Player Projections (1 Viewer)

[mr blond]

I’ve been lazy since discovering FBG some years ago. I always use FBG projections and VBD sheet, coupled with my own football insight for my draft. I’m not complaining about FBG projections, I’m always one of the best teams in my leagues and have won a few championships. I want to start putting together my own projections. With the availability of FF information being what it is, I feel putting together my own projections will give me a more in depth view and an added advantage over others in my leagues.

My problem is that I’m unsure where to start. If any of you experts out there can point me in the right direction I’d appreciate it. My thought is that I should start at the overall team level and work my way down to individual players.

I’m sure there are many in my position, some sort of a how-to guide for projections would be a great resource. I'd also be interested in hearing about any tools utilized out there for projections.

 

[code]

This question has been asked several times over the few months...try a search...

http://forums.footballguys.com/forum/index...e=projection%2A

 

[mr blond]

This question has been asked several times over the few months...try a search...

http://forums.footballguys.com/forum/index...e=projection%2A

Thanks for pointing that out Code. From reading the posts in some of the other threads it sounds like a team stats approach is popular.How about projection tools? I'd be interested in seeing tools or applications used by others.

 

[nerangers]

I used to love doing a lot of my own projections/rankings, but that was before kids! LOL Four kids later, I now let someone else do the first pass. FBG projections are a good start, then I tweak them based on my own research about the teams, and how I think things will play out. Finally I put them into tiering buckets so I cam see where the drop off will be for each position.

 

[Spartans Rule]

I go with a team approach, and I like to break things down by percentages.

First I like to look at each team, estimate how many plays they'll run and what percentage will be runs and passes. Then I estimate what percentage of the carries each RB will get, what percentage of the passes the QB(s) will complete, and what percentage of the catches each WR will get. Then after that, look at each player individually to determine yards per run/pass/catch.

This approach works well for me in keeping the stats realistic and grounded somewhat.

 

[moleculo]

I go with a team approach, and I like to break things down by percentages.

First I like to look at each team, estimate how many plays they'll run and what percentage will be runs and passes. Then I estimate what percentage of the carries each RB will get, what percentage of the passes the QB(s) will complete, and what percentage of the catches each WR will get. Then after that, look at each player individually to determine yards per run/pass/catch.

This approach works well for me in keeping the stats realistic and grounded somewhat.

I do the same basic thing. One twist I use is I want to have historical trends and averages to help guide percentages and YPC figures.

 

[nygiants56]

I take my scoring system and find out who my top 20 at each position are from last year (* if you use CBS Sportsline as a league manager, you can get all this info quickly).then, I like to look into trends..some players perform better early on in the season ( Ron Brown, Drew Bledsoe), some are great cold weather players or guys that seem to get stronger as the season wears on ( Rudi Johnson , for example)..I also look into schedules quite a bit, although they aren't accurate forecasters of late.

the biggest trend I look at is what a player does during the final 8 weeks of the season..if they a performing at a high level, that's likely to carry over into the folowing season.Conversely, if they have a lousy final 8 weeks of the season, its usually a sure bet that that trend will continue the following year and you should bump that player down in your projections..

 

[PennState-JD]

Last year I got put on to a system that worked really well for me. I took the players' stats from the year before, arranged them by position, and then calculated the distribution of total points. Generally, the distribution of points will be remarkably consistent from year to year.

Once I have the point totals for RB1, RB2, RB3, etc., I assign players to each of those totals. That part requires some projections to see if my initial placement is reasonable, a lot of cross referencing, and some luck.

Once you have this, you can draft based on VBD principles as it is very easy to see the projected drop in points for each position you are considering before your next pick will come up.

 

[Rounders]

I believe that FBG will be unveiling a new product this year: Projections Dominator

http://forums.footballguys.com/forum/index...howtopic=242013

 

[code]

Last year I got put on to a system that worked really well for me. I took the players' stats from the year before, arranged them by position, and then calculated the distribution of total points. Generally, the distribution of points will be remarkably consistent from year to year.

Once I have the point totals for RB1, RB2, RB3, etc., I assign players to each of those totals. That part requires some projections to see if my initial placement is reasonable, a lot of cross referencing, and some luck.

Once you have this, you can draft based on VBD principles as it is very easy to see the projected drop in points for each position you are considering before your next pick will come up.

This basically sounds like AVT...I used AVT to model our leagues scoring system, but for individual players simply slotting the players into AVT projections seems less accurate than computing projections by team...

 

[mr blond]

Last year I got put on to a system that worked really well for me.  I took the players' stats from the year before, arranged them by position, and then calculated the distribution of total points.  Generally, the distribution of points will be remarkably consistent from year to year.

Once I have the point totals for RB1, RB2, RB3, etc., I assign players to each of those totals.  That part requires some projections to see if my initial placement is reasonable, a lot of cross referencing, and some luck. 

Once you have this, you can draft based on VBD principles as it is very easy to see the projected drop in points for each position you are considering before your next pick will come up.

This basically sounds like AVT...I used AVT to model our leagues scoring system, but for individual players simply slotting the players into AVT projections seems less accurate than computing projections by team...

I've read up on AVT in the past and it has its merits. If anything it serves as a good way to double check your projections. I'm really looking to take a ground-up approach to my projections. My thought is that it will help me be a more educated owner by helping to organize the info I gather and concepts I learn. In the past I could lurk these boards, read as many posts that interested me, print out the VBD sheet and basically be a top 3 team in my league. With the availability of FF info, I'm looking to stay a step ahead of the competition. I think creating my own projections will help me better understand more than just the players drafted in the first half of my draft.

Thanks for all the replys. I missed the Projection Dominator thread, it sounds like another great tool.

I'm still interested in seeing other tools or applications utilized in projections if anyone is willing to share.