One of the biggest stars of the first week of the 2004 NFL season was Denver running back Quentin Griffin, who ran for 156 yards on 23 carries against Kansas City. The opinion among NFL beat writers was nearly unanimous: Mike Shanahan had once again plugged a low-round draft pick into his rushing system and produced a league-leading running back, just as he did with Terrell Davis, Mike Anderson, and Olandis Gary.

Three weeks later, things look much different. Griffin is close to losing his job as Denver’s starting running back after combining an inability to convert on third-and-short with an acute case of butterfingers. His combined output in the last three games has not matched his yardage in the first week.

In Denver’s second game, his fumble of a handoff in field goal range cost the Broncos a win against Jacksonville. The next week, he gained only 7 yards on 12 carries against San Diego.

The astonishingly quick rise and fall of Quentin Griffin is a good lesson in the need to understand NFL stats in context. When an unheralded player has a big game, one needs look at the defense he played against.

A look at Griffin’s 2003 numbers would tend to suggest that he was ill prepared to replace Portis as the Broncos’ featured back. Griffin had one big game in 2003, a 136-yard effort against Indianapolis; last year’s Colts, like this year’s Chiefs were not exactly known as run stoppers.

Take out the game against Indianapolis, and Griffin’s 2003 numbers drop from a reasonable 3.8 yards per carry to a dismal 3.3 yards per carry. By contrast, Portis averaged 5.5 yards per carry in the same system, behind the same offensive line.

Early in the season, it is sometimes difficult to know which defenses are good and which are bad – nobody would have said after Week 1 that Aaron Brooks’s poor game could be blamed on the great Seahawks defense.

But Kansas City’s run defense was the main reason why the team collapsed in last year’s playoffs after a 13-3 season. After Griffin’s sudden accession to stardom in Week 1, nobody should have doubted the power of the Chiefs defense to turn pedestrian running backs into supermen. Further proof was offered by the performance of Carolina’s DeShaun Foster in Week 2.

Like Griffin, Foster had been one of the league’s least valuable running backs in 2003, mixing a few highlight reel plays with a huge number of runs that went nowhere. Against Kansas City, he ran for 174 yards, including one 71-yard run where he was barely touched.

NFL pundits should have looked at Griffin’s decline and considered the idea that yardage against Kansas City, like Canadian currency, was only worth 80 cents on the dollar. Instead, they were quick to coronate Foster as the league’s next star running back. But Foster looked much less formidable during his next game, when he gained only 51 yards on 19 carries against Atlanta, an average of just 2.7 yards per carry.

A closer look at both performances against Kansas City should have made reporters skeptical that either Griffin or Foster was ready to join the league’s upper echelon of running backs.

The trick to measuring a running back’s success lies not just in total yardage, but how often a running back puts his offense in a good position to continue a drive. Football Outsiders has designed a statistic called Running Back Success Rate, which measures rushing consistency by determining how often a back has a “successful run,” defined as 40% of needed yards on first down, 60% of needed yards on second down, or all needed yards on third or fourth down.

Despite his 156 yards in Week 1, Griffin was successful on only 39% of his runs, compared with a league average of 46%. Eight of Griffin’s 23 carries were stopped for a loss or no yardage. Foster was not much better in Week 2, successful on only 41% of his runs. Astonishingly, Foster lost yardage on eight carries – one of every four. Three more of his carries went for zero yards.

The standard criticism of the Kansas City run defense is that they are losing the battle at the line of scrimmage. But the feast-or-famine performance of running backs that face the Chiefs belies that argument. In Week 3 against Houston, the Chiefs stopped the opposing running backs for no gain or a loss on eight of 20 carries.

Even against the Ravens on Monday night, the Chiefs stopped Jamal Lewis, the best runner they have faced all season, for two losses out of 15 carries. For the season, the Chiefs have stopped opposing runners for a loss or no gain on one of every three carries. The average defense stops opposing runners for a loss or no gain only once every five carries, so perhaps the Chiefs’ defensive line is better than advertised.

When they do let the running back through, however, the linebackers and secondary can’t tackle or pursue in the slightest. That certainly seemed to be the issue with Foster’s 71-yard run, when he slipped through a small hole to the left of guard Tutan Reyes, who was blocking Chiefs’ middle linebacker Monty Beisel, and was barely touched until the Chiefs secondary finally caught up to him three yards from the end zone. This wasn’t 71 yards of cutting and avoiding tacklers – it was poor defense.

When Griffin and Foster have been able to get through the holes against other opponents, they have discovered linebackers and defensive backs that know how to pursue and tackle. Griffin has only one double-digit run this season against a team other than Kansas City, and Foster has none.

Mix poor runs with one big play, and it looks like you’ve had a good day. Mix poor runs with nothing over 10 yards, and the result is mediocrity. By the end of the season, that mediocrity will leave one good game against Kansas City long forgotten.

Mr. Schatz is the editor in chief of Football Outsiders.com.