Slow work day, might as well do a write up for fun.
The following is how I try to work through picks when systems show conflicting information (i.e., your favorite SDQL scenarios pick
both sides of a game and you're not sure who to go with).
So as an example, two plays are active this week for the above string from post #1 in this thread, as we're coming out of the first bye week for NFL teams with
Seattle (-7) going to Washington and
Cincinnati (-1) going to New England.
The Seattle play is up against recent trends where teams coming off a publicly humiliating prime-time beatdown (MNF, TNF, SNF) where they lost by 8+ points or more, typically have rebounded and beaten the spread in their very next game (roughly 66-70% win rate ATS depending on line, week #, other variables).
In this case to go against the original post-bye trend you'd have to believe Kirk Cousins and that awful Washington defense can somehow rise up to handle Seattle in all three phases of the game, who by the way despite the reputation of being a "poor" away team, are actually 8-3 ATS in road games over their last 11 (including even the straight-up loss this season to SD), going back to December of 2012 (Wilson's first season as QB).
They've done even better on the road when they're laying larger points, showing that they do typically dominate the teams they're "supposed" to beat, even if some geographical distance is covered and they're away from their "12th man" cushion.
Regardless, this is a case where I would choose to ignore the 2+ other SDQL scenarios I have showing Washington as the pick due to the post-beatdown scenario that has done well over time, and still go with the other scenarios that are pointing to Seattle.
It's tough because that's a square amateur-hour clown pick that makes the typically losing public feel all warm and fuzzy to bet (choosing a big point lay on the current SB champion is about as square as it gets), but the truth is that the perception of a team being awful on the road versus the stark reality of SDQL data showing it isn't the case makes aligning myself with the dart-throwing public less of a barrier.
In the other example, we're up against nearly the exact same scenario as betting against Washington, with
New England being absolutely embarrassed and taken to the woodshed by Kansas City in a prime-time game (MNF), now hosting a rested and trending Cincinnati team coming off a bye and laying -1 despite being the away team in a hostile stadium.
The first thought I had just off the top of my head is that the Patriots can't possibly lose two straight games ATS that often in the recent Belichick/Brady era, particularly not when that second game is at their house, where they've been strong over time.
But feeding it all into SDQL showed that wasn't the case, and that since 2008 when the Patriots have had an away loss where the ATS loss margin was heavy, they typically have gone home and not outperformed the mean of any home team, winning roughly 50% ATS.
The problem is that the sample size is as you would assume, extremely small, fewer than six games, because the Patriots have been a good enough team the past decade that they don't often endure away losses that exceed the spread by such a large margin followed right away by a home game.
Betting Cincinnati is actually the statistically safer play than betting Seattle over Washington, simply because the post-bye scenario is a 70%+ winner with a great sample size of games with no other top-level scenarios conflicting on the pick.
Tough to pull the trigger on Cincy though, I'm having a hard time with it personally despite already going with Seattle at -7. It's a case of the heart and the head not aligning, just can't go with that similarly square play as easily as I should be able to due too much hero-worship by the media on Belichick/Brady/Patriots the past decade...clouding my decision-making ha.
Anyway, we all need to share more of the thinking of "why" we bet, or don't bet, our own SDQL scenarios, because finding the strings is only 50% of the solution to actually outperforming 52.4% and winning...the other half is figuring out the signal versus the noise, and ignoring the trends that simply don't apply while isolating the ones that do. That's the magic of using SDQL properly.
TLDR VERSION OF THIS POST: Two of my favorite SDQL scenarios are pointing to opposite teams this week, how I chose to bet despite the conflicts comes down to perception versus reality, and SDQL illustrating that reality by helping to cut through the fog.