Tweet
At UFC on FX 5, we had three fights where the fighter who won the first round got finished in the second. Since this is generally an expensive outcome for me live-betting-wise, I thought I should do some research into how often this kind of thing happens. More generally, I decided to look at what happens after one fighter wins the first round, more or less to the agreement of all observers.
The most convenient measure for decisiveness I could think of was: unanimous opinion (3 out of 3) among Sherdog judges. I looked at all their records from 2009 to the present, which meant only main-card fights from the first two years, since Sherdog didn't start using 3 judges for all fights until more recently. I looked only at 3-round fights. I found 338 fights where all 3 judges agreed on a round 1 winner. Their outcomes broke down as follows:
Fights where Fighter_A wins round 1:
Fighter_A wins a decision: 183 times (54.1%, -118)
Fighter_A loses a decision: 42 times (12.4%, +705)
Fighter_A wins inside the distance: 67 times (19.8%, +404)
Fighter_A loses inside the distance: 44 times (13.0%, +668)
Draw: 2 times (0.6%, +16800)
We can also put together some lines for:
Fighter_A wins (draws push): 74.4%, -291
Fight goes the distance: 67.2%, -205
If the remaining rounds were decided by coin-toss, a fighter who wins round 1 would be expected to lose about 25% of the decisions (losing both tosses). What we see is that they actually lose about 19% of them, suggesting that instead of having a 50% chance of winning a given remaining round, they actually have about a 57% chance. We also see that when the fights finished inside the distance, the round 1 winner is the guy getting the finish about 60% of the time. So in a couple of key stats, probability of winning a round and probability of being the one who wins a finish, we can say that "Sherdog unanimity" in winning round 1 statistically takes a guy from being a 50% fighter to a 57-60 percent fighter.
Something else I looked at was the question of what happens when the first two rounds are split. Of course, there is a lot less data for this. I found 71 fights where all Sherdog judges gave the first round to one fighter, and then all three gave round 2 to the other.
Fights where Fighter_A wins round 1 and then Fighter_B wins round 2:
Fighter_B wins a decision: 31 times (43.6%, +129)
Fighter_B loses a decision: 27 times (38.0%, +163)
Fighter_B wins inside the distance: 10 times (14.1%, +610)
Fighter_B loses inside the distance: 1 times (1.4%, +7000)
Draw: 2 times (0.6%, +3450)
Also:
Fighter_B wins (draws push): 59.4%, -146
Fight goes the distance: 84.5%, -545
I was actually thinking the advantage in favor of Fighter_B might be higher. His odds of winning the final round if it goes to the judges is 53.4%, but hard to call statistically better than even. Oddly enough though, the round 2 winner has gotten almost all of the finishes. The only exception was Roland Delorme's win over Josh Ferguson, and a quick glance at FM shows that round 2 was very competitive, if not in Delorme's favor. [http://hosteddb.fightmetric.com/fights/index/3753] If anyone can think of a 3-rounder where a fighter won the first, lost the second, and then finished their opponent in the 3rd, I'd be curious to know.
The most convenient measure for decisiveness I could think of was: unanimous opinion (3 out of 3) among Sherdog judges. I looked at all their records from 2009 to the present, which meant only main-card fights from the first two years, since Sherdog didn't start using 3 judges for all fights until more recently. I looked only at 3-round fights. I found 338 fights where all 3 judges agreed on a round 1 winner. Their outcomes broke down as follows:
Fights where Fighter_A wins round 1:
Fighter_A wins a decision: 183 times (54.1%, -118)
Fighter_A loses a decision: 42 times (12.4%, +705)
Fighter_A wins inside the distance: 67 times (19.8%, +404)
Fighter_A loses inside the distance: 44 times (13.0%, +668)
Draw: 2 times (0.6%, +16800)
We can also put together some lines for:
Fighter_A wins (draws push): 74.4%, -291
Fight goes the distance: 67.2%, -205
If the remaining rounds were decided by coin-toss, a fighter who wins round 1 would be expected to lose about 25% of the decisions (losing both tosses). What we see is that they actually lose about 19% of them, suggesting that instead of having a 50% chance of winning a given remaining round, they actually have about a 57% chance. We also see that when the fights finished inside the distance, the round 1 winner is the guy getting the finish about 60% of the time. So in a couple of key stats, probability of winning a round and probability of being the one who wins a finish, we can say that "Sherdog unanimity" in winning round 1 statistically takes a guy from being a 50% fighter to a 57-60 percent fighter.
Something else I looked at was the question of what happens when the first two rounds are split. Of course, there is a lot less data for this. I found 71 fights where all Sherdog judges gave the first round to one fighter, and then all three gave round 2 to the other.
Fights where Fighter_A wins round 1 and then Fighter_B wins round 2:
Fighter_B wins a decision: 31 times (43.6%, +129)
Fighter_B loses a decision: 27 times (38.0%, +163)
Fighter_B wins inside the distance: 10 times (14.1%, +610)
Fighter_B loses inside the distance: 1 times (1.4%, +7000)
Draw: 2 times (0.6%, +3450)
Also:
Fighter_B wins (draws push): 59.4%, -146
Fight goes the distance: 84.5%, -545
I was actually thinking the advantage in favor of Fighter_B might be higher. His odds of winning the final round if it goes to the judges is 53.4%, but hard to call statistically better than even. Oddly enough though, the round 2 winner has gotten almost all of the finishes. The only exception was Roland Delorme's win over Josh Ferguson, and a quick glance at FM shows that round 2 was very competitive, if not in Delorme's favor. [http://hosteddb.fightmetric.com/fights/index/3753] If anyone can think of a 3-rounder where a fighter won the first, lost the second, and then finished their opponent in the 3rd, I'd be curious to know.
Comment