Single entry and 3-Max contests have a large number of entries that are simply a player’s cash game roster. The problem with that ideology is typically for cash games, we should be looking to maximize floor, which is counter-productive to taking down a large field GPP. But if we view each roster we create as a portfolio, we can intrinsically manage risk/reward, or floor/ceiling, to create an upside roster without sacrificing floor. What this creates for us is a high cash rate in these contests without sacrificing the ability to actually take it down. Leveraging Game Theory doctrine, we’re searching for ways to differentiate ourselves without sacrificing high-probability plays. So what does this look like while we’re actually filling out a roster?
Say we’ve sunk a large portion of our salary into the running back position (again, something we should only be doing if we’ve previously uncovered value plays with a high floor/ceiling combination at other positions), which on any given week could approach 30-35% of overall salary for two roster spots (two of nine roster spots, or 22%). We’re assuming we’ve found value at other positions, but the value we uncover should not simply be cheap plays with high ceilings.
One of the easiest places to find value (without sacrificing floor) is to utilize a mini-stack with a quarterback and one of his pass-catchers (90% solution here is a mid-range wide receiver). The reasoning being :: less actually has to go right for us to see profits, because if a quarterback has a solid game, it stands to reason that one of his pass-catchers will also. If we can find a QB in the sub $6000 range, and pair him with a WR also in the sub $6000 range, we’re bringing our total salary spent to a maximum of 49% for 44% of the salary and roster, respectively (35% salary on RB, 24%, or $12,000 salary, on QB/WR pairing, for a total of 4/9 roster spots, or 44%), and we know we have pay-down positions remaining to fill (TE and DST). The following screenshot for week 1 NFL 2020 is not meant to be binding; it is simply meant to illustrate this thought process for SE/3-Max roster construction. This is an extreme example, assuming we have already found value WRs through our bottom-up roster building approach! In this example, we’re placing a range of outcomes at each position, which would look something like 18-30 expected points from Minshew, 25-35 from CMC, 24-30 from Jacobs, 18-30 for Chark, and 15-20 for Herndon. DST is a completely different animal, in that I am not typically looking for value in a point per dollar mindset at the position, I am looking for the highest raw point total on the week. I personally am very strong at identifying top defenses on a slate, and will not hesitate to be 100% on one defense in SE/3-Max (everybody has what works for them, this is simply a strength of mine). If we had not identified value WRs for this example, CMC would be the first to go, leading us to a more balanced roster construction.