Somehow, rather than accurately depicting reality, mock drafts seem to become reality. Fans are so engrossed in what Mel Kiper has to say that they get upset if their team does not take the player that Kiper speculated the team wanted to take. Think about that. It's true and it's very backwards. Tyson Alualu is one of my favorite players from this draft, but people in Jacksonville are upset because Kiper didn't have him in the first round. Not every team would throw the pundits' wisdom - and thus the fans disposition - out like that for fear of the repercussions.

At PredictionMachine.com, we could attempt to speculate before the draft (we do specialize in predictions). How would that be done? How would that be evaluated? Ricardo Mathews is our most likely pick for Detroit in 6th round at a whopping 0.4% of the time. Detroit doesn't select Mathews. Did we just get that wrong?

The process of predicting the outcome of the draft is not nearly as valuable or as significant as what we are doing - taking the actual results of the draft and evaluating which players will make the most impact. For now, we are focusing on this upcoming season only, but future seasons could definitely be included.

This is my favorite article of the year every year because it shows off just about everything we can do with our technology, from generating inputs for new players, to predicting playing time, health and coaching styles, to playing through entire seasons 50,000 times. It also signifies the beginning of football season. From now through January, 2011, football will be a priority. I spent much of this week preparing the analysis below and will continue to work on analysis and build tools for the NFL and college football seasons until they begin - at which time we will have about four times as much weekly content.

To come up with statistical inputs for rookies, we run a very complex set of algorithms that factor college stats, previous utilization and strength of competition, combine measurables, role and expected utilization of the player's NFL team, previous performance of a similar player in that role for this coaching staff and previous performance of similar rookies at that position in general. This allows us to predict both the player's projected ratio stats such as yards-per-carry, percentage of tackles made while on the field and passing completion percentage, as well as his forecasted usage for the upcoming NFL season. Then we can compare all rookies based on who we think will make the biggest positive impact for his NFL team in his first year. It's a system that I have applied to the previous three drafts and with great success. My projections have highlighted the likely success of Percy Harvin, Jerod Mayo, Louis Delmas, Sean Smith, Ryan Clady and even Super Bowl hero Tracy Porter, while suggesting that Glenn Dorsey, Vernon Gholston, Jerome Simpson, Everette Brown and even Matt Ryan were/are overrated.

Unlike previous seasons, this year we are ranking every drafted player. And the rankings will be based on output from our initial tests of the upcoming season played 50,000 times.

We also took this analysis and applied it to teams to rank the impact of the rookie classes of all 32 teams. See our NFL Rookie Class Rankings for more. 

2010 NFL Rookie Rankings