Dr. Giganto

Well, my first bit of inspection into this gave me mixed results. I took 21 players who either started this season or would have started were it not for injury. Also, I eliminated any player that did not have at least 16 career starts in either the first half or second half (thus, no Tony Romo, no Derek Anderson, etc.)

I then separated the 21 players into two groups...the warm weather group and the cold weather group.

The warm weather group:

Eli Manning
Peyton Manning
Philip Rivers
J.P. Losman
Carson Palmer
Jake Delhomme
Brett Favre
Steve McNair
Daunte Culpepper
Jeff Garcia

The cold weather group:

Byron Leftwich
Marc Bulger
Chad Pennington
Donovan McNabb
Jon Kitna
Drew Brees
Trent Green
Matt Hasselbeck
Kurt Warner
Tom Brady
Ben Roethlisberger

The warm weather group saw an average drop of .85% in their completion percentage. Thus, there was not that much of a difference.

The cold weather group saw an average drop of .755%...a little bit lower than the warm weather group, but too close to really be considered significant.

Next, I looked at touchdowns per game, multiplied times 16. You could also consider it "Touchdowns per 16 games played".

The warm weather group saw an INCREASE of .37 TDs per 16 games. Although it's surprising that it went up, it's such a small amount, it'sessentially zero. This is not very significant

The cold weather group saw an increase of .265 TDs per 16 games...once again, not much of a difference.

Then, I looked at "Interceptions per 16 games". This was the only place where i saw a significant difference. The warm weather group saw an increase of .856 INTs, or just shy of one interception per season.

By comparison, the cold weather group only threw .01 more INTs per 16 games in the second half of the season. Essentially, the time of year had no difference whatsoever on the number of INTs thrown by this group.

There are a few problems with this work I did. First of all, 21 QBs is not that big a sample size...I probably need to look at more players, further back in time, to see if there is any significant difference.

Secondly, there are some people who simply arent appropriate for this study. For instance, Jake Delhomme went to a warm weather school, but he plays in a warm weather division, so at least 75% of his second half games are in warm weather locations. Every year, he has 4 home games in Carolina, 1 or 2 road games in either Atlanta, Tampa, or New Orleans, plus random games in domes or at other southern teams's home stadiums, such as the Cowboys, Jaguars, etc. Only like 1 or 2 of his games per season are in legitimately cold weather.

Then, of course, there's the dome QBs. A guy like Daunte Culpepper played most of his career in frigid Minnesota, but his games were indoors. So, he might skew the numbers, since his games arent truly affected by the weather.

There are two possible solutions I can think of. Yahoo gives you a players' splits based on game-time temperature, and it separates it into 4 categories: Frigid, Cold, Mild, and Hot. I could look at performance in Frigid/Cold games versus performance in Mild/Hot games.

Or, i could look at individual teams that play outdoors in cold weather, and how their QBs performed. And if a player played like 10 years in Chicago, and 2 in Jacksonville, I would throw out the data from the Jacksonville years.

If I did this, these are the franchises whose QBs I would consider:

Both NY teams
Philly
Washington
Chicago
Green Bay
Seattle (only since moving out of the Kingdome)
New England
Buffalo
Pittsburgh
Baltimore
Cincinnati
Cleveland
Denver

Lastly, there's the potential "Brett Favre effect"...a warm weather QB might eventually adjust to the cold, and over a long enough career the differences will shrink to almost nothing. If I observe this, I could possibly use only the first 5 years of a player's career.