The 2026 Rookie Guide is LIVE 🚨
📝 Comprehensive analysis on EVERY fantasy-relevant rookie
🤖 Pre-NFL Draft, as well as Post-NFL Draft Rankings curated by Walter’s data science, as well as our team of experts
🏈 Ideal landing spots for every prospect, as well as ceiling and floor player comparisons
🏆 While Samps is the head writer - the entire WalterPicks team will have “My Guys” in the guide! These are players that we have the most belief in. Every contributor has added reasoning behind their selections that can also be found within the guide.
🏀 March Madness Bracket Analysis
Last week we published an analysis on 28k+ brackets created in the WalterPicks app, comparing the bracket results from those manually completed by users to those that were finished with the WP autofill feature.
With a sample of over 341K Sweet 16 and Elite 8 picks, here is a continuation of that analysis!
Humans Are Pulling Away 🏃➡️
Human picks: 67.7% correct across all rounds
WP Autofill picks: 60.4% correct across all rounds
The Duke Loss was DEVASTATING
The single game that crushed the most brackets was #2 UConn over #1 Duke. Only 11.5% of brackets had UConn winning that matchup.
Duke was the second-most popular champion pick (14.6% of all brackets).
Among the 1,569 brackets that nailed exactly 3 of 4 Final Four teams. 50.9% had Duke as their miss - more than the next 5 teams combined.
The Final Four
Our 2026 Final Four: Illinois, Michigan, Connecticut, Arizona.
Out of 28,478 brackets, only 60 got all four correct - that is 0.2%. WP Autofill and humans were dead even here (0.2% each).
1,569 brackets (5.5%) got 3 out of 4 right. Humans were significantly better here - 7.3% of human brackets got 3 of 4 vs. only 3.2% of WP Autofill brackets.
Among those 3 of 4 brackets:
63.5% missed Connecticut - UConn was picked as champion by only 2.6% of all brackets.
22.1% missed Illinois
7.3% missed Arizona and 7.1% missed Michigan - Everybody had these teams.
WP Autofill Still Better at Spotting Upsets
There were 2 true upsets (seed gap greater than 2) in the Sweet 16 and Elite 8, and WP Autofill called both at much higher rates:
Upset | WP Autofill picked it | Humans picked it |
|---|---|---|
#9 Iowa over #4 Nebraska (Sweet 16) | 8.1% | 0.7% |
#6 Tennessee over #2 Iowa State (Sweet 16) | 18.3% | 3.7% |
Iowa over Nebraska was the biggest miss of the entire tournament so far - only 4.3% of all brackets predicted it. WP Autofill was 11x more likely than humans to have it right. The Tennessee over Iowa State upset was similarly lopsided - WP Autofill was 5x more likely to call it.
Games That Stumped Everyone
#9 Iowa over #4 Nebraska (Sweet 16) - only 4.3% predicted this
#6 Tennessee over #2 Iowa State (Sweet 16) - only 10.9% predicted this
#2 UConn over #1 Duke (Elite 8) - only 11.5% predicted this
#3 Illinois over #9 Iowa (Elite 8) - only 18.7% predicted this
#2 UConn over #3 Michigan State (Sweet 16) - only 30.3% predicted this
#3 Illinois over #2 Houston (Sweet 16) - only 31.1% predicted this
No single game in the Sweet 16 or Elite 8 was predicted correctly by more than 68.6% of brackets!
Best Brackets So Far
Through 4 rounds (max 1,280 points):
2 brackets tied for first at 1,110 points - both fully human-picked, both with perfect Final Fours
At 1,100 points: 1 bracket (human)
At 1,080 points: 1 bracket (human)
The first WP Autofill bracket appears at 1,070 points (tied for 5th)
In the Round 1-2 analysis, WP Autofill held 2 of the top 3 brackets. Now humans hold the top 4 spots and 12 of the top 15.
