← Bulrix

How the Bulrix Score works

The public methodology behind the number. A score people can audit is a score people can cite. Current model: Bulrix Score 1.2 (July 2026).

What the score predicts

The Bulrix Score estimates the probability that a home suffers a major unplanned failure in the next 12 months.

“Major unplanned failure,” v1 definition: an unscheduled repair or replacement event of $1,000 or more, or any emergency dispatch (a burst pipe, no heat in a freeze, no cooling in August, an active roof leak, an electrical hazard). The definition is versioned and published here.

A high score is not a badge for using an app. It is a claim: this home is unlikely to surprise you.

The scale and the math

The score runs 300 to 850, always rendered as “Bulrix Score 742,” never a bare number. It uses standard scorecard scaling, the same mechanism under real credit scores: every 40 points, the odds of a surprise halve (PDO = 40). The anchor: a 700 means roughly 19-to-1 odds against a major failure, about a 5% twelve-month risk.

Formula: Score = 530 + 57.7 × ln(odds against), where 57.7 = 40 / ln(2).

SCORE12-MONTH FAILURE RISKREAD AS
780~1%A surprise is unlikely this decade
740~2.5%Comfortably ahead of trouble
700~5%The claimed-home median anchor
660~10%One meaningful gap
620~17%Several gaps compounding
580~30%A surprise is more likely than not within 3 years

Anchored, never curved. Scores map to absolute odds, not to percentile rank. Your number never falls because a neighbor improved; it only moves when your home’s risk moves. Your percentile rank among Bulrix homes lives in Standing, displayed separately alongside the top homes (first names only), and Standing can drift as the population improves.

The bands

Starting300 to 579Every home starts here.
Documented580 to 669The record exists.
Protected670 to 739Insurance, warranties, and readiness in place. The ceiling for self-reported records.
Current740 to 799Maintenance verified and up to date. Requires pro-verified evidence.
Kept800 to 850The state every listing wants to claim. Requires pro-verified evidence.

The self-report ceiling is 739. Current and Kept require photo- or pro-verified evidence, which keeps the top of the scale scarce and meaningful.

The five factors

Care history10%

On-time completion of due maintenance over the home’s Bulrix life, weighted by consequence and evidence tier. Late is worse than missed once long ago; verified beats self-report.

System condition30%

How much life each major system has consumed against its survival curve (roof by material, HVAC, water heater, panel, plumbing supply type). A funded plan, meaning a saved quote or a scheduled job, buys back up to 60% of the age-lost points, because a planned-for old system is a managed risk.

Record depth15%

Completeness of the home’s blueprint plus the length of the documented timeline. Ten years of records genuinely reduce buyer and actuarial uncertainty.

Protection10%

Insurance on file (an active subscription, a renewal within its grace window, or the policy document itself), warranties in force, inventory done, the home photographed. A renewal date is treated as a reminder anchor, not a coverage status: home insurance auto-renews, so a recently passed date reads as a renewal in progress, never as a lapse.

Open risk signals35%

Active unresolved issues: overdue critical tasks, known defects logged but unaddressed, expired coverage. The heaviest factor since model 1.1; still designed to sit at full points for most homes most of the time.

Every score renders with its top helping and top hurting factors in plain sentences, and the explanation is the to-do list: the same items are your Next Moves.

What never touches the score

The chronological age of the house: never. System age: yes. A 1968 home with a 3-year roof, an updated panel, and verified service outscores a neglected 2015 build. The score reflects stewardship, which the owner controls, not vintage, which they don’t.

Geography: never. Where you live shapes your task list (freeze prep, foundation watering), but it never adds or subtracts points. You cannot maintain your way out of a ZIP code, so a ZIP code cannot move your score.

Money spent: never directly. Dollars only matter through what they change: a system replaced, a task verified. A $30 verified filter change moves the score exactly as the same work performed expensively.

Only the home moves the score. No logins, no streaks, no referrals, no purchases, no sponsor.

Commercial money can sponsor prizes, events, and credits. It cannot buy, boost, or touch a point.

Calibration and stability

Model 1.0 is an actuarially grounded expert scorecard, the same way credit scoring itself began. Failure-odds priors come from published component survival curves and life-expectancy tables used across the inspection and engineering world (roof life by material, water heaters at 10 to 12 years, HVAC at 15 to 20, panel and supply-line risk profiles), plus deferred-maintenance multipliers.

Model 2.0 will recalibrate the factor weights and odds anchors against Bulrix-observed outcomes (emergency dispatches, unplanned jobs of $1,000 or more) and ship as a numbered, dated release.

Stability rules: scores refresh on events plus a monthly tick; single-month movement is capped at 40 points except for verified major events (a roof replaced, a system failed); no day-to-day jitter, ever.

Governance and changelog

Changes to this methodology ship as versioned releases with dates, never silently.

Methodology advisors: [PLACEHOLDER: a licensed inspector and an actuary will be named here.]

CHANGELOG
1.2 (July 16, 2026): insurance recognition corrected. Coverage now counts when an insurance subscription is active or within a 3-month renewal grace, or when the policy document itself is on file; a lapse is flagged only beyond the grace with no document. Filed documents now also count toward record depth.
1.1 (July 16, 2026): factor weights recalibrated. Open risk signals now carry the most weight (10% to 35%); care history reduced (35% to 10%). System condition, record depth, and protection unchanged. The odds anchors, bands, and the 739 self-report ceiling are unchanged.
1.0 (July 2026): the initial expert scorecard. Five factors (35 / 30 / 15 / 10 / 10), PDO 40, the 700 anchor at ~5% twelve-month risk, the 739 self-report ceiling.

The Bulrix Score is a stewardship metric computed from a home’s own record. It is not an appraisal, an inspection, or an insurance rating, and it is distinct from any credit score.