If I should describe my overall goal, it is to feel comfortable about deciding my streaming picks, without the stress of having second thoughts, and wondering if another ranking source was "guessing".  I know that fantasy football so random that I'm guaranteed to pick busts sometimes.  (I am aware that my model accuracy is going to suck-- but it just needs to suck a bit less than the next guy's).  But at least I want to have the feeling of comfort in knowing the underlying method is sound-- so that based on available info, I could not have decided better. 

I think this feeling of comfort can come from:

1. knowing underlying assumptions,

2. knowing that the right predictive numbers are included,

   • therefore knowing that misleading factors are left out,

3. knowing the model is oriented for future prediction specifically (as opposed to back-fitting of historical data)

4. knowing that all numbers in the model are calibrated by the right amounts.

   • If "yards-per-pass-attempt" is 3x more predictive than "sacks-per-play" or whatever, then the past data had better agree.

Towards this goal of not overthinking/doubting/stressing, I can say I feel comfortable for myself. And hopefully I can pass on some of that feeling with you over a season. 

Fantasy projections are no guarantees... and I know some 5% of you will feel like you end up picking the wrong guy all 16 weeks, and I'm sorry for that. But on the whole I really hope I'll help improve the experience overall.

What is Subvertadown NOT?

Let’s be clear that the content and projections here are not about guarantees.  There is no “high certainty” in fantasy or any other form of betting.  Subvertadown is not about guarantees, but it is about understanding how much of a “guarantee” is possible or not possible.

Subvertadown content is also not currently about individual players (RB1/WR2/TE1)—only total team score for those positions.  And Subvertadown is also not currently about helping your draft for these skill positions.