Curious how all the pros here statistically evaluate talent

[The Ball Basher]

OK we all use SoS and basic stats, but does anyone do anything unique or they think is unique?

 

On my end it is pretty simple. I take total points scored and divide that by total opportunities (i.e. targets, carries, passes, ect.). This gets me a point conversion ratio. I then calculate what the league average is at the position and run a simple deviation from the average. This allows me to compare productivity of a player to another when evealuating the draft and trades. A player with a lower conversion ratio and similar opportunites will perform worse and a player with a lower conversion ration, but significantly more opportunities can perform better.

 

Anyone have any tidbits to share as I am curious to see what others do?

[flemingd]

I watch football and prefer players that don't suck.

[CaptainHook]

I watch football and prefer players that don't suck.

this

[The Ball Basher]

I watch football and prefer players that don't suck.

 

Well that is a novel idea...LOL

[MikesVikes]

It looks like your theory is based on the fact that you have coaching authority over the players. The actual coaches may have a different opinion of his players. We don't have control over the number of targets that a player gets.

 

But I will add that performing any type of number crunching will give you a working knowledge that you didn't have before.

Edited November 2, 2010 by MikesVikes

[Brentastic]

I watch football and prefer players that don't suck.

+1

 

I know this response may come across as 'smart-assed' but it's really the only way. Watch all the games (even pre-season) and you know who the real players are and those that are inflated by easy opponents or those around them.

[Phantom Knight]

It looks like your theory is based on the fact that you have coaching authority over the players. The actual coaches may have a different opinion of his players. We don't have control over the number of targets that a player gets.

 

But I will add that performing any type of number crunching will give you a working knowledge that you didn't have before.

I actually like the OP's formula. Having a bit of an idea of who will do better with more touches is a nice piece of info to have. Obviously we cannot control who WILL get more touches, but we can identify who is likely to get more touches - due to injury, ineffectiveness of player in front, etc. Often times there may be 2 or 3 players that seem to be trending toward more touches... this formula can make the decision of which to pick up a bit easier.

 

I know it is a lot less scientific, but I like to look at average ffp points scored in my leage over the last 3 weeks. If we are in week 9, what a player did in week 2 or 3 is of no interest to me (exception being if the player was injured along the way and missed games). Likewise, looking at highest scores from only the last week does not give enough of a data set and leaves many chasing points. But a 3-week window is enough of a data sample in my opinion to spot a player truly trending upward. Many times over the years it has brought a player to the forefront for me who otherwise would not have been my top choice.

 

As an aside, Yahoo used to provide this report (last 3 weeks average score), I used it to evaluate player in all my leagues - Yahoo, MFL et al. Sadly, Yahoo has taken this handy tool away.

[Company of Heroes]

One thing to consider for trades (which is easier to do as the season wears on) is look for specific weeks you need matchup help at different positions. Start looking week by week and unload players (even if they are solid) that are likely to ride your bench for players with better matchups in the specific weeks you need matchup help.

 

Even if the acquired player's ability and end of season production is inferior you can come out ahead if you got more points from the player in the weeks you were hurting. Even better if said player is coming out of a tough schedule stretch and therefore has some upside potential.

[keggerz]

Stats? Who needs 'em.

[Big John]

Stats? Who needs 'em.

[wolfer]

Stats? Who needs 'em.

 

I need your IDP breakdowns! Me, me!!

 

If you think TDs decide the difference b/t studs and everyone else, red zone targets/goal line opportunities are worth looking at.

 

Oh and sometimes the 'players that don't suck' are stuck in the dumbest situations. See: Charles, Jamaal.

[G.K.Trey]

I need your IDP breakdowns! Me, me!!

 

If you think TDs decide the difference b/t studs and everyone else, red zone targets/goal line opportunities are worth looking at.

 

Oh and sometimes the 'players that don't suck' are stuck in the dumbest situations. See: Charles, Jamaal.

 

How is Charles in a "dumb" situation? Because he is on your team and you think he should touch the ball 40 times a game? Considering that KC is

5-2 , they seem to be doing ok without your input on the offense.

