***2023 Week 5 Results***

With Week 4 behind us, let's look back and see how the rankings performed with details on each metric and last season's results available here:

• Performance by tier (Ideal/Solid/Solid(Deeper)/Avoid)

• JohnyTheGreek-style accuracy against projections

• FantasyPros-style accuracy against FantasySharks projections and a new "Expert X"

A little up and down this week with many of the expected top performers underperforming and some of the deeper recommendations putting up solid numbers. Overall, Week 5 tied season highs in terms of Solid Tier players over 10 and 20 pts and massive hit rate across all recommended plays and lows in terms of Ideal Tier average and players over 20 pts. Also, none of the 45 recommended plays put up a 0 so I'm very happy about that for all of our sake! 

***2023 Week 5 Rankings***

Welcome to the Week 5 edition of the DT Deep Dive! Apologies for running a day late but my family and I were traveling Mon-Wed which threw everything off! 

For the season, the model is performing very well after some small tweaks as the season has worn on (more on those changes below). Average scoring, startable players, hits, and misses mostly follow expectations along tiers (Tier 2 has been overperforming expectations a bit). Recommended plays across all tiers are outperforming projections nearly 60% of the time with a 28% massive hit rate. If I give myself the same grace as others and call recommendations who scored within an assist of their projection "Accurate*", my accuracy rate jumps to 68.6%. Finally, the top end of my ratings are still outperforming  both "Expert X" and FantasySharks across 2 of 3 accuracy measures (average and % correct) with progress made last week to close the FantasyPros-style metric. Full 2023 season and weekly results available here.

Moving into Week 5, I have dug into the data a bit to model some new variables such as PFF's measure of opponent offensive line quality (not just pressure rates) and PFF's measure of overall team offense (alongside pace of play) to try to estimate team defensive snap counts. I have also added a negative value for the top performing O-Lines in the league. These remain much smaller values than those placed on snap shares, pass rush grades, and run defense grades (in that order). More analyses of all these metrics coming soon!

Come back over the weekend for some commentary on selected players but in the interest of getting this out before TNF...let's get to the players!

Ideal Tier is guys you would start in any format (and even consider in DL leagues that don't explicitly require DT).

Solid Tier is players are also expected to perform well and you can usually start to find some potential streamers in this group.

Solid (Deeper) Tier is mostly going to be relevant in 16 team leagues, leagues where you have to start 2 DT, or leagues with very deep benches.

Avoid Tier is mostly players who did not have strong enough snap counts in Week 1 to warrant consideration in leagues other than Best Ball that you are managing in-season.

Good luck!