The RotoViz suite of Roster Construction Explorer tools are awesome. Because we have both tournament data and a long history of data, we are able to rigorously test out roster construction ideas for historical success.
When it comes to the Underdog BBM III and Puppies however, there is one important caveat to keep in mind; our pre-2021 data is compiled exclusively from best ball leagues that use full PPR scoring. Underdog is half-PPR and (mostly) involves tournaments. We need to be careful about translating lessons from other game types.
Looking through historical Fanball data tells us that even the most optimal structures are not precise prescriptions ordained by the best ball gods.
FANBALL 2015-2017, RB1 BEFORE RD 3, RB2 AFTER RD 5
Taking five total running backs in an Anchor RB (RB1 before Round 3, RB2 after Round 5) works well in full PPR leagues, and so does six. Moving off that range clearly sacrifices win equity, but it is a range. Sharp tournament players have leaned towards the lower end of that range to stockpile more wide receivers. The logic is that tournaments reward weekly ceilings and wide receiver scoring is more volatile on a weekly basis. With wide receiver ADPs higher than previous years and half-PPR scoring, is this actually the best approach? I’ll be going through Zero RB, Anchor RB, and Hyper Fragile to answer this question within each construction.