View Single Post
Old 08-25-2013   #19
Texanmike02
Site Contributor
 
Texanmike02's Avatar
 
Join Date: Apr 2007
Age: 36
Posts: 7,794
Rep Power: 58272 Texanmike02 is a quality contributor and well respectedTexanmike02 is a quality contributor and well respectedTexanmike02 is a quality contributor and well respectedTexanmike02 is a quality contributor and well respectedTexanmike02 is a quality contributor and well respectedTexanmike02 is a quality contributor and well respectedTexanmike02 is a quality contributor and well respectedTexanmike02 is a quality contributor and well respectedTexanmike02 is a quality contributor and well respectedTexanmike02 is a quality contributor and well respectedTexanmike02 is a quality contributor and well respected
Default Re: FantasyFirstRoundStrategy

I have a spreadsheet that I built to draft for me. I can send it to you if you want (as long as we don't play in a league together). But I will give you a basic gist of how I draft. I started doing it several years ago (and then stopped playing until last year I started again) and this thing has been money. The most important thing to do is rank players correctly AND have accurate projections. Obviously this is the most difficult thing but I think the biggest problem people have on draft day is adjusting to the draft.

Everyone has a philosophy on drafting. Draft RBs early, draft QB's later... etc. They are all hit and miss. Draft based on opportunity cost and nothing else.

To get ready to do this you need to establish values for your players. Usually I will gather 5-6 cheat sheets, weigh and average them and then make adjustments based on where I think errors are.

Once you do that you need to set up a need coefficient. I am in a league that plays 1 QB 2 RBs 2WRs 1K 1DEF and then a WR/RB and a WR/RB/TE.

Assuming that every team needs 1 QB and 1 backup QB I will use 1 (for a starter) and .5 for a backup so the QB coefficient is 1.5. It is more difficult for WR's because every team will start 2WRs but can start up to 4. To deal with that I have all WRs ranked with a coefficient of 3 (2 starters and 2 backups) and I have the same thing for RBs. WR/RB/TEs also have a second column in which they are all dumped into one pool. In that pool WR/RB have a coefficient of 3 and TEs have a coefficient of .75 (Since TE isn't a mandatory position I don't even think every team will draft one of them). K/DEF have a coefficient of 1.

The coefficient changes as people draft. If they draft a starter I will deduct 1 point. If they draft a backup that .5 points comes off of the board. Every team starts off with a total coefficient of 13.25 (this year).

You wind up with a board that looks like this:
PLAYER 1
QB 1.5
WR 3
RB 3
WR/RB 3
TE .75
K 1
D 1

Take the number in the second column and divide it into the total of the coefficients to get the likelihood that that person drafts that position. That is done for each person left to draft between you and your next pick and then you total the numbers for each position to get how many players should be taken. I zero out the coefficients for K/D in the first 5 rounds because I have never seen anyone take a player that early and I round anything more than .3 up to the next whole number.

So if I am drafting now and I have a pick in 5 more picks I will wind up with what I call "likely picks" between now and my 5 picks for each position.

I also "worst case" for each pick which is just the number of picks between mine and the next one.

I have all players ranked in total number of points I think they will score by position and I have a link to my first worksheet that shows the most points possible at that position (the total number of points I think the BPA will score) and then subtracts the total of each successive player down to the last player I bothered to predict.

Then I have MY needed picks.

So say my "likely picks" for QB is 2 and the worst case is 5 and for RB it is 4 and 5.

If the differentials look like this:

QB RB
10 15
15 10
1 15
35 1
20 0
8 14

For QB I would take 25 (total of the first two) and 81 (total of the 5) and average them and 41 and 41 for the RB. Then multiply that by the percentage of needed players I have (1 for QB 3 for RB including starting the RB in one of the remaining Spots) I do the same thing for 2 or 3 picks later and display them as well for each position left. I usually pick based on the lowest number first but I keep an eye on the 2nd or 3rd pick later to see if anything big is developing and if there is I might override the auto pick.

This works pretty well until I have drafted all of my starters. For any sleeper I have (players I think that will be undervalued) I will usually go in and either drop their difference to zero or subtract one from the worst case number if they fall inside of the worst case picks. This year I am working on a second sheet that will rank players in likelihood they get drafted and work that into the equation.

I'm not sure if that helps or not. I've played with a lot of formulas and used them in competing leagues a couple of times and this seems to yield the best results. Having the numbers laid out like this also allows me to spot positions of need for other players which lets me target players they might need for trade later. If you want the sheet I'll throw it your way just PM me but overall remember you're not drafting players and you're really not even looking to score points, you are trying not to lose points from what is possible in each pick.

Mike
__________________

Sig by False Start - Thanks man

Govern my country, NOT MY LIFE!
Texanmike02 is online now   Reply With Quote