In one category, each starter in a race was given points based on the following: six points for a grade I win at any time in his career; four points to a grade II winner or a horse that was second in a grade I; three points for a grade III win, second in a grade II, or third in a grade I; two points for a black-type win in a race carrying a purse of $50,000 and higher, second in a grade III, or third in a grade II; and one point for winning an open black-type race or second in a $50,000 and up race. A second category listed the total number of starters in a race that had won or went on to win a race of equal or higher grading. A third category divided the number of graded stakes winners by the size of the field, giving the committee the percentage. Finally, beginning in the 1980s, the committee had access to a computer-driven rating of each horse in a race developed by Bloodstock Research Information Services. Ten years ago, the computer ratings were replaced by annual classification ratings produced by the North American Rating Committee, a panel of racing secretaries who assign theoretical weights to all starters in black-type races. A more recent change was the decision by The Jockey Club of Canada to create its own grading system in 1999, thus causing a change in the name from the North American to the American Graded Stakes Committee. Morris and others charged that the system was flawed because of the statistical reliance on the graded stakes themselves. If one state or region had a high number of graded races for one division, the critics said additional races would likely be upgraded or graded because of the presence of so many graded stakes winners. Likewise, if a region had few graded races in a division, the chances were slim any races would be upgraded, they said. The other criticism centered on the committee's supposed reliance on the statistics, and not their own common sense or wisdom. "The numbers get scrutiny," said C. Steven Duncker, the committee's current chairman. "But it's not the most important factor when grading a race." The criticism "comes with the territory," Duncker said. "I liken it to the NCAA men's basketball selection committee. There's always someone saying this team or that team should have gotten into the tournament." One of the things Duncker hopes to accomplish as chairman is to make people better understand how the process works. "We need to do a much better job of transparency," he said. For starters, for the past two years, representatives from three industry publications have been invited to attend the committee's summer meetings, when procedures and guidelines are discussed. Duncker said he is considering opening up the grading sessions later this month to the same publications. Additionally, following this year's grading session, Duncker said a conference call will be scheduled for all interested members of the press. The committee has made several changes. First, serious consideration is being given to create something that more resembles Europe's Pattern Races for the 2-year-old division. Modifications also have been made to some of the statistical analysis that might prevent one region from getting top-heavy in a division. The committee also voted to require more stringent drug-testing for all graded races beginning in 2004. Duncker is the first to acknowledge the system isn't perfect. He recognizes grades can be perpetuated in a region, specifically mentioning some wintertime grade III events at Aqueduct that may not be deserving of their current status. Mark Johnston, a trainer based in England and a frequent buyer at American yearling sales, said he doesn't have complete confidence in the system. "I don't feel it's absolutely accurate," Johnston said. "I don't feel confident that I can take it at face value that a grade I at one track is the same as a grade I at another track." Yet Johnston and even the Racing Post's Tony Morris certainly would agree that the system is better than the chaos that existed before Jacques Wimpfheimer took a European idea and gave it an American spin.