About The Round Robin

I'm Grant. The Round Robin is what sixteen years of second-guessing the selection committee turned into. I wanted to know how they made their calls, built some tools to work it out, and then couldn't stop myself from doing the same for the Hall of Fame and the Oscars.

Background

I've filled out a bracket projection every year for sixteen years, long before any of it was a website. For most of that time it was called Primetime Bracketology, and I renamed it The Round Robin in 2024. Bracket Matrix, which grades bracketologists against the real field, has tracked my picks the whole way. In my best year I finished in the top ten on their Best Overall Bracketologist Composite. In plenty of other years I was nowhere close. I kept doing it anyway, mostly because I wanted to get it right. What I never had was a tool that worked the way I did.

The problem

Every season went the same way. I'd have NET rankings, quad records and team sheets spread across four windows and try to keep all of it in my head while I placed teams. The data was all public. I just had nowhere good to use it.

Bracket Studio

So I built the tool I wanted, and tried to make it usable by someone who hasn't spent sixteen years on this. Bracket Studio runs on live data: Warren Nolan's NET resume data during the season and ESPN's BPI in the preseason. You work through all 76 teams the way the committee does, with every metric on one screen.

Bracket Performance Calculator

Picking the right 76 teams is the easy part. The Bracket Performance Calculator also grades how close you got each seed, with the most points for an exact line. It runs on the Paymon Scoring System, which I built and which now lives here.

The name

In a round robin everybody plays everybody, which makes it about the fairest format there is. That's how I think these decisions should be judged, and it's how I judge my own models. Every forecast on the site gets scored against what really happened, and the scores stay up.

Right now that covers three decisions. For March, there's Bracket Studio and the Bracket Performance Calculator. For the Hall of Fame, there's HoFW, a position- and era-adjusted model tested against every real class since 2005. For the Oscars, there's a model that follows the race from the preseason through every precursor award.

How it gets built

It's just me. I use AI for work that would otherwise need a team, which is how one person can keep three models and their data running. The picks, and the judgment calls behind the models, are mine.

What's next

No roadmap to promise. If I find another decision made by a committee or a voting body that people argue about every year, I'll probably build a model for it, and you'll see it here while it's still half-finished.

Data and attribution

Bracket Studio is powered by publicly available team-resume data from WarrenNolan.com, including the NET Nitty Gritty report and team-sheet context used throughout the tool.

The Hall of Fame database and HoFW model are built on nflverse/nfl_data_py (which bundles Pro-Football-Reference draft and seasonal stat data) for career stats and draft pedigree, plus Wikipedia for award tables, Pro Bowl/All-Pro honors, and Hall of Fame induction history. Both are free, public data sources; the model's known limitations are listed on the HoF Forecasts page itself.

My Oscars forecast model is built on precursor award results (PGA, DGA, SAG, WGA, BAFTA, Golden Globes, Critics' Choice), plus Wikipedia for historical nomination and win data, and the TMDB and OMDb APIs for film metadata. This product uses the TMDB API but is not endorsed or certified by TMDB.

I pull that data together so you can work with it in one place. Bracket Studio's data refreshes nightly, and the timestamps in the app show the last successful update.

The Round Robin is not affiliated with, endorsed by, or sponsored by the NCAA, the Pro Football Hall of Fame, the Academy of Motion Picture Arts and Sciences, Warren Nolan, or any member institution or league. All bracketology, NFL Hall of Fame, and Academy Awards terminology is used editorially and for informational purposes.

If you spot a data problem or an attribution issue, email me and I'll look into it.

Support the site

This is a side project I do because I like it. If you get something out of the tools, chipping in helps cover the hosting and data costs.

Support the site

Contact

Found a bug, spotted bad data, or think the committee got it right and I got it wrong? Email me.

Email Me