How Often Does the Higher-Rated Boxer Win? Upsets by Rating Gap
How often the higher-rated boxer wins, how the rating gap shapes upsets and draws, and why short fights are the weakest ground for favourites.
Every card rests on one quiet assumption: the boxer with the better rating should win. I wanted to know how often that holds and how it bends as the gap between two fighters widens. So I took 170,874 professional bouts from public records, 1990 to May 2015, where both boxers carried a pre-fight computer rating, and sorted them by how far apart those ratings sat.
The headline answer is that the higher-rated boxer won 71.2%, the lower-rated boxer 24.2%, and 4.6% were drawn. That is a solid edge but not a crushing one, and it shifts sharply with the size of the gap and the scheduled distance. When I build my fight night predictions page, this is the baseline I start from before styles come into it. One caveat up front: these records carry no betting prices, so "favourite" here means the higher-rated boxer, not the market favourite, and the rating is a strong but imperfect stand-in for a price.
1Upsets shrink steadily as the rating gap widens
The stacked chart splits every bout by the ratio of the higher rating to the lower one. When the two numbers sit within 1.25x of each other, it is close to a coin toss: the higher-rated boxer won 50.3% and the lower-rated boxer 43.8% across 14,600 bouts. At a 2-3x gap the rated boxer wins 66.6%, and at 10x or more he wins 90.0%.
What strikes me is the floor. Even with a 10x gap, the lower-rated boxer still won 7.6% of 28,045 bouts. That is the bad cut, the flat night, the overlooked journeyman with a real right hand.
Ratio = higher pre-fight rating divided by the lower one. Bouts in brackets.
View the data: results by rating gap
| Higher-rated won | Draw | Lower-rated won | |
|---|---|---|---|
| Under 1.25x (14,600) | 50.3% | 5.8% | 43.8% |
| 1.25-1.5x (14,380) | 54.8% | 5.7% | 39.4% |
| 1.5-2x (22,444) | 60.6% | 5.6% | 33.8% |
| 2-3x (31,055) | 66.6% | 5.4% | 28.0% |
| 3-5x (31,673) | 74.2% | 4.6% | 21.2% |
| 5-10x (28,677) | 81.3% | 4.0% | 14.7% |
| 10x or more (28,045) | 90.0% | 2.4% | 7.6% |
| All rated bouts (170,874) | 71.2% | 4.6% | 24.2% |
Draws move the same way. They fall from 5.8% in closely rated bouts to 2.4% at 10x or more. Evenly matched fighters go the distance and split the judges far more often than mismatches do.
2Fair three-way prices drawn straight from the rating gap
The table turns those frequencies into decimal odds with no margin, simply one divided by the historical rate. For the higher-rated boxer the fair price runs from 1.99 under a 1.25x gap, to 1.50 at 2-3x, down to 1.11 at 10x or more. On the other side, the lower-rated boxer is fair at 4.72 with a 3-5x gap and 13.16 once the gap reaches 10x.
This is an illustrative model, not a price feed. It tells me roughly where a number should start before I weigh southpaw angles, reach, activity and who owns the jab.
| Rating ratio | Bouts | Higher-rated | Draw | Lower-rated |
|---|---|---|---|---|
| Under 1.25x | 14,600 | 1.99 | 17.24 | 2.28 |
| 1.25-1.5x | 14,380 | 1.82 | 17.54 | 2.54 |
| 1.5-2x | 22,444 | 1.65 | 17.86 | 2.96 |
| 2-3x | 31,055 | 1.50 | 18.52 | 3.57 |
| 3-5x | 31,673 | 1.35 | 21.74 | 4.72 |
| 5-10x | 28,677 | 1.23 | 25.00 | 6.80 |
| 10x or more | 28,045 | 1.11 | 41.67 | 13.16 |
The draw column is where I spend time. Fair at 17.24 in closely rated bouts and 41.67 in deep mismatches, it gives me a benchmark: when two boxers are rated near level, the draw deserves a proper look rather than a reflexive dismissal.
3Four-rounders are the weakest ground for rated favourites
Scheduled distance matters more than most people price. In 4-round bouts the higher-rated boxer won only 61.2% of 13,822 bouts, against 69.5% over six, 73.2% over eight, 73.3% over ten and 72.3% over twelve.
My reading is stylistic. Over four rounds there is no time to solve a problem: a fast starter, an awkward opponent or a single knockdown can settle it before the better boxer adjusts. Give him eight or ten and the jab, the conditioning and the corner's second plan start to tell.
View the data: higher-rated win rate by scheduled rounds
| All rated bouts, by scheduled distance (bouts in brackets) | |
|---|---|
| 4 rounds (13,822) | 61.2% |
| 6 rounds (40,322) | 69.5% |
| 8 rounds (31,348) | 73.2% |
| 10 rounds (35,207) | 73.3% |
| 12 rounds (20,482) | 72.3% |
The edge flattens after eight. Twelve rounds do not hand the rated boxer more than ten do, because the championship distance brings its own risks, from fading legs to close cards.
4Upset rate drifted slowly downward from 1990 onwards
The line tracks the share of rated bouts won by the lower-rated boxer each year. It was highest at 28.0% in 1990 and lowest at 21.6% in 2012, finishing at 22.9% in 2014.
The slide is gentle rather than dramatic, and the wobble between years is real. I read it as the rating itself getting sharper as more results feed into it, though matchmaking habits may play a part too.
View the data: upset rate by year
| Lower-rated won | |
|---|---|
| 1990 | 28.0% |
| 1991 | 26.7% |
| 1992 | 25.2% |
| 1993 | 26.3% |
| 1994 | 25.6% |
| 1995 | 25.4% |
| 1996 | 24.8% |
| 1997 | 24.9% |
| 1998 | 24.1% |
| 1999 | 23.4% |
| 2000 | 25.0% |
| 2001 | 25.4% |
| 2002 | 25.4% |
| 2003 | 24.5% |
| 2004 | 24.4% |
| 2005 | 23.8% |
| 2006 | 23.0% |
| 2007 | 24.4% |
| 2008 | 24.0% |
| 2009 | 23.3% |
| 2010 | 23.1% |
| 2011 | 23.1% |
| 2012 | 21.6% |
| 2013 | 22.3% |
| 2014 | 22.9% |
For a bettor, the lesson is that the early years slightly overstate the upset rate. When I lean on this data, I give the later seasons more weight.
Betting notes
- Respect close ratings. Under a 1.25x gap the lower-rated boxer won 43.8%, so a short price on the rated man there is rarely justified by the numbers alone. That is where the styles have to do the work.
- Shade four-round favourites. The higher-rated boxer won just 61.2% over four rounds against 73.3% over ten. I want a wider rating gap before backing him at a short price over the short distance.
- Benchmark the draw. Use 17.24 as the fair draw price in tight pairings and 41.67 in mismatches. Anything far shorter than that range is mostly margin, not insight.
Quick answers
How often does the higher-rated boxer win?
Across 170,874 rated bouts from 1990 to May 2015, the higher-rated boxer won 71.2%. The lower-rated boxer won 24.2% and 4.6% were drawn.
How often do big underdogs win in boxing?
Even when the higher rating was 10x or more the lower one, the lower-rated boxer won 7.6% of 28,045 bouts. That frequency implies a fair price of 13.16 before any margin.
How common are draws in professional boxing?
Draws made up 4.6% of all rated bouts. They ran at 5.8% when the boxers were closely rated and fell to 2.4% when the rating gap was 10x or more.
I'm Chloรฉ Marchand, based in Paris, and I write the boxing betting tips at bettips.vip โ reading styles and rounds rather than the names on the marquee.
All of Chloรฉ Marchand's work โ
