Predicting boxing upsets

[Jersey Joe]

Ever since James Buster Douglas shocked the boxing world by defeating Mike Tyson in Tokyo, I’ve wondered if there is any way to predict boxing upsets. This isn’t just a passing interest either - Douglas was a 40-1 outsider, so even giving him a 10-20% chance of winning would have made it a very lucrative bet to back him at the bookie. But most fights obviously don’t have anywhere near those odds. Question is, can we actually figure out in advance if an upset is on the cards?

There are some reasons to think so. Firstly, boxing is a “two-horse race” and has the randomness factor of the knockout (or TKO) - this means that even the rankest underdog usually has at least a “puncher’s chance”. It might only be 5%, but it’s still there if the weaker fighter has any kind of KO ability at all. This means that big underdogs can be a good bet if the odds are big enough (e.g. Douglas/Tyson). Secondly, boxing is famous for being a sport where “styles make fights” - in other words, pure ability is not always as important as the matchup between 2 different opponents. If A beats B, and B beats C, that does *not* mean that A can beat C. There have been many fights where a better fighter has been defeated or held to a draw by a weaker opponent who has an “awkward” style for him. So if there is a way to find out the styles of each fighter, and then see which style each fighter is “weak” against, you could wait to be on fights between a favourite and a weak underdog but one who has the style to counteract the better fighter. Finally, boxers are renowned for having primes, and starting to “lose it” after a run of success - whether it’s the slick speedster hitting 32+, or the playboy who slacks off once he achieves his ambition and then comes up against a hungry, angry, hard-training opponent. Closely watching the lifestyle, training regimen, comments, and media reports about a boxer might give some insight here. There are other factors too like ring rust, change of trainer, personal problems, psychology etc.

So, it looks to me like there could be some interesting possibilities. Imagine a situation where a protected but big-name glamour fighter is champ, and some fairly unknown contender fights him. Chances are the unknown would be a big underdog. But let’s say we’ve analysed the glamour boy’s record and see he is a bit chinny and has a problem vs swarmers, then see the up & comer has good power and a swarming style - not only might the underdog be worth a bet, he might actually be the *favourite* in this matchup.

Before I go on, I’d like to ask is anyone here interested in discussing this subject? I'd be especially interested in hearing from anyone who would be willing to work seriously on this.

[Beeston Brawler]

Count me in :good

[Boilermaker]

Ever since James Buster Douglas shocked the boxing world by defeating Mike Tyson in Tokyo, I’ve wondered if there is any way to predict boxing upsets. This isn’t just a passing interest either - Douglas was a 40-1 outsider, so even giving him a 10-20% chance of winning would have made it a very lucrative bet to back him at the bookie. But most fights obviously don’t have anywhere near those odds. Question is, can we actually figure out in advance if an upset is on the cards?

There are some reasons to think so. Firstly, boxing is a “two-horse race” and has the randomness factor of the knockout (or TKO) - this means that even the rankest underdog usually has at least a “puncher’s chance”. It might only be 5%, but it’s still there if the weaker fighter has any kind of KO ability at all. This means that big underdogs can be a good bet if the odds are big enough (e.g. Douglas/Tyson). Secondly, boxing is famous for being a sport where “styles make fights” - in other words, pure ability is not always as important as the matchup between 2 different opponents. If A beats B, and B beats C, that does *not* mean that A can beat C.


ACtually, i think that this theory is very interesting and is a lot stronger than people agree. I think that that A beats C far more than C beats A. Obviously, there are always exceptions, with the Ali, Frazier, Foreman triangle one of the most high profile examples usually given. But i think that theory actually holds up.

Firstly. there will always be random results due to factors such as aging, deteioriation, training or even just bad performances. IN fact, even when A beats B, it does not mean that A beats B. Look at Louis v Schmelling, or the Ali -Frazier fights or countless others where fights can go either way.

If you look at three fighters record, i think you would find that more often than not that where fighter a beats b, and b beats c, Fighter A will beat Fighter C. I think that this can be a huge tool in looking at and predicting upsets.

