What's Your Process? (1 Viewer)

[Ack88]

One of the great, long term values of this board to me has been getting the perspectives of other players. More specifically, having been an active poster here for 20+ years (Thanks, FBG), I've noticed much evolution in the way managers build their rosters, be it for re-draft or dynasty. As an example, Value Based Drafting (VBD) was an early strategy debated and explained here. Additionally, I think Joe Bryant (and I'm not certain of this) was the creator of VBD. VBD entailed creating projections for the draftable players in your league; if you have a 12 team league and 15 roster spots per team- there are 180 total players that you create projections for. You would then weigh the total points scored by the top player and compare to a baseline player (last starter) to develop a comparison. For example, if Patrick Mahomes scored 400 points in your scoring format (top scoring QB) and your 12th ranked QB was Kirk Cousins and he scored 250 points, Mahomes would have an X value of +150. You could then compare across positions by calculating the difference between the top scorer and baseline player (last starter). If Justin Jefferson (top WR), scored 200 points and the 36th WR (12 teams/ 3 starters) scored 85 points. Let's say that's Adam Theilen. Jefferson then has an X value of +85. By this metric, Mahomes would be considered "more valuable" in the cross comparision.

With all the other metrics ADOT, Yards per attempt, yards per snap, etc...There are literally dozens of quantitative analytics out there, my question to the board is- What is your process?

I still use VBD as a significant indicator and create my draft shell board based upon VBD. Within the VBD model, I also use tiers if the player buckets have super close projections.

I create my own VBD projections independently, without looking at any other sources first, because I don't want to be biased by outside thinking.

Then, I look at other sources and read about the narratives that inform the projections and sometimes, based upon the quality of the narrative argument, change my projections.

Thanks, in advance, for your thoughts.

 

[Gally]

For those of you that do projections for each player how do you translate that to fantasy points? I can see it being done in various ways but since this is a weekly game somehow translating the yearly projection into weekly points seems very meaningful. Do you do it as simply plugging the total projection into the scoring system? So 1 pt for 10 yds and you project 1000 yds so that is 100 pts. Then divide that by 15 weeks to come up with 6.67 pts per game? Do you some how factor in the boom/bust factor to come up with a weekly value assessment?

I don't have a good answer for any of this and have gone back and forth trying to come up with some formula but nothing I have been happy with. Just looking for other thoughts from people that may have figured something out for this.

 

[acarey50]

For those of you that do projections for each player how do you translate that to fantasy points? I can see it being done in various ways but since this is a weekly game somehow translating the yearly projection into weekly points seems very meaningful. Do you do it as simply plugging the total projection into the scoring system? So 1 pt for 10 yds and you project 1000 yds so that is 100 pts. Then divide that by 15 weeks to come up with 6.67 pts per game? Do you some how factor in the boom/bust factor to come up with a weekly value assessment?

I don't have a good answer for any of this and have gone back and forth trying to come up with some formula but nothing I have been happy with. Just looking for other thoughts from people that may have figured something out for this.

The quantitative part is the "easy" part - do projections (or take from a source like FBG), plug into your scoring system and get your fantasy projections. Pretty straight forward. Worrying about further breaking that down into PPG is majoring in the minors unless you can semi-accurately predict who will get injured,

The qualitative part is the "hard" part. I'll note guys I think have higher upside potential (boom players), those that have bust potential (ie WR that lost a stud QB, RB with a new guy there to take reps), and may adjust my quantitative rankings a bit from there, but I'm not likely to jump a guy up a huge number of rounds because of gut feel.

Lastly, depending on the league/site, I'll look at ADP data to see where my rankings differ greatly from ADP. These are the players that I will look at closer - what did I miss in my analysis, what are others seeing. If I have a guy ranked well ahead of ADP and I still like him, I know he is someone that I may be able to let slip a round but would still take ahead of ADP to ensure I got him. If a guy is ranked well below ADP, then he is likely not a player I will end up with as he will go before I'd be comfortable taking him.