DB 8

Freakonomics Man 1: People who are being corrupt are always trying to actively cover their trail. So corruption is, by its

nature is hard to identify; hard to prove. Murder is really great, cause almost always when someone’s murdered there’s a corpse.

Man 2: You might say "How would you know someone's cheating?" and the answer is – well it’s in the

data. I don't have to ever have seen a sumo match. I can go in the data, I can look at it, and I can tell you, with almost complete certainty, that there was rampant cheating going on.

[Muffled announcer and crowd noise.]

Narrator: In professional Sumo tournaments, the wrestlers fight one bout per day for 15 days.

Man 3: If you win 8 out of the 15 matches, you can move up in rank - half a slot. The difference of half a

rank can be maybe 5,000 dollars in paycheck a month. The respects you get in the Sumo Association. So when you talk about stuff like that, that 8th win is real critical.

Narrator: Arikshi entering a tournament’s final 15

th match with a 7 and 7 record, has far more to gain from

a victory than an opponent, with a record of say 8 and 6 has to lose. If a wrestler has 8 wins under his belt, he's guaranteed to advance. Even if he loses that last match. So he can afford to take a fall. In Japan, there's a term for match rigging – Yaocho. Many suspected that Sumo matches might be rigged. But it is nearly impossible to prove, unless you look closely at the numbers.

Man 2: Two wrestlers that I would expect to have an even match, when 1 of them needs the eighth win and the other doesn't, the one who needs it wins 75% of the time rather than 50% of the time; that is a huge deviation.

Man 1: I (8-6 wrestler) let you (7-7 wrestler) win this deciding match cuz you, my friend, are gonna fall down the pyramid if you don't. In return, the next time those 2 guys meet lo and behold, the 8-6 wrestler almost always wins those matches.