YOUR FIRST LOOK INSIDE

Try it. Make up your own mind.

No account, card, or setup. Pick the part you want to explore.

MODEL / HARBOR VS METRO

A number you can question.

Start with the uncertain injury report. Change its chance, then its impact. They answer different questions.

Harbor wins56.5%

Metro wins43.5%

Calculated from the sample assumptions.
Not a real prediction or a verified edge.

Read the notes behind this model ↗

Fictional basketball notes for this sample. Harbor creates shots at the rim through its two ball handlers. Metro often allows the first defender to be beaten and has been slow helping from the weak side. Harbor also finds open shooters when the defense collapses. Its shooters move into space before the pass arrives, but shooting results can vary. Metro has an offensive rebounding advantage: it sends two players to the glass while Harbor tends to leak out early. Harbor has handled ball pressure well, although Metro forces more mistakes with its second unit. Metro’s starting center is questionable. We estimate a 35% chance he misses this game; if absent, the backup offers less rim protection. The injury report is not confirmed. These are authored assumptions, not real scouting or statistics.

How this sample was made: manually authored to let you inspect and edit a working tree. It is not an AI-generated result or evidence of prediction accuracy.

Follow the reasoning

Choose any factor to read its notes and inspect its numbers.

INSPECT THIS FACTOR

Metro’s starting center misses this game

Unconfirmed injury report. The 35% estimate is a judgment about availability. The impact concerns reduced rim protection if he is absent.

Chance this happens 35%

How likely is this exact statement to be true in this event?

Influence +30

On: Harbor has a shot-quality advantage

How strongly does it support (+) or oppose (−) this parent? This is an influence score, not percentage points added to the result.

Try this: make the injury absence more likely without changing its influence. Then reset and change only its influence. Watch the current model result below.

Current model result56.5% Harbor · 43.5% Metro

How does this factor affect the number?

This is an uncertain statement without children. You supply its chance. The model weights its influence by that chance before it reaches the parent.

35.0% × +30 = 10.5 weighted influence on Harbor has a shot-quality advantage.

That score is not percentage points. The model combines evidence and converts it into a bounded probability. Moving one injury estimate does not directly add that many points to the chance of winning.

This is a modeling rule, not evidence that the output is calibrated to real events. Your inputs and their relationships still need judgment. The sample does not establish AI quality.

Not sure what your percentage should be?

You can start with notes, build manually, or ask AI for a draft with a paid plan. Review the assumptions before treating the number as useful. You can also skip modeling and just track bets.

See a planned slate →