Built by someone who wants to help you look before you jump.

I spent two years working through it. I bounced through every stat you can think of — efficiency, success rate, drive quality, third-down conversion, every flavor of pace and tempo. I landed somewhere I didn't expect: time of possession.

Not just time of possession itself — but how elastic each team's TOP is against different opponents. Some teams hold the ball no matter who they play. Some teams' possession gets pushed around. That elasticity, weighted by opponent strength, became the basis of what's now Pace+.

For a full year I bet a dollar on every spread, total, and moneyline I could get to. Real money, real markets, every game. Pace+ was strong. It was also incomplete — there were specific kinds of matchups where it had blind spots.

That's what made me build the rest of the system. The question wasn't whether Pace+ could replace the public models. It was whether putting five reads on every game — SP+, Elo, PPA, advanced efficiency, and Pace+ — could fill the holes Pace+ alone couldn't.

The answer turned out to be yes. But not in the way I expected.

The A+ picks hit 75%. The moneylines were near-perfect. The featured plays cleared 90%. I'm a statistician. I knew immediately the numbers were contaminated. Anyone whose backtest looks that good has a bug, not a thesis.

What was broken was look-ahead bias. The system was using full-season ratings to "predict" games whose outcomes the ratings already knew. We rebuilt the rating pipeline so every Week N projection uses only data from Week N−1 and earlier. The numbers came in lower, and they were real.

Then we caught another version of the same bug in different parts of the model. Rebuilt again. Numbers came in lower again.

Then we caught data corruption in the rating refresh — duplicate rows polluting the inputs. Fixed it. Rebuilt the backtest a third time.

Then we added a stability filter. Every signal in the locked system has to hold up year-over-year — not just hit a single-season win rate, but produce comparable results in 2024 and 2025. Signals that didn't pass got cut.

What remains is the locked spec. The A+ Board hit 68.4% across two seasons with a 0.8 percentage-point difference between years. Three audit passes. Two rebuilds. The number that remains is real.

Most operations would have shipped the first model. We didn't, and that's the point.

There are a lot of people selling picks on the internet. Most of them cherry-pick their wins, hide their losses, and charge a monthly fee for the privilege. PressBox is built on a different premise entirely.

Every stat below is backtested against opening lines from tracked sportsbooks. No look-ahead bias, no favorable assumptions, no cherry-picking the good weeks.

Live today: the Board, the History, the Breakdown, and the Upcoming slate with live scores. Picks flow through the system on a 30-minute regrade cycle. Every grade change is journaled. Every line move is on the record.

What's coming after launch:

• A bet allocator — sizing each play against your bankroll and the tier's historical hit rate
• A parlay builder and grader — assemble multi-leg bets and see them graded in the same journal as everything else
• A user picks ledger — log your own bets alongside ours, with running profit/loss
• Other sports — NFL is the natural next step

The 2026 college football season is the first season under the locked spec. Picks drop in time for Week 0 (Aug 29). Release pricing — $3.99/wk, $11.99/mo, $34.99 Season Pass — locks for the life of your subscription. Subscribe today and your first charge is anchored to Aug 23, when picks start to drop. Pay nothing until then.