01
What can happen?
Start with the entire outcome distribution: floor, likely range, ceiling, and probability above your own line.
Explore a distribution → Full player ranges · Week 4
The average
is not
the player.
Don't ask what he'll score. Ask what he can be. We publish the floor, the ceiling, and the inputs we will not hide.
16
Games this slate
1,688
Player forecasts
P10—P90
Decision range
Visible
Inputs + limits
Carol M. Highsmith · Library of Congress
How to use Backfield IQ
Start with the range.
Then inspect why.
Fast answers for users in a hurry; an inspectable chain for anyone who wants to challenge the inputs. Both begin from the same versioned forecast.
02
Why that shape?
Follow the football chain from participation and team volume through teammate allocation and conversion.
Open the explanation →03
What do I believe?
Change an explicit assumption and see a clearly labeled stress test—never a disguised model rerun.
Test an assumption →04
What supports it?
See the historical record, the five pick markets, and what we will not call a bet.
Inspect the evidence →
The entire game · one probability system Kenneth C. Zirkel · CC BY 4.0
What point projections hide
The average is not the answer.
A 12.4-point projection cannot tell you whether this player has a usable floor or a boom-bust tail. A market line cannot show how much probability sits on either side. The distribution can.
01
Set your threshold
Ask the question your lineup or market actually needs: what is the chance this player clears my number?
02
Read the shape
See the floor, central range, ceiling, and tail probability instead of treating the mean as destiny.
03
Audit the football
Trace availability, team volume, teammate competition, protection context, and conversion—with unavailable inputs called out.
Market evidence, with boundaries
Five markets. No borrowed credibility.
- +153.8u
- 2025 net units
- 8.1%
- Portfolio ROI
- 5
- Portfolio markets
Five-model portfolio, 2025 replay: 1,896 Sunday picks, one unit each. 2026 Weeks 1–3, today's models re-run on each week's pregame inputs: +25.2u on 325 picks, scored at every book. Past results are not a staking forecast.
Inspect model evidence →Access
The research is free.
The picks are not.
Every forecast, every explanation, the methodology, and the full 2025 track record are open, with no account. A distribution loses nothing by being read. A qualified pick does — publishing one at scale moves the number it was priced against — so the selections and the API sit behind a $50/month subscription.
Everyone, no account
The research
Freeno card, no sign-up
See the range and the reasoning.
- ✓Weekly player outcome distributions
- ✓Start / sit and ceiling-versus-floor comparisons
- ✓Participation, teammate competition, and matchup explanations
- ✓Editable assumption stress tests
- ✓The complete methodology and public model validation
Published openly · Frozen and dated
Props and prediction markets
By request
Beat the Market
$50per month
Price the whole distribution.
- ✓Everything published openly, plus:
- ✓Qualified, timestamped picks when a frozen rule clears
- ✓Displayed line, fair probability, threshold, and frozen pregame signal
- ✓API access to every relevant player–market pair on the current slate
- ✓Distribution summaries, exact PMFs where served, availability, and model status
No self-serve signup · Set up by hand
Every roster
Every team. Every role.
Current rosters, player histories, and the limits behind every forecast.