Oct 17, 2021
2 mins read
Access can be requested here: tinyurl.com/ocean-NFL21
NEW LOOK CHEAT SHEETS:
These sheets are my favorite due to their versatility. They work for quick builds, as a starting point for your builds, as an end point to fill that last roster spot, and of course visualize the slate and how each position shakes out.
For starters, I added two new columns: VEGAS which shows the favorites and dogs, along with home or road. SPRD which shows the spread for each game. I debated on also including the team totals, but declined because they're already reflected in the projections and providing them again could cause people to give more weight to them than is proper (they're available on other tabs if needed).
The other change I made was to the color coding across the board. Of prime importance here is the changes made to the coding for projections and in particular, CREAM. The problem with numbers is that people tend to view them in absolutes and play them as such, even when they know they're not. This is also the problem with optimizers (among other things), though the optimizer by it's purpose must assume they are. Anyways, I've tightened up the range to upper ranges of the standard distribution curve ( top 68.2%) and then softened it within that range with the darker red noting approximately the top 84.1% and the blue filling in for those below the 68th percentile. This should hopefully circumvent some of the mental high jinx that numbers give us and allow us to view things in tiers of color rather than hard numbers.
The reason for including the spread is that it can be useful when building rosters based off of perceived game scripts or getting a feel for the likely game script the field is considering.
I've chose to note the favorites and dogs, along with home or road, because nearly 35% of performances over 15, 20, and 25 points come from HOME FAVS with approximately another 27% and 22% coming from ROAD DOGS and ROAD FAVS respectively. In other words, HOME FAVS and ROAD DOGS are 1.5x as likely to meet these thresholds as the two alternatives, with HOME FAVS being the most likely. Additionally, when it comes to performances over 35, HOME FAVS and ROAD DOGS account for nearly 70% of all such games and do so at a very similar rate, though again, HOME FAVS are the most likely by a slight margin. If this is a lot to remember, I've colored HOME FAV and ROAD DOG in varying shades of red (which as always means "better"); HOME FAV is of course darkest because it's the best. HOME DOG is in blue because it's the worst and should probably be avoided under most circumstances.
Obviously, such things are not certainties and players from all four categories surpass these thresholds, but continually exploiting +EV (positive expected value) edges such as these, week in and week out, are often the difference between profitable and break even, and break even and losing in the long run.