How Scoring Works — REAC Photo-Check

The short version

Every property starts at 100. Each failed check subtracts points. The projected score is 100 minus the total points lost, divided by the number of units inspected.

What a fail costs

A fail costs more when it is more dangerous, and most inside a unit. These are HUD's published deduction values (Table 1 of the NSPIRE Scoring Notice) — the same ones behind every “−X pts” on a fail in the app:

Severity	In a unit	Inside (common areas)	Outside	HUD correction window
Life-Threatening	−60.0	−54.5	−49.6	24 hours
Severe	−14.8	−13.4	−12.2	30 days
Moderate	−5.5	−5.0	−4.5	30 days
Low	−2.4	−2.2	−2.0	30 days

A worked example

One unit, walked by your crew — so the sample is 1 unit. Two fails found:

Refrigerator not keeping food cold — Severe, in a unit → −14.8 points
Cracked window glass — Moderate, in a unit → −5.5 points

Total deduction: 14.8 + 5.5 = 20.3 points. Divided by the sample — 1 unit inspected — that's 20.3 points per unit.

Projected score (As Found): 100 − 20.3 = 79.7 → 80 · B · Passing (60 or above passes). One check first: is any single unit carrying 30 or more points per inspected unit? 20.3 is under 30, so HUD's 59-point cap does not apply here.

The crew fixes the refrigerator on site and photographs the repair. As Left now counts only the cracked glass: 100 − 5.5 = 94.5 → 95 · A. The gap between the two numbers is the day's work, proven with photos.

Sample size is why the same fail can look small or enormous: one Life-Threatening fail in a 1-unit sample costs the full 60 points (score 40); spread over a 10-unit sample it costs 6 (score 94). The app always shows the sample next to the score — and its sample-size simulator shows how the projection moves as the sample grows.

The rest of the rules

The same fail counts once

If the same kind of fail shows up more than once in one unit, HUD deducts it once for that unit. Building-wide fails deduct once per building, and outside fails deduct once for the whole property.

One bad unit can cap the score at 59

If a single unit's fails average 30 points or more per inspected unit, HUD caps the property score at 59 — a fail — no matter how good the other units are.

Passing is 60 or above

90–100 is an A, 80–89 a B, 70–79 a C, 60–69 a D, and below 60 is an F — the property fails. At 30 or below, HUD may also refer the property for enforcement action.

Smoke and CO alarms: a flag, not points

HUD deducts 0 points for smoke alarm and carbon-monoxide defects. Instead the score carries a mark — * for smoke, + for CO. They are still life-safety items on a 24-hour fix clock.

As Found vs As Left

“As Found” is the projected score with the fails exactly as your crew found them. “As Left” recalculates after the fixes marked done on site or verified by a manager — it shows where the property stands now. HUD itself only scores what its inspector observes; As Left is this app's readiness view.

The units you walked are the sample

HUD inspects a sample of units and divides by that sample. Here, the units your crew actually walked are the sample — walking more units makes the projection steadier.

Projected estimate using HUD's published NSPIRE scoring model — not an official HUD score. Only HUD issues official scores; your official result also depends on which units the inspector samples. What the projection gives you is every fail priced the way the inspector will price it — no surprises about what matters most.

Want to price a fail yourself? The free point calculator runs this same math — pick a severity and a location, set the sample size, and see the deduction and the projected score. And every checkpoint behind these numbers is listed on the free reference page.

How scoring works.