Top NFL Statisticians?

Sounds like logistic regression models.

Excel is pretty good at converting any sort of tabular data on a website into columns and rows.  You don't even need to have a .csv file.

I majored in Mathematics/Actuarial Science in college.  Regression models were used to determine what stats were important in modeling past results and what was most predictive of future performances.  After that, I make projections based on current performance level of teams adjusted for SOS and test those against criteria I've measured over the past season and a half to significantly predict ATS winners.  I don't use regressions to model future outcomes.  I think they help point the right direction in what stats to use, but a predictive model solely based on regressions would not be consistently successful in my opinion.  Unless those models could be tested against certain criteria, the respective weights of each variable would hold too much variance to be consistent - just my opinion.  What I mean is that although a pure statistical model could be used to better predict winners compared to simple moneylines, beating the spread is another story entirely.  I think that's what needs to be tested against projections and compared to past results.  I think far too often regression models are great at modeling the past but fail to predict the future accurately.  I agree about combining historical data with current data, but don't think a linear or evenly weighted model will work.  What I mean is that a team with a 5 yard per pass attempt advantage would not be linearly comparable to a team with a 2.5 YPA advantage with different strengths/weaknesses.  You can project how efficiently a team will pass or run just fine, but that doesn't mean the flow of the game will progress the same.  The team playing against the 5 YPA advantage may drop 8 men in coverage every play and force the run if that's what they're good at controlling.  In this way, a team playing against the 2.5 YPA advantage may actually perform worse than the team playing the 5 YPA advantage if their run defense is worse.  I don't think you can predict performance in each category and add them together.  I think every area needs to be integrated or measured collectively, which is where statistical models fail.  An example - I play unders in games where my system projects high yards per carry and a high number of sacks, regardless of points scored or points allowed.  It seems counterintuitive, but a team rushing the ball well will continue to run and thus run down the clock quicker.  I don't care if they rush for 5 yards every play if it takes them 16 plays and almost 10 minutes to go down the field and score.  A team projected at 2 yards per carry and a low QB rating may seem like a bad team to contribute to an over, but if that means they run the ball 10 times and throw 50, the clock is stopping on every incompletion or they're gaining 8 or so yards with the completion - good for an over.  Now those may seem obvious, but my point is that statistics alone (in a regression model for example) may be able to project a team's stats for a game fairly accurately, but if it doesn't take every category into consideration as a whole, it won't be successful ATS.  I hope this all makes some sense - if not, feel free to ask questions and I'll clarify what I can.

As far as injuries are concerned, the only offensive injuries I worry about are at quarterback - and possibly offensive linemen.  I completely ignore all other positions and feel completely justified in doing so.  On the defensive side of the ball, however, I'm not sure what positions are most significant and at what relative level.  Certainly the MLB or a lock-down corner, but again, I don't know a way to consistently adjust based on those injuries - especially if you go deeper and address the specific ability of each injured player in question.

As far as emotions and subjective factors, I don't think they exist (at least not significantly).  In college maybe, but in the NFL where players are playing for a salary, I think motivation levels aren't issues that need to be accounted for consistently.  In certain cases, they probably are present, but for the most part I think every player is completely motivated to play at a predictable level every time.  It's their job. 

Not so wise..... Emotion is the biggest factor that matters in the NFL

Have all three of these as well, among others.

Sides