I also think that the common opponent theory is another which may hold up. If A beats common opponents easier than B then A usually beats B. Again, things such as KO styles may affect this, but i think as a general rule it can be a good guide and may actually help predict upsets, which can tend to be based a lot more on media reports.

[Dempsey1238]

Well to tell you guys the truth, The movement Zab try to low blow Baldomir, I kind of thought if Zab didnt show Bladomir ANY respect in the fight like this, he would lose.

So the first 6 rounds was Zab, I did give Baldomir 2 of the 6 rounds. Than the fame 7th round, and it was Baldmir all the way from 7th to 12, and Zab lost because of his complete disrespect for Baldomir.

Baldomir at the time didnt have the power(Little did we know lol)
Speed,
foot work,
Foot speed,
defense
Jab
He had nothing over Zab out side of guts and a amazing chin. But some how he did it.
I not saying I can say when a upset is happing, but on this ONE fight after the low blow of the intros, I kinda of had this gut feeling a upset was going to happen.

[lefthook31]

Hype can make a fight not what it seems for sure as well. Unless you have insight into a fighters personal life that would be difficult to assess, but all those things you mentioned are true. Tyson shouldnt have lost to Douglas, but if someone was close to Tyson and saw how he was conducing himself, getting knocked down in sparring, amatuer trainers, not training seriously, etc, the writing was on the wall, certainly worth a couple hundred dollar bet.
I remember reading somewhere that odds makers in Vegas will never have a fighter as big of a betting underdog as Douglas was, ever again. They just wont support the fight. I cant remember in recent years anything even close to those odds in a fight anyway. Now 5-1 is more realistic but hardly worth betting if it seems like a one sided blow out. I would have not bet Hatton with the odds of the fight. Its easy after the fight to say you should have.

[janitor]

There is always some old hand who looks at the fight and sees the bigger picture.

Jack Johnson famously picked Max Schmeling to beat Joe Louis.

Louis in turn was one of the few people who genuinely picked George Foreman to beat Joe Frazier (before the fight).

Willie Pep predicted prety much how the rumble in the jungle would go and why.

[Jersey Joe]

Yeah I agree that *most times* A beats B who beats C, A wins over C too. That's because most times A wins because he is simply better, ditto for B. C in most cases is just a much worse fighter.

But that is exactly why, on rare occasions, the odds get distorted, because people draw the wrong conclusion - they think this works *all the time*. The right conclusion is "If A is a great, B is very good, and C is a contender, then A beats B and C, and B beats C". The wrong conclusion is "If A, B, and C are all similarly good, then if A beats B and B beats C, then A is a huge favourite over C". See what I mean? I am comparing fighters on a closer skill level. The ABC theory means people sometimes vastly underestimate a fighter in a matchup.

Regarding the long-outsider thing - after the Douglas fight, HOLYFIELD was for a while a 25-1 long shot vs Tyson. Holyfield obviously being a far superior fighter to Douglas, Tyson having been locked up for 3 years - those odds were UTTERLY INSANE. And it doesn't matter if the bookies think the odds are nuts, because bookies don't set the odds, the weight of public betting money sets the odds. No bookie is big enough to go against a stampede of public bets in one direction. Also now with things like Betfair, you don't even have bookies.

I think maybe I didn't describe my idea very clearly in the first post. Effectively what I'm trying to do is find situations where the betting public thinks one thing, enough to push the odds far in one direction - and then not only are they mistaken, but they are mistaken to a huge extent. To be that wrong, either you need to know some insider information (e.g. that someone has been boozing and partying, and sucks in training), or you need to be much better than the public at interpreting the public information (e.g. realising that 25-1 or 40-1 odds are insane in most head to head fights; or knowing that a world class boxer cannot be a 10-1 underdog in any fight; or that a style matchup actually shifts the odds significantly). The public seems to be on reputation, rather than a real analysis of the matchup ahead.

[Jersey Joe]

There is always some old hand who looks at the fight and sees the bigger picture.

Jack Johnson famously picked Max Schmeling to beat Joe Louis.

