Every score is a weighted average of five checks, each built from the numbers you enter — never a black box. Here's exactly what goes into it, and what it doesn't cover. For the principles behind these numbers rather than the formulas themselves, see how House Autopsy thinks.
The weights below are the default. Picking a strategy (Rental hold, House hack, BRRRR) or a weighting for a rental hold (Cash flow, Appreciation) shifts these weights to match what actually matters for that strategy — a BRRRR deal is scored 40% on Deal Structure, for instance, because the ARV discount is what makes the refinance work. Your own deal's How the score works, under The five checks, shows the exact weights it was scored against. Flip doesn't reweight these five checks at all — it uses a different set entirely, covered further down.
Cash Flow Health30%default
Your monthly cash flow after every expense — vacancy, management, CapEx reserve, repairs, taxes, insurance, and the mortgage — plus your cash-on-cash return. Properties with a rent-to-price ratio 50%+ above the market benchmark in a High or Very High crime area also get a 4% collection-loss deduction (unpaid rent, evictions, turnover) — a real cost the entered numbers alone don't capture.
This is the biggest single driver of your score, because a deal that loses money every month isn't a rental investment — it's a monthly bill with a house attached.
Debt Safety22%default
Debt Service Coverage Ratio (DSCR) and loan-to-value (LTV) — whether the property's income can cover its own mortgage payment, and how much equity cushion you have. All-cash purchases have no debt to measure, so this check scores cash-on-cash return instead and is labeled Capital Efficiency.
Lenders generally want DSCR above 1.0 for a reason: below that, the property can't pay its own debt from rent alone, regardless of how the rest of the deal looks.
Market Strength22%default
Your rent-to-price ratio and cap rate, each compared to a regional benchmark, plus a conservative baseline appreciation rate for the market (a planning assumption, not a forecast) — though appreciation only carries real weight in this check for strategies built around it; see the scoring table for the split.
The market sets the ceiling on what's realistically achievable — a strong deal in a weak market is still constrained by that market's fundamentals.
Property Risk12%default
Roof age, HVAC age, overall property age, foundation condition, and neighborhood crime rating.
These drive future maintenance costs and tenant quality. A 20-year-old roof isn't a maybe — it's a bill with an approximate due date.
Deal Structure14%default
How far below After-Repair Value (ARV) you're buying — the ARV discount is the only input this check scores. It's what creates equity at closing, but that equity figure and your interest rate aren't separately factored in here; rate instead feeds the Debt Safety check above.
Buying with a real margin of safety is what protects you if any other assumption in the deal turns out to be wrong.
Flip scoring — a different model
Selecting Flip on the Analyze tab doesn't reweight the five checks above — it switches to a completely different formula, built around what actually determines a flip's outcome instead of a landlord's monthly cash flow. Verdict thresholds (BUY/MAYBE/CAUTION/SKIP below) are the same for both; the checks underneath them aren't.
Margin of Safety35%
Your projected profit — sale price at ARV, minus purchase price, rehab costs, holding costs, and selling costs — expressed as a percentage of ARV.
This is what actually protects you if a rehab runs over budget or the sale takes longer than planned — a thin margin has no room to absorb either.
Rehab Risk20%
How well-scoped your rehab estimate is, adjusted down if your contingency reserve is thin.
An unverified guess and a contractor-quoted number carry very different risk even at the identical dollar figure.
Exit Risk20%
The same property-risk model rentals use — foundation condition, crime rating, system age (see Property Risk above) — reused here because a hard-to-sell, high-risk property is exactly what threatens your ability to exit.
A flip's risk isn't just financial — it's whatever makes the property harder to resell once the rehab is done.
Financing & Holding15%
Your holding costs — mortgage, taxes, insurance, utilities during rehab and sale — as a fraction of the equity the flip is creating.
A flip that takes twice as long to sell can quietly eat its own margin in holding costs alone.
Market Liquidity10%
How fast homes typically sell in your market (median days-on-market) — a faster market means a more predictable exit if your timeline slips.
The same rehab in a 25-day market and a 60-day market carries very different risk, even when every other number is identical.
From score to verdict
The five weighted checks combine into a single 0–10 score, shown to one decimal, which maps to a verdict:
BUY
7.8 – 10 Scores well on the numbers entered. Verify the rent comps and run the stress test before committing.
MAYBE
6.8 – 7.7 Solid deal with one or two areas to address — often the case on a first analysis, not a reason to walk away by itself.
