Even though some might refuse to admit it, we have all allowed the projections of our respective fantasy sites to influence our decisions on whom to sit and whom to start in any given week.
To be honest, I have even relied on such projections outright to decide certain spots of my starting lineup. I would be surprised if a majority of fantasy footballers haven’t done the same themselves, for whatever reason, be it that they are too busy that weekend to properly research their players and matchups, or that they have two players competing for a single starting spot that is a toss-up as to who will produce better.
Given the realities of life and the fact that fantasy football is a hobby for most people, a hobby that could very easily be a full-time job if one was lucky enough, what weight should we really give to these fantasy projections? I might be alone here, but I certainly do not know what formula these sites use to come up with their numbers.
It would have to be some combination of past performance and the defenses that each team is going up against that week, but exactly what factors in is a mystery. Do they include weather conditions? Is it a conference rivalry? What are the past performances versus the same opponent? These are all valid considerations, yet I suspect they are not factors in the projections.
To determine whether we are doing ourselves a disservice by relying on fantasy projections, I have put together a table of five players from each offensive position (they are actually the top five projected players from FFToday.com’s Week Six fantasy projections) and compared their actual stats with their projected stats. I took the Week Six projections of Sportsline.com, Yahoo! Fantasy Sports and FFToday.com to compare.
Why did I choose these three specifically? Because they are the ones I access most frequently.
The points are calculated using standard scoring rules: six points for a touchdown, -two for an interception, and one point for every 25 yards passing and 10 yards rushing or receiving.
I considered a projection to be accurate, or a “hit”, if the total points of the projection was +/- two points of the player’s actual performance. These successful “hits” are signified by being marked in red.
To the left are the projections of Sportsline for five starting quarterbacks, and as you can see, all of whom were supposed to produce decent fantasy numbers. Unfortunately for Sportsline, none of their projections proved to be accurate. If you started either of Drew Brees or Tony Romo, you likely won’t be as upset because both players exceeded Sportsline’s projection: 30.8 points for Brees and 30.84 points for Romo. Unfortunately for those that started the other three, they were all under their projections.
Now what about the others? Well, both Yahoo! and FFToday.com fared better, producing two and one hits, respectively. However, what we are looking for here is accuracy and a hit rate of less than 50 percent is not all that encouraging. A quarterback that completes less than 50 percent of their passes is looked down upon, so too should we upon poor projections.
Here are the actual stats of the above quarterbacks for Week Six:
Running backs are typically the bread and butter of the fantasy realm. If they aren’t part of a platoon system, then they get frequent touches and often have two spots for them on your roster.
With all of the stats that are calculated, such as yards-per-carry and yards-after-contact, you would figure that projections would tend to be quite accurate. Unfortunately that isn’t the case. Although better than the quarterback projections, overall we still have a hit rate of less than 50 percent.
Sportsline does redeem itself somewhat with a 60 percent hit rate for running backs.
FFToday.com does win the award for the most accurate projection of the week for predicting that Matt Forte would score 17.5 fantasy points by rushing for 115 yards and one touchdown. He ended with 110 yards and a touchdown, resulting in a difference of only 0.5 points (this was good news for me, as I had started Forte in two of my leagues):
WIDE RECEIVERS / TIGHT ENDS
For the sake of brevity I will not provide details of the remaining two categories of players here, simply because the result is much of the same, if not worse. Of the 10 players that remain, Sportsline only gets three hits, Yahoo! gets four, and FFToday.com gets one. That is nine hits out of a possible 30 for a hit-percentage of 30. Please click HERE for a copy of the table in PDF format to peruse on your own time.
What can we take from this analysis? I am not too certain. True enough that it was only done on five (almost) randomly selected players from each position for Week Six, so the sample size is rather small. But it does go to show you that success at fantasy football takes more than simply looking at your league’s projections and clicking on “top lineup” and dusting your hands clean for the week.
It takes consideration of matchups, team chemistry, weather, and numerous other factors that may or may not influence a player’s performance any given week. These sites, relied upon by many, had hit-percentages of 43 (Sportsline and Yahoo!) and 33 (FFToday.com). You could probably do better with only 15 minutes of time spent reviewing your team. I suppose that is the lesson that could be learned.
If anything, you can take from this analysis that a law student will do almost anything to procrastinate reading another archaic case on bills of exchange.