Weighted scoring

When you compare properties, the AI layer scores each property in each category on 0–100. MARA's application code then does all of the arithmetic. The AI never calculates the weighted score or decides the ranking.

The formula

Weighted Score = Σ(Category Score × Category Weight)
                 ÷ Σ(Active Category Weights)

Because the divisor is the sum of the ACTIVE weights only, the result always stays on a 0–100 scale, and any category left at 0% is excluded entirely.

Worked example — one category

Investment Potential = 100%, everything else = 0%

Property A: Investment 82  ->  Weighted Score 82 / 100
Property B: Investment 65  ->  Weighted Score 65 / 100

The unselected categories do not drag the score down to 8.2 or 5. They are not part of the calculation at all.

Worked example — two categories

Investment Potential = 60%, Price = 40%

Investment 82 x 0.60 = 49.2
Price      70 x 0.40 = 28.0
                       ------
Weighted Score         77.2 / 100

Automatic normalization

Your sliders do not need to add up to 100. If you set Investment to 5 and Price to 3, MARA normalizes them:

Total selected weight = 8
Investment: 5 / 8 = 62.5%
Price:      3 / 8 = 37.5%