VCU news

By Tom Gresham

The VCU men’s basketball team’s run to the Final Four represents a highly unlikely statistical event, but Laura McLay, Ph.D., an assistant professor in the Department of Statistical Sciences and Operations Research at VCU, cautions against taking the Rams’ odds-busting success as an argument against using statistical analysis to predict sports outcomes.

Approaching one’s tournament bracket with an eye on statistics still will serve you best in your office pools, McLay said, but during some years, such as this one, a series of unlikely events will emerge and defeat the statistics-based approach. Statistical analysis will produce the most accurate predictions on average for what will occur over many tournaments, but results in a single tournament will always divert in some way from what the numbers can predict.

“I encourage people to use statistical information when they’re filling out their brackets, but they should also allow some randomness in the process,” said McLay, a self-described “sports nerd and lover of numbers.” McLay recently wrote about VCU’s current streak of success on her blog, Punk Rock Operations Research.

VCU’s five-game winning streak against favored teams in the NCAA Tournament marks only the third time that a No. 11 seed has ever reached the tournament’s Final Four. Odds against what the Rams have done proliferate in media reports, including analyst Ken Pomeroy’s 1-in-3,333 chance that VCU would make the Final Four and Nate Silver’s 820-to-1 measurement on his closely followed New York Times blog. Silver, in fact, speculated that VCU reaching the Final Four this year might be the most unlikely event to ever occur in the NCAA Tournament and clearly the most surprising – statistically – in the past 25 years.

McLay compared VCU’s striking tournament streak to a coin being flipped repeatedly and landing on heads over and over again. At some point, she said, one wonders if instead of seeing a rare event, the coin is biased toward landing on heads.

“Maybe this isn’t a fair coin we’re flipping,” McLay said. “Maybe VCU is better than the numbers going into the tournament indicate.”

McLay said the numbers available before the tournament all were stacked against a VCU run. Indicators such as RPI, Ken Pomeroy’s statistics, the Sagarin Ratings, win-loss record, team offensive and defensive efficiency and margin-of-victory did not rate VCU’s squad favorably alongside past Final Four entries.

However, McLay said, “there are more possible outcomes in the tournament than there are stars in the galaxy.”

This vast number of possibilities means that something such as all four regional No. 1 seeds reaching the Final Four is still not a safe bet at the beginning of the tournament. In fact, that event has happened just once in the tournament’s wild history.

“Even with likely events, they are still unlikely to occur,” McLay said.

McLay, in her blog post on the Rams’ tournament run, noted that at the outset of the tournament one estimate put VCU’s chance of winning a national championship at 1-in-203,187. However, she said now that VCU has reached the Final Four, odds have improved dramatically for the team, ranging in observations she’s seen from about a 1-in-10 chance to a 1-in-4 chance. She believes the best measurement lies somewhere between those figures – a figure, she said, that “means VCU has an excellent chance of being the national champion.”