Louis in turn was one of the few people who genuinely picked George Foreman to beat Joe Frazier (before the fight).

Willie Pep predicted prety much how the rumble in the jungle would go and why.

Yeah, that's the sort of thing I'm talking about. Granted, some of these might just be lucky guesses. But I suspect that there must also have been something that they saw (e.g. Max Schmeling himself said he knew the flaw in Joe Louis and how to beat him - turned out it was weakness to the counter right hand) which gave them the ability to go against the consensus and predict an upset.

There are a lot of knowledgeable posters on here. We should be able to work out some of the "hidden factors" which result in boxing upsets that the mainstream media & fans get surprised by.

[Mendoza]

Great thread. I have called a few upsets before. My best one was Rhaman over Lewis in the first fight.

Predicting upsets is not easy, but you can have a better chance of sniffing out upsets if you understand boxing. I'll share ten ideas:

1 ) Styles makes fights. Sometimes one fighter has a weakness that plays right into the other fighters strengths. In general the boxer has the edge over the puncher, the puncher has the edge over the swarmer, and the swarmer has the edge over the boxer...all in a style sense of course.

2 ) Pay close attention to outside the ring distractions, and make a best guess if the fighter is really motivated or not.

3 ) You can make a solid guess if a fighter is past his prime. Older fighters, even great ones can be upset by much younger and faster challengers. Other times a fighter who just came off a real tough match ( win or lose ) might lose something might not have his best stuff in his next fight.

4 ) Watch the scales. Some fighters have to drain themselves to make weight, and pay for it in the ring. Other times the fighter is too fat, and his speed, timing and stamina suffer.

5 ) Sometimes a good corner / trainer can help. Check the trainers record in high stakes matches too. Some like Steward are excellent, others like McGrit are below average.

6 ) The ring size can matter. The smaller the better for the bigger and more durable punchers. The larger the better for the fleet footed boxers with good jabs.

7 ) The location of the fight. It the fight in a hostile environment? Is the arena hot ( Bad for older fighters ). Is the fight in a high altitude environment ( Bad for fighters with suspect stamina )

8 ) This one stinks. AHEM, the promoter picks the judges. Picking an upset on a fight that looks to go the distance with a controversial ending can happen.

9 ) Check a fighters experience level. Sometimes a red hot prospect has not been tested, and in such a case a live under dog who also has not been tested will beat him.

10 ) And this is the might be the most important point of all....wait for Teddy Atlas to make a pick, then select the other guy.

[Jersey Joe]

Now 5-1 is more realistic but hardly worth betting if it seems like a one sided blow out. I would have not bet Hatton with the odds of the fight. Its easy after the fight to say you should have.

Yeah, you wouldn't take 5-1 on someone who actually IS a big underdog. You want someone who is a 5-1 betting underdog, but who your analysis shows is actually perhaps an evens bet. Or someone who is a 3-1 dog but you actually rate a slight favourite.

The fact is, like any betting, 95% of fights you simply won't bet on because the odds will not be far enough out of line. You need to wait for the most favourable opportunities, and not touch anything else.

[Jersey Joe]

I have called a few upsets before. My best one was Rhaman over Lewis in the first fight.

Predicting upsets is not easy, but you can do better if you understand how styles can make fights. Sometimes one fighter has a weakness that plays right into the other fighters strengths. You can also pay close attention to outside the ring distractions, and make a best guess if the fighter is really motivated or not.

Yeah, I like these ideas. Another one is something like a change of trainer & handlers - if it's a change from a pairing that was good (e.g. Tyson/Rooney), this is often a sign of problems. Equally, changing from a mediocre to skilled trainer can sometimes predict a big improvement in a fighter, especially if he has great intangibles but is a bit rough around the edges (e.g. Pacman).

[Beeston Brawler]

I cannot believe Holyfield was 25/1 to beat Tyson.

Both were past their best at that time - Tyson more so..... but even their peak versions would encourage me to place a week's wage on Evander.

Most of the guys that stood up to Tyson..... and had skills, chin and a bit of power emerged victorious.