Every number on this page is drawn from the most recent regression run, reflecting 15 market tiers and 31,546 verified residential sales. The interactive diagnostics below let you explore how each community scores on COD, PRD, R², and the median ratio.
Updated November 27, 2025 3:38 PM with the latest verified sales.
Trusted data pipeline
Integrity and privacy drive every decision. We pull from public Skagit County sources, keep raw files in a locked PostgresGIS environment, and log the transformations so anyone can follow what changed. Only aggregated, purpose-built outputs leave the secure workspace.
We download parcel characteristics, sales registers, and spatial layers straight from the Skagit County assessor and GIS websites. In the same pass we ingest public restaurant reviews and permitting data to grow the Skagit Flavor Index.
Automations in our management commands normalize formats, handle missing values, reconcile duplicate parcels, and flag outliers. Each task records counts and warnings so it is clear what stayed, what was trimmed, and why.
Every property receives terrain features—elevation, slope, aspect—as well as distances to rivers, ferries, schools, utilities, and more. Those tags make neighborhood comparisons and site planning more grounded.
Clean parcels feed into a PostgresGIS layer joined with census demographics plus city, state, and federal budget data. That view keeps local economic pressure, capital projects, and revenue context next to each parcel.
The structured data powers Property Helper, Neighborhood Reports, the Flavor Index, Job Outlook, and custom reports for you—always scoped to the questions at hand.
AI assists with coding, research, data cleaning, parcel narratives, and automated LLM tooling, but humans verify every narrative because models can be wrong. When in doubt, we slow down and double-check.
Diagnostics snapshot
A compact, responsive view of the newest diagnostics. Colors key off IAAO targets so you can quickly scan which segments are on track.
| Segment | Sales | R² | COD | PRD | PRB | Median | Price band |
|---|---|---|---|---|---|---|---|
| Adjustment run data is still being prepared. | |||||||
Histogram of sale-to-value ratios so you can confirm the center, spread, and skew of the model.
A balanced model has a narrow bell curve, meaning extreme under- or over-valuations are rare.
Ideal: ratios stack between 0.95 and 1.05 with the median pinned near 1.00.
Features explained
Each predictor now lives in a single responsive row with room for future stats, so you can skim their roles at a glance.
| Predictor | Why it matters | Driver group | Direction | Importance | Stat A (TBD) | Stat B (TBD) |
|---|---|---|---|---|---|---|
|
Living area
LOG_AREA
|
We take the natural log of finished square footage so the model reads size as a percent change. It keeps very large hom… | Size & Layout | Pulls ratios lower |
0.6%
Model share
|
— | — |
|
Effective age
LOG_AGE
|
Older homes often sell at a discount, but the impact tapers as properties age. Using the logged age captures that quick… | Age & Depreciation | Pushes ratios higher |
4.4%
Model share
|
— | — |
|
Build quality
QUALITY_SCORE
|
Quality scores summarize materials, finishes, and workmanship. Higher scores usually translate to higher values even af… | General | Mixed influence |
—
|
— | — |
|
Condition
CONDITION_SCORE
|
Condition measures upkeep and recent renovations. Well-maintained homes sell closer to market benchmarks than deferred-… | Quality & Condition | Pulls ratios lower |
1.4%
Model share
|
— | — |
|
Time trend
T
|
Monthly time steps keep the regression synced with market movement. They also prevent stale sales from skewing a hot ma… | Time & Market Cycle | Pulls ratios lower |
4.3%
Model share
|
— | — |
|
Land share
LAND_SHARE
|
This feature captures how much of the total value sits in the land component. It helps explain valuation bias between v… | Lot & Land | Pushes ratios higher |
3.4%
Model share
|
— | — |
|
Garage amenity
HAS_GARAGE
|
Simple indicator variables such as garages, basements, or views still matter. They make sure basic amenities stay value… | General | Mixed influence |
—
|
— | — |
|
Size × time interaction
AREA_TIME
|
Interactions let us test if certain home types appreciate differently. Here we watch whether larger homes move faster o… | Time & Market Cycle | Pulls ratios lower |
7.8%
Model share
|
— | — |
|
Log Total Mv
LOG_TOTAL_MV
|
Predictor surfaced in the latest regression run. | Other | Pulls ratios lower |
15.3%
Model share
|
— | — |
|
Log Total Mv Sq
LOG_TOTAL_MV_SQ
|
Predictor surfaced in the latest regression run. | Other | Pulls ratios lower |
13.0%
Model share
|
— | — |
|
Mv T3
MV_T3
|
Predictor surfaced in the latest regression run. | Other | Pulls ratios lower |
6.0%
Model share
|
— | — |
|
Mv Sq T3
MV_SQ_T3
|
Predictor surfaced in the latest regression run. | Other | Pulls ratios lower |
6.0%
Model share
|
— | — |
|
T Sq
T_SQ
|
Predictor surfaced in the latest regression run. | Time & Market Cycle | Pulls ratios lower |
5.4%
Model share
|
— | — |
|
Mv T1
MV_T1
|
Predictor surfaced in the latest regression run. | Other | Pulls ratios lower |
4.1%
Model share
|
— | — |
|
Mv Sq T1
MV_SQ_T1
|
Predictor surfaced in the latest regression run. | Other | Pulls ratios lower |
4.1%
Model share
|
— | — |
|
Area Quality
AREA_QUALITY
|
Predictor surfaced in the latest regression run. | Quality & Condition | Pulls ratios lower |
3.3%
Model share
|
— | — |
|
Area Condition
AREA_CONDITION
|
Predictor surfaced in the latest regression run. | Quality & Condition | Pulls ratios lower |
2.9%
Model share
|
— | — |
|
T Sq Mv T2
T_SQ_MV_T2
|
Predictor surfaced in the latest regression run. | Time & Market Cycle | Pulls ratios lower |
2.7%
Model share
|
— | — |
|
T Sq T2
T_SQ_T2
|
Predictor surfaced in the latest regression run. | Time & Market Cycle | Pulls ratios lower |
2.6%
Model share
|
— | — |
|
Mv T2
MV_T2
|
Predictor surfaced in the latest regression run. | Other | Pulls ratios lower |
2.5%
Model share
|
— | — |
|
Value Time
VALUE_TIME
|
Predictor surfaced in the latest regression run. | Time & Market Cycle | Pulls ratios lower |
1.5%
Model share
|
— | — |
|
T Mv T1
T_MV_T1
|
Predictor surfaced in the latest regression run. | Time & Market Cycle | Pulls ratios lower |
1.4%
Model share
|
— | — |
|
Mv Sq T2
MV_SQ_T2
|
Predictor surfaced in the latest regression run. | Other | Pulls ratios lower |
1.3%
Model share
|
— | — |
|
T T1
T_T1
|
Predictor surfaced in the latest regression run. | Time & Market Cycle | Pulls ratios lower |
1.3%
Model share
|
— | — |
|
T Sq Mv T1
T_SQ_MV_T1
|
Predictor surfaced in the latest regression run. | Time & Market Cycle | Pulls ratios lower |
1.2%
Model share
|
— | — |
|
T Mv T2
T_MV_T2
|
Predictor surfaced in the latest regression run. | Time & Market Cycle | Pulls ratios lower |
1.2%
Model share
|
— | — |
|
Log Area Sq
LOG_AREA_SQ
|
Predictor surfaced in the latest regression run. | Size & Layout | Pulls ratios lower |
1.1%
Model share
|
— | — |
|
Land Time
LAND_TIME
|
Predictor surfaced in the latest regression run. | Time & Market Cycle | Pulls ratios lower |
1.0%
Model share
|
— | — |