Rotosource: Your source for Fantasy Sports information, reviews, guides and more

Home Guide Site Reviews Projections

Football Reviews Baseball Reviews Basketball Reviews Projection Competitions



2006/2007 Fantasy Basketball Player Projection Competition

2006/2007 Fantasy Basketball Player Projection Competition Results

What an ugly season 2006/2007 turned out to be - the "Season of the Injury". The list of major fantasy players that missed over 25 games due to injury is simply stunning: Dwayne Wade, Ray Allen, Michael, Redd, Yao Ming, Joe Johnson, Paul Pierce, Lamar Odom, Brandon Roy, David West, Jason Richardson are just some of the names, and there are a host of major players that missed 20, 15 or 10 games as well.

It is notoriously difficult to predict player injuries, and of the ten projection sets that we looked at, none of them predicted most of these injuries. In fact, some of the players that were expected to miss large chunks of the season, Pau Gasol and Amare Stoudamire, played early and often, respectively. Few websites got everything right, but some did better than others, and we did our best to rank their usefulness to you - the fantasy basketball manager. Note - we only looked at websites with player projections. If a website didn't include player projections in their draft prep materials, we didn't include them in our competition. We explain our focus on player projections in our fantasy sports guide.

One thing to take special note of is the process we followed to rank the projection sets we received. Because most fantasy basketball leagues use a rotisserie scoring format, that is how we evaluated the projections. The problem arises because player valuation under roto scoring is very difficult to accomplish. We evaluated each player according to the method discussed in our fantasy sports guide, and we evaluated each player according to actual statistics for the 2006/2007 NBA season. We used standard Yahoo settings to determine waiver-wire player value, and to determine position depth. We then took the difference between player value according to a projection set, and according to actual results, and a projection set's score was equal to the sum of the squares of the differences for all players. There are many different ways to evaluate projections, including basic correlation of variables between reality and a projection, but we like our method for the simple reason that it most closely approximates a projection set's usefulness during an actual fantasy basketball draft.

Originally, we planned to take games played into account when setting player values, both for the real player values, and player values as determined by a set of projections. This quickly hit a snag this season, as it became clear that those projection sets that didn't try to project games played were really rewarded this year. Everyone who tried to project games played this season was so far off that they would have been better off just predicting that everybody would play 75 games apiece and be done with it. Rather than reward the projections who didn't try to project games played (because we're certain that they would not have predicted equal games played if they had made the attempt), we decided to evaluate all players on a per-game basis this season.

Fantasy basketball sites are a little different from other fantasy sports websites, in that many fantasy basketball sites offer their content for free. Because fantasy basketball has a smaller following than fantasy baseball or fantasy football, many sites just don't charge for their projections. Did the paid sites produce better projections?

2006/2007 Fantasy Basketball Player Projection Competition Results

Direct any questions regarding the submission, scoring, or rules of this competition to ppcbasketball0607@rotosource.com.



Home Guide Site Reviews Projections

Football Reviews Baseball Reviews Basketball Reviews Projection Competitions