What the current model actually measures
Version 1.0 compares your answers with ten hand-set, subjective area profiles. It is a preference-sorting demonstration, not a measured neighborhood-quality index. The profile numbers are not government statistics, survey results, live housing prices, travel times, or independently validated ratings. NHScore has no documented empirical calibration for these initial values. Use the result to choose places to investigate, not to decide where to buy or rent.
Nine profile dimensions describe urban pace, quiet, walkability, food and evening activity, outdoor access, water access, central Tampa access, newer housing, and private space. A separate cost tier is also an editorial assumption, not an affordability test. None of these values evaluates who lives in an area.
The calculation, step by step
- Each answer supplies one or more target values and weights from the table below. Where two answers affect the same dimension, their targets are combined using a weighted average and their weights are added.
- For each area and each chosen dimension, calculate
abs(target − profile) / 100 × weight. Add those distances. - Add
abs(budget tier − area cost tier) / 4 × 4. Rent uses the rent tier; buy and “just exploring” use the buy tier. This symmetric distance can penalize a less-expensive area as well as a more-expensive one: it does not enforce a maximum spending limit. - Divide the total distance by the sum of the lifestyle weights plus the budget weight of 4. Subtract from 1, multiply by 100, round, and constrain to 0–100.
- Sort by that rounded score and show the first three. Equal scores retain the profile order below. Three results appear even if every profile is a poor approximation.
similarity = round(100 × (1 − weighted distance / total weight))For example, target walkability 95 versus profile 68 at weight 4 contributes 27 / 100 × 4 = 1.08 distance and 4 weight. A cost-tier difference of 1 adds 1 distance and 4 weight. Higher similarity means only a closer match within this model—not a probability of satisfaction or a statement that an area is better.
Display labels are also product choices: 82–100 is “Higher profile similarity,” 70–81 is “Moderate profile similarity,” and below 70 is “Lower profile similarity.” These thresholds are not validated decision cutoffs.
Answer targets and weights
Notation: dimension target / weight. These are explicit product choices, not evidence-based estimates of importance.
- Pace: city = urban 95/3, quiet 20/2; village = urban 65/2, quiet 52/1; suburban = urban 28/3, quiet 78/2; space = urban 10/3, quiet 95/3.
- Walking: essential = walk 95/4; nice to have = walk 68/2; low priority = walk 25/2.
- Food: high = food 95/3, quiet 38/1; balanced = food 65/2, quiet 65/1; retreat = food 30/1, quiet 90/3.
- Outdoors: water = water 98/4, outdoors 78/2; parks = outdoors 95/4, water 35/1; both = outdoors 82/3, water 78/3; low priority = outdoors 45/1, water 45/1.
- Central Tampa: very important = commute 98/4; somewhat = commute 68/2; rarely = commute 30/2. This does not model your workplace or actual travel time.
- Home: character = newer 18/3, space 48/1; new = newer 95/4, space 70/1; yard = space 95/4, urban 18/1; compact = space 20/3, urban 85/1.
- Budget: the five displayed ranges map in ascending order to tiers 1–5, with weight 4. Purchase ranges are under $350k, $350k–$500k, $500k–$750k, $750k–$1.1M, and $1.1M+. Rental ranges are under $1,800, $1,800–$2,300, $2,300–$3,000, $3,000–$4,000, and $4,000+ per month. These are interface bands, not current market quotations.
Every initial profile value
Values are subjective 0–100 settings; cost tiers are 1–5. “Commute” means central-Tampa proximity in this editorial model. Water access is not flood safety. Quiet is not crime safety. Higher is not universally better.
| Area | Urban | Quiet | Walk | Food | Outdoors | Water | Commute | Newer | Space | Buy / rent tier |
|---|---|---|---|---|---|---|---|---|---|---|
| Safety Harbor | 35 | 82 | 68 | 56 | 78 | 90 | 48 | 35 | 58 | 4 / 3 |
| Westchase | 25 | 78 | 48 | 52 | 70 | 34 | 72 | 78 | 72 | 4 / 4 |
| Hyde Park & SoHo | 82 | 42 | 90 | 92 | 46 | 72 | 90 | 28 | 25 | 5 / 5 |
| Seminole Heights | 72 | 46 | 62 | 88 | 48 | 24 | 86 | 20 | 43 | 3 / 3 |
| Dunedin | 48 | 67 | 82 | 72 | 90 | 98 | 32 | 22 | 44 | 4 / 4 |
| FishHawk Ranch | 12 | 92 | 35 | 28 | 93 | 20 | 22 | 92 | 88 | 4 / 3 |
| Carrollwood | 33 | 72 | 32 | 60 | 58 | 35 | 70 | 34 | 75 | 3 / 3 |
| Downtown & Water Street | 100 | 16 | 95 | 90 | 45 | 82 | 98 | 88 | 10 | 5 / 5 |
| Brandon | 38 | 55 | 20 | 58 | 36 | 12 | 60 | 55 | 68 | 2 / 2 |
| Riverview | 18 | 68 | 16 | 38 | 55 | 42 | 38 | 88 | 82 | 2 / 2 |
Official sources and what they do not establish
The Tampa address-research workflow links to FEMA, county evacuation resources, school districts, transit agencies, and insurance information. These support independent, address-specific checks. They do not supply or validate the numeric quiz profiles above. We do not import crime, school grades, resident demographics, or political data into the current match calculation.
Excluded criteria and the legacy system
Race, color, ethnicity, religion, national origin, sex, familial status, disability, age, marital status, sexual orientation, gender identity, and political affiliation are not model inputs. We do not infer them from addresses, schools, crime, or other proxies. The old prototype included political and demographic filters; that retired system does not power the current quiz and its public endpoints are removed.
Limitations and independent checks
- Ten broad areas are not a comprehensive set of Tampa Bay housing options. Labels combine neighborhoods, commercial districts, cities, and unincorporated areas; they are not equivalent statistical units.
- No property inventory, inspection, accessibility assessment, price feed, flood analysis, or route calculation runs in this quiz.
- Names and boundaries overlap. A downtown, subdivision, or street may differ greatly from its broad profile.
- The listed “closest dimensions” are the smallest raw differences, not the largest weighted contributions; they need not be close in absolute terms.
- A sponsor cannot pay to change the model, area order, or a user's score. A paid placement is separate from editorial research.
Version and corrections
Model 1.0 launched in August 2026. This September 25, 2026 review makes its assumptions explicit; it does not change the numbers or calculation. Labels now describe profile similarity rather than an “excellent” or “best” place to live. No automatic data refresh is implied. Material future input or algorithm changes should be versioned here.
Report a questionable value or factual error to mickey@nhscore.com with the page, area, disputed statement, and supporting source. We distinguish correcting a documented fact from changing an editorial assumption. See ownership and corrections.