In Athens, the study predicted overall medal counts with 94 percent accuracy and gold medals with 86 percent accuracy. The overall medal accuracy for the Sydney and Salt Lake City Olympics were 95 percent and 94 percent, respectively.

      Johnson’s paper, “A Tale of Two Seasons: Participation and Medal Counts at the Summer and Winter Olympics,â€� was first produced in 1999 with Ayfer Ali, while Johnson was teaching at Wellesley College and Ali was a student. It was published in Social Science Quarterly in December 2004.

      Some of the surprises this year, from Johnson’s perspective:

• More nations — 26 — won medals at these Winter Games than at any previous Winter Olympics. At the Salt Lake City games, 24 nations won medals.
• Canada won 50 percent more medals than forecast, and nearly displaced the U.S. for second-most medals. Canada shattered its previous record of 17 medals in 2002, with 24 this year.
• South Korea won 11 medals, equal to China’s haul. Johnson didn’t forecast South Korea due to data issues, but he still didn’t expect the country to break into the top 14 nations.
• China continues to impress, and for the fourth consecutive Olympics beat the forecast by a large margin. China’s investment in its athletics infrastructure continues to cause this aberration; the country is clearly not following historical trends. Johnson expects this to be even more pronounced in 2008’s Beijing Games, when he expects China will top the podium most often.

   Italy was a very generous host in these Games, giving away more gold and other precious metals than Johnson’s model predicted. The home field advantage didn’t seem to help; Italy actually won three more medals in Salt Lake City than it did while hosting in Torino. Finland and Norway both had disappointing results as well.