A question for those who study the NFL draft

[VarlosZ]

[ QUOTE ]
What did it say about David Klingler and Andre Ware?

[/ QUOTE ]

It doesn't go back that far, but it shouldn't be too hard to get a rough idea yourself. Go back and look at their college stats. If they started a lot of games and had a high Comp%, then the system liked them.

[Jeff W]

[ QUOTE ]
It's accuracy seems pretty impressive. It would have pegged Ryan Leaf, Akili Smith, Jim Druckenmiller, Joey Harrington, Kyle Boller, Marques Tuiasosopo, Quincy Carter, Charlie Batch, and J.P. Losman as total busts.

[/ QUOTE ]

Since the system is based on past data, we should expect it to have pegged past QBs who have already been incorporated. You can't build a function from a dataset and then use that same dataset to test its accuracy.

[VarlosZ]

[ QUOTE ]
[ QUOTE ]
It's accuracy seems pretty impressive. It would have pegged Ryan Leaf, Akili Smith, Jim Druckenmiller, Joey Harrington, Kyle Boller, Marques Tuiasosopo, Quincy Carter, Charlie Batch, and J.P. Losman as total busts.

[/ QUOTE ]

Since the system is based on past data, we should expect it to have pegged past QBs who have already been incorporated. You can't build a function from a dataset and then use that same dataset to test its accuracy.

[/ QUOTE ]

Yeah, of course results after the creation of the function are preferable, but in lieu of those, it's not like past results are meaningless. The system and its hypothetical results provide pretty good evidence that completion percentage and games started in college correlate quite strongly with professional success.

[VarlosZ]

Here's the FO article that first introduced the QB projection system. The chapter in the book is much more in depth.

[TheNoodleMan]

[ QUOTE ]
[ QUOTE ]
What did it say about David Klingler and Andre Ware?

[/ QUOTE ]

It doesn't go back that far, but it shouldn't be too hard to get a rough idea yourself. Go back and look at their college stats. If they started a lot of games and had a high Comp%, then the system liked them.

[/ QUOTE ]

Ware was a 61% passer and I think he started for 3 years. Klingler 57% and 2 years as a starter.
Just curious about them since you mentioned throwing out QBs taken past the 2nd round to get rid of the Texas Tech guys and the Houston guys are similar except they were drafted higher.

[VarlosZ]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
What did it say about David Klingler and Andre Ware?

[/ QUOTE ]

It doesn't go back that far, but it shouldn't be too hard to get a rough idea yourself. Go back and look at their college stats. If they started a lot of games and had a high Comp%, then the system liked them.

[/ QUOTE ]

Ware was a 61% passer and I think he started for 3 years. Klingler 57% and 2 years as a starter.

[/ QUOTE ]

I can't be sure at all, but comparing those numbers to similar players, it looks like the system wouldn't like Klingler (J.P. Losman started 27 games and completed 57.8% of his passes, and he's projected as a very marginal starter, or a pretty good backup).

It would probably like Ware a lot more. Kellen Clemens started 32 games with a Comp% of 61%, and the system projects him as a slightly above league-average QB. However, scrambling QBs frequently throw very few passes, and Lewin has stated that QBs who throw an unusually small number of passes per game tend to underperform their projection. I don't know if this is the case with Ware.

[SoloAJ]

[ QUOTE ]

I strongly recommend their latest book, Pro Football Prospectus 2006, which contains a lot more information on the projection system, plus just about anything else you might want to know about the NFL.

[/ QUOTE ]

I would label myself slightly above a "casual fan." I don't get into statistical stuff like this too much....

Question: Would I still get enjoyment out of this or is it mostly a book for the nerdery?

[Jeff W]

[ QUOTE ]
Question: Would I still get enjoyment out of this or is it mostly a book for the nerdery?

[/ QUOTE ]

I dunno, might as well buy it and see, it's only $12 or something.

[VarlosZ]

$19 at the book store, but yeah, a great value. Lots of stats you won't find anywhere else make it a great reference (e.g, did you know that the Bears' offense lined up with 3 WRs 36% of the time, or that their defense rushed six or more defenders 13% of the time, or that Brian Urlacher's average tackle occured 4.1 yards past the line of scrimmage?).

There are all sorts of research essays, such as the frequency and efficacy of various types of blitzes at various times, whether NFL teams suffer from away games thousands of miles from home, whether big RBs actualy wear down a defense, whether age or workload is more responsible for a RB's eventual breakdown, the 4th Down tendencies of the various head coaches, etc., etc.

If you're not interested in stats or research, there's still plenty of straight up football analysis, with chapters for each team analyzing recent performance and off-season moves, paragraphs and stat projections on all offensive skill position players, and lots more. Makes for excellent bathroom reading.

[Jeff W]

I bet the projection system will love Chris Leak. 4-year starter, >60 career completion%.