One team a week, and you never get to use it again.
See the full 32-team league across the whole season, what each week is worth, and what
picking the biggest favourite today costs you in November.
The grid is deliberately not drawn when the data fails its checks. A grid built on broken data would look correct and be wrong.
Data
Your pool
The week you are picking for right now.
How deep you expect the pool to run. Plan to that week, not to week 18 by default.
Recorded with your settings for context. It does not change the numbers below, because this version defines no pool equity model.
Teams you have already used
Select a team to take it out of every week. You can also select the team name in the grid below. Selected teams are struck through.
Games you will not pick
These do not change a single probability. The posted price has already paid for the road trip, the divisional familiarity and the short week. A filter only tells the optimizer which games to choose from, so switching one on can only make the returned path survive less often, and the tool shows you by exactly how much. A pick you make yourself in Freestyle always overrules a filter.
Pick rates, this week only
Three alternative paths
Three ways to read the same numbers, not one answer pretending to be certain. Choose which one to mark in the grid below.
Freestyle: pick your own path
Choose any week yourself and the optimizer fills every week around it. Select a team in the grid below, or use the menus here. Each card then shows what your choices cost against leaving that week to the optimizer, which is the number worth looking at.
The season grid
The grid scrolls sideways. On a small screen, swipe it or use the arrow keys once it has focus.
Everything you enter lives in the page address. There is no account and none of your picks are stored. Copy the link to come back to exactly this state, or send it to someone in your pool. Email signups are not open yet.
How the numbers are made
| Setting | Meaning | Value |
|---|
Questions
Why is the biggest favourite not always the right survivor pick?
Because you only get to use each team once. A team that wins 78 percent of the time this week might win 90 percent of the time in week 12, against a much weaker opponent. Using it now spends the stronger week for the weaker one. That difference is the opportunity cost, and it is what the Balanced path is built to weigh.
How is the win probability calculated?
Where both moneylines are posted, they are converted to implied probabilities and the market margin is removed, so the two sides of a game sum to one. That is a market price. Where they are not, an Elo-style projection is used instead, with 55 rating points of home advantage, built from completed results in recent seasons. That is a model projection. Every number on the page is labelled with which of the two it is, per game rather than per week.
Is this the optimal path?
It is optimal for the objective you picked, and that is a narrower claim than it sounds. Choosing a whole season of teams is a maximum-weight assignment problem, so it is solved exactly rather than approximated. What that does not mean is that the season will follow it. The probabilities are estimates, so you get three alternative paths rather than one falsely certain answer, plus a sensitivity check showing how little it takes to change the current week.
Where does the pick rate come from?
You enter it yourself. Pick popularity is not redistributable, so nothing is scraped. The Contrarian path takes your current week entries only, and every later week is scored exactly like the Balanced path.
Is anything saved or sent anywhere?
No. Everything runs in your browser: there is no account, no signup, and none of your picks are stored. Your settings live in the page address, so copying the link is what saves your work.
Is this betting advice?
No. This is a decision science tool. It is educational, it never tells you to wager anything, and it links to no wagering service. The decisions are yours.
Signups are not open yet.
