At work our fantasy football draft is tomorrow and I've been busy this weekend on my pick sheets. I think there's something inherent in most who play fantasy football that we think we can outsmart the experts, find sleeper picks, find over-valued players, etc... For some reason we think we can go home one weekend and come up with a better list than several dozen professional websites - who've probably spent 100s or 1000s of hours putting player projections together... Heck, we might not have even seen one game last year by some of the players we're projecting, but we've read somewhere that a player's having a great pre-season...
But I guess that's the fun of it. We'd rather go down in flames with our own mis-guided player list than just blindly using a carefully researched cheat sheet from a professional website.
Well, today I've been looking at Team Defenses and have come to the conclusion that they're very difficult to project - at least not with the certainty to warrant a high draft pick.
For example:
In 2005 the top scoring defense in my league was Carolina with 220 points - about 60 points better than average. So it should make sense that they would've been good in 2006? Nope! Carolina had a below average defense score in 2006 - ranking 19th in the league. (Defenses in fantasy football include things like sacks, interception, fumble recoveries, return game TD, defense scored TD, safeties, blocked kicks/punts, points allowed, etc)
In 2006 the top scoring defense was Baltimore scoring a whopping 275 points - a full 109 points better than the average defense. So should I step up and draft such a powerful defense early? I don't know, because they only scored 172 points in 2005.
Here's another thing. The 3rd best defense in 2006 was San Diego - scoring 208 points. Was 2005 any indicator that they'd be that good? Not really. In 2005 the San Diego defense ranked 23rd in fantasy defense only scoring 148 points.
Why should defensive scoring be so volatile and seemingly unpredictable? The Baltimore defense would seemingly be drafted very early - probably a first round draft pick if they could be relied upon to score 270 points again this year, but in our league the top defense generally doesn't get drafted until the 3rd round, and in most leagues the defense doesn't go until much later - 5th or 6th round. It does seem they're highly unpredictable.
Some stats: The top 10 fantasy defenses in 2005 scored an average of 195 points, an average of 35 points above average. In 2006 those same 10 defenses averaged 167.4 points - less than 2 points better than the average for the league. In fact, 5 of the top 10 2005 performers were below average in 2006. In fantasy defense, past performance seems to often have little to do with future performance.
It makes me think fantasy scoring for defenses is based largely upon random or at least unpredictable results and that the level of certainty of future defensive scoring is very low.
There is one defense that has been consistently strong over the past 2 years: Chicago Bears. They performed well above average with 216 points in 2005 and 248 points in 2006. That seems more the exception than the rule though.
My plan? Unless the best defenses drop farther than they normally do in my league I'll probably try to take two defenses late and play the best match-up in a given week.