CAUTION
5.3 – 6.7 Multiple flags that need to improve before committing. Useful for understanding exactly what a deal would require.
SKIP
Below 5.3 Doesn't pencil at current numbers. Not necessarily a bad property — just not a deal at this price and terms.
The exact cutoffs are 7.75, 6.75 and 5.25 on the unrounded score. Because scores are shown to one decimal, 7.8 is the lowest score you will see on a BUY.
Here's what that looks like for the sample deal:
Sample — real output, same engine you're about to use
MAYBEMost checks clear the bar — one needs a closer look.
-10.9%
asking price
$255,000
BUY threshold
$227,191
That's how far the price would need to drop to become a BUY.
Scored on five checks — cash flow, debt safety, market strength, property risk, deal structure.
If rent comes in 15% below estimate, cash flow drops to $108/mo — still positive, but the verdict drops to SKIP.See full analysis →
Directional analysis from House Autopsy's scoring method, applied to the numbers you entered. Not financial or legal advice. How the score works
What can override the verdict, regardless of score
The weighted score above is the default path to a verdict — but five conditions cap the verdict at CAUTION by default, no matter how high the score is otherwise (one of the five can be acknowledged and overridden — see below):
Severe or moderate foundation concerns
A very-high crime rating
A BRRRR refinance that doesn't recover your capital at standard terms (75% LTV, 3% closing costs)
A DSCR loan with coverage between 1.0× and 1.25× — DSCR loans typically require at least 1.25× to qualify
Cash flow at or below −$300/unit/month — caps the verdict by default; you can acknowledge and proceed past it if you have a specific reason (e.g. a house hack you're subsidizing intentionally), which is the only condition on this list you can override
Foundation and crime data do double duty here: they're part of Property Risk's 12% weight, but a severe reading on either one bypasses that weighting entirely rather than just dragging the average down.
Two more conditions don't change the verdict letter but attach an explanatory note so you're not left guessing why — by that point the weighted score is already low enough on its own to reflect it:
Slightly negative cash flow.
DSCR below 1.0× on any loan type.
The Cash Flow Health check maxes out once your monthly cash flow clears roughly 0.15% of purchase price per unit, with a $400/mo floor for lower-priced markets — so a $300,000 single-unit property needs about $450/mo to max this check, while anything under roughly $267,000 needs a flat $400/mo. Multifamily deals multiply that ceiling by the number of rented units — for a house hack, every unit counts, because the score rates the property as if it were fully rented (the rent for the unit you occupy is estimated from your other units when you don't enter it).
Separately, the risk warnings on the Checklist tab (flood zones, HOA rental restrictions, lead paint, tax reassessment risk, and similar state- and property-specific flags) are informational — they don't change the score or verdict directly. Several of them (short-term-rental legality, HOA rental restrictions, and state-specific tax or insurance risk) do lower your confidence rating instead, so an unverified risk can't read as "High confidence" even when the underlying score looks strong.
The rent field also shows HUD's Fair Market Rent for the property's ZIP and bedroom count (the ZIP's own Small Area figure where HUD publishes one, otherwise its metro or county). It's roughly the 40th-percentile rent, utilities included, for a standard unit: a sanity check, not a comp or an appraisal. A rent 25% or more above it is called out under the result and in Check before you offer, and it never changes the score or verdict. This product uses the HUD User Data API but is not endorsed or certified by HUD User.
What's estimated vs. what's known
Verdicts are directional signals, not guarantees — real estate outcomes can't be predicted with certainty. Some inputs (your purchase price, loan terms) are numbers you know exactly. Others (market rent-to-price benchmarks, appreciation trends, a property's true condition before inspection) are estimates that improve as you gather better information. The app surfaces a confidence level alongside every score specifically so you can tell which situation you're in — low confidence means the verdict may shift once real data (a contractor quote, a signed lease, an inspection report) replaces an estimate.
If you leave property taxes for us to estimate, we use the property's state: its average effective tax rate on owner-occupied homes (Tax Foundation, Property Taxes by State and County, 2026, March 2026, from the US Census Bureau, 2024 American Community Survey). In California, Illinois, New Jersey, New York and Texas, where a sale usually resets the assessment, we use the typical post-sale rate instead when it's higher. With no state, we use the region's rate. A state average can be well off for one county or one house, so the estimate stays marked as estimated and the checklist asks you to look up the real bill.
Want the formula behind a specific number instead of the whole score? See the cap rate calculator for one worked example. For definitions of the terms used above, see the glossary.