There are, to my mind, four key passing statistics: yards per attempt, completion percentage, interception percentage and touchdown percentage. There are various ways of combining these and other statistics into a QB "efficiency" rating, but these are mostly useless as they overvalue touchdowns and understate the significance of interceptions.
Often, touchdown passes are treated as a sort of "offset" interceptions when rating a QB. For example, a QB who throws 15 TDs and 9 interceptions is usually assumed to be superior to a QB who throws 7 TDs and 6 interceptions over the same number of passes. However, whether a drive is finished off by a TD pass is often irrelevant to the progress of a game, while interceptions are frequently backbreaking. Consider that ND's touchdown drive to tie the game against Navy was set up by a long pass with Powers-Neal finishing it off with a short run. As a practical matter, whether Holiday completed the 2-point conversion pass turned out to be much more important than whether he threw a short TD pass on second down. However, in a close game like the Navy contest even one interception probably would've doomed N.D. That's not to say that quarterbacks shouldn't get some credit for throwing touchdowns. But usually the "ratings" give them too much credit.
In my past efforts, I've explained the important of yard per play statistics and with passing this translates to yards per attempt. An average of 7 yards per attempt is above-average, about like a 4-yard-per-carry rushing attack.
So, let's look at Holiday by these critical measures, broken down three ways. We'll start with his 2001 campaign, his 2002 season as a whole and then his 2002 season after returning from the shoulder injury he suffered late against M.S.U. (that is, Holiday's last five games). Here's the breakdown:2001
Yards per attempt: 5.44
Completion %: 50.7
Touchdown %: 2.08
Interception %: 4.86
2002 (whole season)
Yards per attempt: 7.24
Completion %: 52.3
Touchdown %: 3.05
Interception % 1.02
Yards per attempt: 8.00
Completion %: 57.9
Touchdown %: 4.39
Interception %: 0.87
Instantly it becomes clear that there has been dramatic improvement in Holiday's passing prowess on every front. While, to be sure, Holiday deserves a huge amount of credit, consider the role that coaching plays here. N.D. last year averaged a pitiful 5.1 yards per attempt (literally the worst in Division I-A) and lost two starting wide receivers and a tight end to the NFL and nonetheless has improved to 7.1 yards per attempt in 2002. Stanford, despite returning most of its apparently key offensive personnel, has fallen from 8.3 yards per attempt last year to a pitiful 5.4 this year.
Now, one thing that is amazing about Holiday's statistics this year is his low interception percentage. Even for the whole year, it is only slightly above 1 % (2 interceptions in 197 attempts). Anything under 3% is excellent and anything under 4% is good. His touchdown percentage over 3 is reasonable (over 4 if you take his post-return numbers), but to the extent that other QBs exceed this rate it probably has more to do with play calling than anything else.
The most significant improvement, however, is in yards per attempt. Last year's 5.44 was a very weak figure, one of the worst among starting quarterbacks. This year's 7.24 is robust and the 8.00 post-return is excellent.
Let's compare Holiday with a few quarterbacks from apparently more passing-oriented attacks to see how Holiday stacks up.Wallace (Iowa St.)
Yards per attempt: 8.15
Completion %: 57.7
Touchdown %: 3.98
Interception %: 3.98
Yards per attempt: 6.71
Completion %: 57.4
Touchdown %: 3.94
Interception %: 4.21
Yards per attempt: 5.98
Completion %: 58.4
Touchdown %: 3.35
Interception %: 4.02
Yards per attempt: 7.94
Completion %: 62.7
Touchdown %: 6.36
Interception %: 2.31
Post-return, Holiday closely resembles Iowa St.'s Wallace, except that Wallace throws interceptions about four times as often. Even taking account of the whole year, Holiday is clearly superior to Florida's Grossman in yards per attempt and trails him only slightly in completion percentage and touchdown percentage and, as with Wallace, Holiday throws only about one-fourth the number of interceptions. Holiday is vastly superior to Stanford's Matter in all but completion percentage.
Of this group that I compare with Holiday, Palmer from U.S.C. is having the best year. Palmer has an excellent touchdown percentage, though this may be a function of a weak Trojan running game (U.S.C. averages under 3 yards per carry and has not faced the most fearsome run defenses in the country, to put it mildly). However, even Palmer throws interceptions more than twice as often as Holiday. Palmer's completion percentage over 62 is spectacular, however.
How does Holiday stack up with N.D. quarterbacks of the last 30 years or so? The best yards per attempt figures posted by N.D. quarterbacks in the recent past were 9.69 per attempt by Kevin McDougal in 1993, 9.12 by Rick Mirer in 1990, 9.05 by Mirer in 1991 and 8.71 in Jarious Jackson's heroic efforts to rescue the 1999 season. Holiday falls short of these excellent figures, but not hugely so, especially if his post-return figure of 8.00 is used. Moreover, the interception rate of all them is considerably higher than Holiday's. Mirer's 3.00 in 1990 and McDougal's 3.14 in 1993 were excellent figures, but roughly three times Holiday's rate and Jackson's 4.43 (not a bad figure) in 1999 was roughly 5 times Holiday's post-return rate.
In terms of comparing Holiday's effort of this year with QB performances of strong N.D. teams in the past, Holiday fits in well. In 1988, Rice completed 50.7 percent of his passes for 8.52 yards per attempt with a touchdown percentage of 5.84 and an interception percentage of 5.07. Rice's strong per-attempt average often catches people by surprise; but while Holtz's championship teams didn't throw often, they threw with devastating efficiency.
Montana's two best years were the Cotton Bowl championship in 1978 season and the national championship in 1977. In 1978 Montana averaged 7.73 per attempt, 54.2% completions, 3.84% touchdowns and 3.46% interceptions. In 1977 he averaged 8.48 per attempt, 52.4% completions, 5.82% touchdowns and 4.23% interceptions.
Clements in the Sugar Bowl championship year of 1974 averaged 7.20 per attempt, 56.7% completions, 3.72% touchdowns and 5.11% interceptions. In the national championship year of 1973 he averaged 7.81 per attempt, 58.1% completions, 7.08% touchdowns and 5.30% interceptions.
Thus, while Holiday trails these legends by a bit so far in some categories (assuming that his whole-season figures are used), none of them come close to his low interception percentage.
Actually, the only Irish QB to come close to Holiday's interception percentage was LoVecchio, who had a 1.86% interception rate in 2000, including the Fiesta Bowl. LoVecchio, however, trailed Holiday with a 6.94 yards per attempt figure. Morever, though LoVecchio clearly played well that year, most of his statistics were run up against weak opposition. Of the 7 regular season games he started that year, only one was against a team that finished the regular season with a winning record (6-5 West Virginia). Holiday, by contrast, this year has faced 6 teams in his 9 starts that seem certain to finish the season with winning records (Maryland, Michigan, Pittsburgh, Air Force, Florida St and Boston College), two others could finish at .500 (Purdue and M.S.U.).
Neither N.D.'s offense nor Holiday are finished projects, but huge improvements have been made.