Week 4 Let's Bring It In

September 29, 2026

We’re in full swing now!

While the chaotic season rolls on, the website meanwhile has a few updates to help you:

1. TAPsheets let you tailor your ROS rankings! No longer do you need to rely on default ROS player rankings. You can use “Edit Rankings” to drag-n-drop players— for example into an order that matches your favorite expert ROS rankings.

2. TAPsheets have “Trade Chart” view! Now you can choose “Trade Chart” view— which some of you are used to— and was much requested.

3. Kicker “Hand-Picked” Rankings. Now there is a new sorting option, to view Kickers in my own order, modified by my personal discretion. This goes against a long-standing feature of the website— that it’s always shown the raw model projection outputs with live updates: as Vegas lines change and weather forecasts shift. My Hand-Picked rankings get updated maybe once per day, but they might still help some of you struggling to decide, especially one a lot of kickers are very close in projection— forming single, long tiers.

If these are useful for you, or if you have other ideas, let me know at the “Ask on Reddit” post, at the bottom of any page!

Week 3 Accuracy Recap

For all expert rankers, week 3 was absolutely terrible for accuracy.

It was the hardest and worst week to rank correctly, for all. That applies to D/ST, QB, and Kicker.

From my experience with prior years, it’s surprising that we haven’t hit on a nice, clean predictable week yet.

This feels eerily more like the 2022 season, when it was such a struggle to figure out the strengths and weaknesses of each team.

So, more than ever, I think this is when we need to feel we can trust in the process.

In weeks 2 and 3, the Subvertadown models were keeping well ahead of the others, but now let’s look at week 3:

• D/ST: This was the only sore spot for my models, in week 3. After 2 stellar weeks, the models made a bad call for the Bengals, and the Vikings proved that the other rankers were more-correct. The only reason I ended up middle-of-the-pack for accuracy was based on other calls (like being lower on the Eagles). The learning I take away from this is being too confident in a new defensive unit after only 2 weeks. Note that Vegas lines were upset by this, too. The Bengals were supposed to hold the Steelers to 19 points, not 30. It was a surprise for all; but we could have protected better against the risk.

• Kicker: Of all things, Kickers are the position performing best for Subvertadown. Everyone else is really struggling here. The only top-ranked bust was Smack, last week— otherwise you should have done okay or great.

• QB: This week, I saw only 1 expert do clearly better than my model with his QB rankings. In a strange week, QBs averaged a very high 17.7 points. This meant that most expert rankers failed in accuracy mostly with the bottom-ranked QBs: A lot of low-ranked QBs actually performed quite well.

• Survivor: Finally a week with no lost pathways— All survived.

• Betting Lines: Returned +18% ROI, bringing the season to net +6% ROI, which is normal and on target.

Now I can also comment more on the FLEX rankings

• RB: Nothing special, just kept pace with other top sources. I’ve noticed everyone tends to do similarly in RB accuracy.

• WR: A good week! I was ahead of the curve with good accuracy.

• TE: So far, this is where the model really shines. Two weeks in a row of not seeing anyone else come close, on accuracy measurements.

That’s pretty cool to me! This is an all-new methodology this year, and I’m glad it’s working out.

Good luck everyone, now we really need to focus and make Week 4 a blowout win for streaming. Fingers crossed. :-|

/Subvertadown