NFL MVP - Formula
 

I've made a formula to rate QBs, RBs, WRs and TEs and come up with a score per game. I haven't come up with a way of rating defensive players yet.

I'm sure that it's like the BCS computer rankings - it would cause more disagreement than anything else. It's not meant to be perfect but at least it separates a guy with a 900 from someone with a 700. Actually I can't see anyone outside of my top 5 with a chance.

Updated for end of regular season - actual number of votes included.

Name	Pos	Team	Total	Actual Votes
Shaun Alexander	SEA	RB	1320	19
Larry Johnson	KAN	RB	1176
Tiki Barber	NYG	RB	1156	6
LaDainian Tomlinson	SDG	RB	1112
Carson Palmer	CIN	QB	1109	2
Steve Smith	CAR	WR	1104
Peyton Manning	IND	QB	1016	13
Tom Brady	NWE	QB	1003	10
Edgerrin James	IND	RB	996
Larry Fitzgerald	ARI	WR	960
Marvin Harrison	IND	WR	936
Chad Johnson	CIN	WR	932
Matt Hasselbeck	SEA	QB	928
Drew Brees	SDG	QB	916
Trent Green	KAN	QB	915
Santana Moss	WAS	WR	912
Joey Galloway	TAM	WR	912

I would say your top 3 are probably the clear-cut top 3 contenders in real life... so at least it passes the eyeball test.

What is your formula? I wanted to do something like this for the NAFL and determine who the front runners are for the MVP in the league.

I think you would be relatively safe to say his top 5 are arguably the top 5 in real life also.. so that also passes the eyeball test...

I see problems lower on the list...for instance, I know LaMont Jordan is the man on offense for the Raiders (leads them in rushing and catches), but he's on a losing team...as is Collins. I wouldn't rank either of them before Hasselbeck or Bledsoe. Also, Steve Smith should be higher. As was said before, I don't have a problem with your top 5.

Based on the list, I'm guessing that team record may not be a large part of his formula... for instance, I wouldn't expect Fitzgerald to even make a list this long because his team is so bad...

I was just curious because I need a starting point to develop the formula. I tried and it didn't work out, so I gave up on the formula.

Would be pretty cool to have this for MP Leagues...

That is what I am trying to do. Rather than have a voting system where not all GM's participate, have a formula driven MVP calculation which shows you the front runners as the season progresses.

Like I said, the formulas are not perfect but they are a starting point. Actually I did something similar for an online baseball league. I came up with some rankings to give owners an idea of the frontrunners for the MVP, Cy Young, Rookie and Reliever of the Year. It was never definitive but got people talking. So yes, I think these formulas could be used for online football leagues as well. The formulas can also be used to determine a Game Score to help to decide a Player of the Game award.

It all started as a way of simplifying the QB rating system

Formulas:

QB = 2*Completions - Incompletions + Yards/10 + 10*TD - 10*Int
Note that the [2*Completions - Incompletions] part can be done using [3*Completions - Attempts]
I also round the Yards down so that 153/10 = 15 and 157/10 = 15 as well. Just helps when working things out in my head when there's no spreadsheet around.

RB = Yards/10 + 10*TD - 10*Fumbles Lost
The result is doubled to get it comparable to QB.
Note that rushing and receiving stats are used for Yards and TDs.

WR/TE = Same as RB ( except no rushing stats used ) except result is multiplied by 4

Actually I discovered a couple of mistakes in my original formulas. I originally used straight up fumbles instead of lost fumbles and I had a bracket in the wrong spot on the receivers.

Corrected values:

Thanks 21c. I am going to try and toy around with this and get something going for the NAFL to use through excel outputs. I agree, the one thing missing is team ranks, in which a player on a better team gets a higher adjustment.

EDIT: Should there also be something to adjust for the position itself. The chances of a TE getting league MVP over a QB/RB/WR are slim and none.

Feel free to adjust it yourselves. My formulas have no scientific validity to them except they come out as reasonable numbers that can be easily compared.

Sorry, I wasn't knocking your formula at all, just opening up to discussion on how to add to it.

On the other hand, for a fictional league you have the opportunity to correct this gross injustice ;)

Don't worry, I didn't see that you were knocking it. I just thought that anyone could use my initial formulas to build on for themselves.

Does anyone have an idea as to how the game figures out who the MVP's are? There has to be some sort of code written to determine it, or does it just take the league leaders?

I'll give anyone a buffalo nickel if they can figure the in game MVP formula out.

It has to be a formula. I think running a formula like this on the past season of an FOF league would go a long way to figuring it out. However i think player of games and player of weeks also might be a factor in FOF (and that makes a lot of sense).

Updated as of Week 15 pre-MNF.

Updated for end of regular season.