretch of the imagination. RJ gave Gus a slight edge at 54.3%. I scored the round for Gus, too, as did all three official judges.
Round 3 was the "what do you count more" round. There was back and forth throughout and Gus landed a nice knee in the first minute, but I don't see how DC isn't the clear winner through the first 4:15. Then comes the knee and strike combo that drops Cormier. It's absolutely the most effective moment of the round and Gus lands three more strikes while DC's on the ground getting up. But that's pretty much it. There's nothing after those seven seconds except Gus getting some clinch control and DC landing a few light strikes.
So what counts more, seven seconds of the single most effective striking moment or 4:15 of more effective striking? Even though Gus had a knockdown, it wasn't, as a whole, that Jon Jones-esque turning point that would steal a round in my book. I gave the round to DC and RJ says to basically flip a coin. At 50.7% for DC, we should expect a lot of disagreement and that's exactly what we got. The official judges were split along with most everyone watching the fight.
Round 4 was a tough one. I scored it for Gus. He felt more active which, as it turned out, was statistically untrue. Gus was the more active jabber, landing and missing, but DC was far more active with power shots. He threw 15 more power shots to Gus' 6 more jabs. While 14 of DC's extra power shots missed, who knows if any of them affected the judges. One judge went with DC, the other two with Gus. RJ has this round at 60.4% for DC but if we take away how judges tend to value missed strikes, it's closer to a coin flip.
Here's the overall RJ report.
At 99.5 and 96.4%, DC was the clear winner of the first and fifth rounds as long as Adalaide Byrd wasn't in the building. It was the middle rounds where a reasonable case could be made for Gus every single time. If Sal D'Amato changes his round 3 score to Gus, we've got a new light heavyweight champion. That feels pretty damn close, but this is after-the-fact closeness.
Another way of looking at closeness is when the round 5 bell rings, what are your odds of winning the fight? The further they are from 50%, the less close the fight.
From this perspective, judge's scorecards are about odds and even though three rounds may feel close, it doesn't necessarily make for a close fight, especially when one guy clearly wins the other two rounds. All DC needed to do was get just one of the middle rounds from at least two judges to take the belt home - an enormously strong statistical position.
Put another way, Gus had to win all three of the middle rounds from at least two judges, and he's basically a coin flip in any given round. Take a coin, pick heads or tails and flip it three times in three separate sequences. See if you pick all three flips correctly in at least two of the sequences. It's not an easy task.
The instant the round 5 bell rings, DC has an 88% chance of winning on an individual judge's scorecard and a 96% chance of winning overall. While he didn't get the unanimous decision that would be expected, he ended up taking the fight on 2-of-3 scorecards in the 28% part of the outcome distribution.
This fight may have seemed close after the fact and there's definitely fodder for debate about how rounds should've been scored. But from a judging odds perspective, Daniel Cormier owned Alexander Gustafsson that night.