Heels, Sox & Steelers

Is starting your rookie QB a bad thing?
In the last week I've talked a few times about Eli Manning and Philip Rivers (here and here), the two QBs taken early in the first round, and the only two rookies QBs who will probably begin the season as starters. With all the pressure that Manning has already faced from his shaky minicamp debut, I got to wondering if starting a rookie had ill-effects on his career long-term -- particuarly when compared to QBs who didn't start until later in their career. Most people think that any experience is good experience and as a result, rookie QBs gain more in exposure than they lose in bruised egos.

I took a look at all QBs that have been in the league since 1990 through 2001 and I grouped them into those that started as rookies and those that didn't (specifically, I looked at rookies that started at least 6 games their first season). To get a very general measure of QB success I calculated TDs - INTs for each QB and I also calculated each QBs winning percentage (modified to reflect wins scaled by number of games started -- I'll call it 'expected wins').

I modeled a QB's expected wins per season being dependent on:

* whether a QB started as a rookie (the theory being that more playing time means long-term dividends in the way of experience, poise...etc)

* the difference between a QB's TDs and INTs (call it TD_INT)

* the number of years a QB plays in the league

The results show that for every additional TD a QB throws while keeping INTs constant, increases his expected wins by 0.24. For example, if Tommy Maddox throws four more TDs than INTs he'd be expected to win one more game than he would have won if he threw the same number of TDs as INTs.

For each additional year in the league, QBs are expected to win 0.12 more games. So a QB in his sixth season is on average expected to win almost a half a game more (0.48) than a QB in his second season -- over the course of the year.

The biggest effect on winning percentage however is if a QB started as a rookie. Simply by starting as a rookie, a QB's expected wins per season increases by 1 full game.

Of course there are a lot of variables that aren't taken into consideration here. First, many of the QBs that start as rookies were first round draft picks and therefore more likely to succeed as professionals than players taken later in the draft. Second, it follows that if a team uses a high pick to take a QB then that team must have had a pretty poor record the previous season. In all likelihood, winning percentage will increase just because the team won so few games the previous season -- and that's how the draft should work, creating some sense of competitive balance. Finally, the players surrounding a QB and they type of offensive philosophy are very important variables in determining a team's success and that is something I didn't consider here.

While none of this is earth-shattering, it is at least interesting to see how starting a rookie QB affects the number of wins a team can expect. Given this, on average, there doesn't seem to be any deleterious long-term effects to starting a rookie QB, and in fact it looks to be a good investment. Knowing that, it will be interesting to see how Manning and Rivers (especially Manning) cope in their first seasons. And one more thing, unless Roethlisberger plays like a veteran All-Pro during training camp, don't expect to see his mug on the field after the preseason. Of course if he does manage to make it on the field in 2004, it can only help the Steelers long term, right?

Send me an Email