Methodology · v2.3 · published April 12, 2026

How we score North Carolina landlords.

Six factors. Public sources. Nightly recalculation. We do not accept money from landlords. The math is open — here it is.

On this page01Overview02The formula03Each factor in detail04Data sources05When the score changes06Disputes07What we do not measure08Changelog

Overview

Every North Carolina landlord in our database is assigned a numeric score between 0 and 100 and a corresponding letter grade. The score is calculated entirely from observable behavior — not from reviews, ratings, or anything that can be bought. We do not accept payments from landlords for any reason.

This page documents how the score is calculated. It is updated whenever the formula changes. The most recent version is v2.3 (published April 12, 2026). Prior versions are available in the changelog.

Want to dig in? The scoring code is open source. The data sources are listed below. If you find a bug, write us at methodology@lawpassport.com.

The formula

The final score is a weighted average of six factors. Each factor returns a value between 0 and 10. We multiply by the weight, sum, and scale to 0–100.

Factor
Weight
Max points
01
Responsiveness
30%
300
02
Resolution Rate
25%
250
03
Resolution Speed
15%
150
04
Complaint Density
15%
150
05
Severity Adjustment
10%
100
06
Recency Decay
5%
50
Total
100%
1,000

Letter grades are assigned from the final score using fixed thresholds, with one exception: any landlord with fewer than three observable cases in the past 24 months is shown as provisional and not graded.

A
90–100
B
80–89
C
60–79
D
40–59
F
0–39

Each factor, in detail

01

Responsiveness

weight 30%

Did the landlord respond when contacted about a serious issue?

We measure the median time, in calendar days, between a verified complaint being raised and the first substantive landlord response. A response is substantive if it acknowledges the issue and proposes an action — even if that action is to refuse. Form auto-replies do not count.

Score
Means
9–10
Median ≤ 3 days
7–8
Median 4–7 days
5–6
Median 8–14 days
3–4
Median 15–30 days
0–2
Median > 30 days, or no response at all
02

Resolution Rate

weight 25%

When a tenant raised an issue, did the landlord eventually fix it?

Of all complaints we observed in the past 24 months, what percentage ended in a tenant-confirmed resolution? A resolution is the deposit being returned, the repair being completed, or — in the case of disputes — a written settlement signed by both parties.

Score
Means
9–10
≥ 80% confirmed resolutions
7–8
60–79%
5–6
40–59%
3–4
20–39%
0–2
< 20% confirmed resolutions
03

Resolution Speed

weight 15%

When they did resolve, how fast?

Median time between issue raised and confirmed resolution, in calendar days. Cases that did not resolve are excluded from this factor.

Score
Means
9–10
≤ 7 days median
7–8
8–21 days
5–6
22–45 days
3–4
46–90 days
0–2
> 90 days
04

Complaint Density

weight 15%

How many complaints, normalized by portfolio size?

Total verified complaints in the past 24 months, divided by the number of rental units in the portfolio. Smaller landlords are not punished for portfolio size; large landlords are not given a pass for it.

Score
Means
9–10
< 0.05 complaints per unit per year
7–8
0.05–0.15
5–6
0.16–0.35
3–4
0.36–0.75
0–2
> 0.75 complaints per unit per year
05

Severity Adjustment

weight 10%

Was anyone living in unsafe conditions?

Complaints involving habitability emergencies — no heat in winter, no water, mold with documented health impact, structural hazards — are weighted more heavily. We use a 1–5 severity rubric published in §4 of the data charter.

Score
Means
9–10
No severity-3+ complaints in 24 months
7–8
1–2 severity-3 events
5–6
3+ severity-3, or 1 severity-4
3–4
2+ severity-4 events
0–2
Any severity-5 event in past 24 months
06

Recency Decay

weight 5%

Are problems recent, or in the rear-view mirror?

Complaints in the past 6 months count 100%. Complaints 7–12 months ago count 75%. Complaints 13–24 months ago count 50%. Older than that, they fall out of the active calculation but remain in the public record.

Where the data comes from

The score combines public records and observed activity on LawPassport. Every data point on a profile is sourced. Click any number on a landlord profile to see where it came from.

NC Administrative Office of the Courts
refreshed nightly
Civil filings, summary ejectment (eviction) records, magistrate-court judgments
NC Secretary of State — Business Registration
refreshed weekly
LLC ownership, registered agents, principal addresses
County tax assessor offices (urban NC counties)
refreshed monthly
Property ownership, parcel records, transfer history
Municipal code-enforcement databases
refreshed nightly
Open and closed code violations, complaint history
HUD REAC inspection scores
refreshed quarterly
Federal inspection results for subsidized properties
LawPassport demand letters
refreshed real-time
Tenant-side filings sent through this platform — both the demand and the landlord's response (or lack of one)
Verified tenant reports
refreshed as submitted
Tenant-submitted complaints, verified against lease and address

When the score changes

Scores recalculate nightly. A new complaint, a new resolution, a new court filing, a new code violation — anything observed in the previous 24 hours is reflected by 6 AM the following morning.

Provisional landlords (fewer than three observable cases in 24 months) become graded as soon as a third case is observed and verified.

If a landlord disputes their score

Every profile has a Respond link. Landlords can dispute a specific complaint or the score as a whole. Disputes are reviewed by our records team within 5 business days. Three outcomes are possible:

  • Confirmed. Source documents support the dispute. We correct the record. The correction is logged and timestamped.
  • Annotated. The landlord's response is added to the public profile alongside the original complaint. Both views remain visible.
  • Rejected. The dispute does not survive review. The score does not change.

Our error-correction rate (the percentage of disputes that resulted in a confirmed change) for Q1 2026 was 11.4%. We publish this number every quarter.

What we don't measure

We deliberately do not include:

  • Tenant reviews or star ratings. Too easy to game in both directions.
  • Race, national origin, or any protected class of either landlord or tenant.
  • Eviction rates in isolation. Eviction filings appear on the profile as a data point, but do not directly change the score. A landlord with many filings and high responsiveness is treated differently than a landlord with many filings and low responsiveness.
  • Property type or neighborhood. We do not adjust by ZIP code, building age, or housing program participation. The behavior is the behavior.

Changelog

v2.3
Apr 12, 2026
Recency Decay added (5% weight). Complaint Density brackets tightened. Removed a depreciation factor for landlords with provisional-status histories.
v2.2
Jan 18, 2026
Severity Adjustment introduced. Replaces the old "emergency multiplier" that operated outside the weighted average.
v2.1
Oct 04, 2025
Resolution Speed brackets shortened by ~30% across the board after audit found grade inflation in fast-moving disputes.
v2.0
Jul 21, 2025
First version published openly. Prior internal versions are not graded — affected profiles re-scored at launch.
If you spot a bug

We want to be wrong as rarely as possible.

If a number on a landlord profile looks off, or you think the formula treats a real case unfairly, write us. Real names get real replies.

methodology@lawpassport.com