- Joined
- Mar 5, 2016
- Messages
- 8,349
- Reaction score
- 4,048
Round 3 seems to be the deciding round. I was pretty surprised to see all 3 judges give it to Ankalaev. I thought Pereira won.
Pereira landed more strikes in round 3 when you look at the numbers. Seems like recency bias to me. All rounds should be scored in a void with no prior knowledge of the previous rounds. This is obviously difficult for humans, which is why judges have extensive training to avoid bias.
For those who don't understand, recency bias (in this context) is when a fighter dominates or does very well in a round (like Ankalaev did in round 2), and therefore you are already more likely to score rounds for him before the next round even begins. If the following round is close, amateur incompetent judges will always score the following round for the winner of the previous round, due to recency bias.
There's something similar in mathematics called the "halo effect" where a professor that gives a student a 0 on one test question is far more likely to give a 0 on the next question. For this reason, professionals scoring math tests actually score question 1 on each test for every student, before moving onto the 2nd question scoring on every test, and so on, so they have no memory of how each student did on the previous question when scoring the next question.
Pereira landed more strikes in round 3 when you look at the numbers. Seems like recency bias to me. All rounds should be scored in a void with no prior knowledge of the previous rounds. This is obviously difficult for humans, which is why judges have extensive training to avoid bias.
For those who don't understand, recency bias (in this context) is when a fighter dominates or does very well in a round (like Ankalaev did in round 2), and therefore you are already more likely to score rounds for him before the next round even begins. If the following round is close, amateur incompetent judges will always score the following round for the winner of the previous round, due to recency bias.
There's something similar in mathematics called the "halo effect" where a professor that gives a student a 0 on one test question is far more likely to give a 0 on the next question. For this reason, professionals scoring math tests actually score question 1 on each test for every student, before moving onto the 2nd question scoring on every test, and so on, so they have no memory of how each student did on the previous question when scoring the next question.
Last edited:
