Dr. Adam Tennant, assistant professor of engineering, grew up watching boxing. Over time he began questioning why one guy got a chance to fight for a championship when another boxer seemed a better contender. This curiosity stayed with him and when a graduate project in a complexity course came up concerning ranking, he decided that rather than ranking infrastructure or bridges, he’d rank boxers.
Tennant was already of the opinion that boxing organizations had their share of corruption. In his paper, “Complexity analysis in the sport of boxing,” he cites the corruption of major boxing organizations. For example, The International Boxing Federation founder was convicted of racketeering for taking bribes to move fighters up in the rankings, a boxing promoter paid officials to boost rankings for his guys and the World Boxing Organization was known to manipulate ranking rules to benefit their business interests. The complexity experiment gave Tennant a way to alleviate possible corruption using a calculation bank-based scientific approach to build a dataset of the top boxers over a 10-year span in the welterweight divisiion (140 – 147 pounds).
He gathered a dataset from the website, BoxRec, which holds the records of current and retired boxers from the 1800s through today. His set consisted of more than 11,000 boxers to analyze. He ran the data through the same algorithm, PageRank, initially developed and used by Google Search to rank websites in their search engine results. “It is a probability distribution that uses the weighted network to optimize rankings and past success of an individual and produces a cumulative advantage that continues to grow as one’s career continues,” said Tennant. “It’s a model validated by studying the careers of both scientists and athletes and how early success generally predicted future success.”
Tennant only took into account the number of wins and losses. The results showed a marked difference between how the major boxing organizations ranked the boxers compared to how they ranked using BoxRec, total wins and PageRank.
Tennant plans on doing more research on rankings in boxing with datasets spanning the entire history of boxing, and looking at pound for pound, who the top boxer in history is. He also offered that if some of the ranking organizations wanted a ranking system less prone to corruption, perhaps they should give his a try.
Photo Credit: USI Photography and Multimedia
Dr. Adam Tennant