Plain-language glossary

Every word on the demo translated into something you can hand to a non-expert. The left column is what you see on the pages; the middle column is the technical name engineers and researchers use; the right column is one sentence of what it means.

Back to the start.

What you see	Technical name	What it means
Your goal	Objective	A bundle of numbers that says what a good deal means for you: how much money you have, how much risk you accept, what kinds of deals fit, which states you'll buy in, what return you want.
Filter step	Gate	One of the four stages a parcel goes through: hard exclusions (e.g. flood zone) → extra data lookup → estimated value → buy/look-again/skip recommendation.
Buy	BUY	The math says the deal clears both your profit floor and your annual rate-of-return floor with room to spare.
Look again	REVIEW	The deal clears one of your two floors but the other is within striking distance. Worth a second pair of eyes.
Skip	PASS	The deal doesn't clear your floors, or it's outside what you said you'll buy (wrong state, wrong size, wrong wetlands percentage, etc.).
Estimated value	AVM / valuation_estimate	The system's best guess of what the parcel would resell for retail. Derived from comparable sales nearby; carries a low/high band.
Buy price	acquisition_cost	What the operator pays to acquire the parcel. Typically a fraction of the estimated value — that fraction is one of the dials on your goal page (default 40% for a flip).
Profit (estimated)	margin	Estimated value minus buy price minus the cost of selling. Doesn't account for taxes or carry costs.
Annual rate of return	IRR	If you put $1 in, what fraction comes back per year — annualised. 25% means $1 turns into $1.25 over one year; over two years it'd be ~$1.56.
Total return on money spent	ROI	Total profit divided by total dollars deployed. Doesn't annualise — a 50% ROI over five years and a 50% ROI over one year both report 50%.
Chance you lose money	P(loss)	Out of every 100 pretend markets the simulator runs, in how many would the portfolio end up in the red.
Chance you run out of money	P(ruin)	Out of every 100 pretend markets, in how many would the portfolio lose so much it can't keep operating (loses >= 50% of what you spent).
Pretend markets / market dice rolls	Monte Carlo scenarios	The simulator rolls dice 5,000+ times for each parcel — different sale prices, recessions, catastrophes. Each roll is one possible future; together they form the range you see.
Range of outcomes	Reservoir / scenario distribution	The full spread of what happened across all the pretend markets. The average hides too much — the range tells you whether the average is reliable or volatile.
Best 5% of futures	P95	If you ran 100 pretend markets, the 95th-best result. Tells you what the upside looks like in a good market.
Worst 5% of futures	P5	If you ran 100 pretend markets, the 5th-best result. Tells you what the downside looks like in a bad market.
Real data	REAL provenance	The number was fit from actual data the system has loaded (county sale records, NOAA storm history, etc.). Higher trust.
Best-guess number	ASSUMPTION provenance	The number wasn't fit from data yet; it's a sourced placeholder (we cite where it came from — paper, federal table, research prior). Lower trust until upgraded to real.
Your input	USER_INPUT provenance	You typed this on the Your-goal page. The system trusts what you say; nothing to fit.
Where each number came from	Provenance table	Every important number the system uses is tagged Real / Best-guess / Your input with a link to the source. So you always know which numbers to trust most.
Confidence	overall_confidence	The advisor's honest summary of how well-grounded its proposal is. High = most numbers are real-fit. Medium = several numbers are still best-guesses. Low = it can't ground its answer right now.
Catastrophe (cat event)	cat_event_per_month_hazard	How often a damaging weather/loss event hits the county per month (calibrated from NOAA Storm Events when loaded).
Catastrophe severity	cat_severity	When a catastrophe hits, what fraction of the deal value gets wiped out. Still a best-guess pending more real data.
Sale ratio	sale_ratio	How much of the estimated value the parcel actually sells for. Below 1.0 = sold below estimate; above 1.0 = sold above. Still a best-guess pending 2+ sales per parcel.
Hold time	hold_months	How long the operator owns the parcel before selling. Still a best-guess pending consecutive sales per parcel.
Deal box	deal_box	The kinds of deals you'll accept: price range, acreage range, allowed flood zones, wetlands percentage cap, whether HOA-restricted parcels count.