[ajfalcone]

Many facets to drafting. 1) your leagues scoring rules 2) players true talent 3 ) opportunity for the player

 

Having Torain, Blount hit fantasy revelance was no secret if you knew Portis and cadillac wouldnt sustain. Even Lynch is an example. The guy is talent and now is given the opportunity. Of course luck is needed for where hes gonna end up but any bucket head knew his potential.

 

winning your championship isnt in the first couple rounds. Its rounds 8-9 and beyond grabbing true talent/ opportunities that win championships. Also not being active on wiver wire will for sure eliminate you from contention.

 

Also statistically evaluating talent is wrong because it doesnt work. players who have not been give the opportunity to play have no stats. So you cant base their potential. Unfortunately players who get drafted erly on get paid millions which means the team is commited to them regardless if someone else being more talented is behind them. See Jamall Charles situation

Edited November 3, 2010 by ajfalcone

[keggerz]

Many facets to drafting. 1) your leagues scoring rules 2) players true talent 3 ) opportunity for the player

 

Having Torain, Blount hit fantasy revelance was no secret if you knew Portis and cadillac wouldnt sustain. Even Lynch is an example. The guy is talent and now is given the opportunity. Of course luck is needed for where hes gonna end up but any bucket head knew his potential.

 

winning your championship isnt in the first couple rounds. Its rounds 8-9 and beyond grabbing true talent/ opportunities that win championships. Also not being active on wiver wire will for sure eliminate you from contention.

 

Also statistically evaluating talent is wrong because it doesnt work. players who have not been give the opportunity to play have no stats. So you cant base their potential. Unfortunately players who get drafted erly on get paid millions which means the team is commited to them regardless if someone else being more talented is behind them. See Jamall Charles situation

maybe that can be said on offense but that isn't even 100% true..you can look at the production based on their opportunities but finding snaps played isnt the easiest thing to do...on the IDP side of the ball I have metrics I am working on that sort of act like a batting average for a baseball player....Player A hits .310 in 600 at bats and Player B hits .335 in 100 at bats....big difference but still shows promise...IDP Side Player A logs 125 tackles on the year and his "batting average" is .190 having played in 900 snaps and Player B logs just 25 tackles on the year but his "batting average" is .250 on 100 snaps...still shows promise...of course that doesn't mean the produciton will translate but it is still something to consider

Edited November 3, 2010 by keggerz

[historymike]

I have a few number crunchers I trust (like NumberFire) that consistently produce solid results, but I also also watch trending players, and I am much more interested in upward trends over the previous three weeks than I am with seasonal averages or deviations from league averages. This would be the statistical version of the hot hand theory.

 

Oh, and like some of the other posters I prefer to choose players that do not suck. The suck quotient can be defined many ways, but here is my formula:

 

 

-b ± SQRT(b^2 - 4ac)

 

 

x = --------------------------------------------

 

 

2a

 

 

 

Where X is the suck quotient, b is the number of times a player has been injured in the past two years, a is the total number of drug/alocohol incidents up before the NFL disciplinary committee, and c is the number of stupid comments that the player has made that might cause him to get benched.

Edited November 4, 2010 by historymike

[Big John]

I have a few number crunchers I trust (like NumberFire) that consistently produce solid results, but I also also watch trending players, and I am much more interested in upward trends over the previous three weeks than I am with seasonal averages or deviations from league averages. This would be the statistical version of the hot hand theory.

 

Oh, and like some of the other posters I prefer to choose players that do not suck. The suck quotient can be defined many ways, but here is my formula:

 

 

-b ± SQRT(b^2 - 4ac)

 

 

x = --------------------------------------------

 

 

2a

 

 

 

Where X is the suck quotient, b is the number of times a player has been injured in the past two years, a is the total number of drug/alocohol incidents up before the NFL disciplinary committee, and c is the number of stupid comments that the player has made that might cause him to get benched.

aka the quadratic formula.

[the chef]

Wow ,,,

My Brain Hurts!!!!!

[historymike]

aka the quadratic formula.

 

+2 for knowing the math.