What shipped
Setback is a thermostat schedule painter with a heating-bill model behind it. Pick a brush temperature and drag across a 7-day by 24-hour grid to paint setpoints, colored from cold blue to warm orange. Copy Monday to the weekdays, Saturday to Sunday, or reset to a flat 68. Choose a climate (eight cities with approximate monthly normals and a day/night swing), a month from October to April, the house size and insulation, and the heating system: a gas furnace, a heat pump with electric backup strips, or electric baseboard. An hourly model lets the house coast down slowly with the heat off when the setpoint drops, then pays to recover. On a heat pump, a cold morning or a big jump in setpoint brings in the backup strips, which cost about three times as much per unit of heat. You get the weekly, monthly and October–April season cost against holding one flat temperature, a one-week chart of setpoint, indoor and outdoor temperature with hourly cost bars (strip-heat hours in red) that you can scrub, and a monthly cost comparison. The verdict chips are BIG SAVER, SOLID, MODEST, NO REAL SAVINGS, COSTS MORE and EMPTY, plus STRIP HEAT, COLD WAKE and PIPE RISK warnings. You can copy a markdown report stamped in America/Chicago time, and state saves under setback-v1.
How to try
- The demo opens on Classic 9-to-5 setback: a 1,800 sq ft Chicago house in January with a 92% gas furnace, 68 at home and 62 overnight and during the workday. It comes out to about $40 a week against $53 for a flat 68, a BIG SAVER at about 24%.
- Load Minnesota heat pump trap. A deep 60-degree night with a 6 AM jump to 70 sends the backup strips in every cold morning, so the schedule costs about 6% more than leaving it alone: COSTS MORE, STRIP HEAT and COLD WAKE. Soften the morning ramp and watch the red hours disappear.
- Work from home on a Nashville heat pump only saves about 3% (MODEST). Weekend cabin mode saves about 27% and Drafty old house saves about 47%, because a leaky house loses the most heat when it's held warm.
- Tap or hover any cell to see that hour's outdoor temperature, heat load and cost against flat, then switch the month to see how the savings change across the winter.
Why this (vs neighbors)
Heating season is starting, and "should I turn the heat down at night?" has an answer that depends on your equipment. With a furnace a setback nearly always saves money. With a heat pump, a big morning recovery can switch on resistance strips and wipe out the savings. Earlier money builds like Payback Clock and Margin Math are spreadsheets with a chart. Setback lets you paint the schedule directly and see the cost of each hour.
Scope trim
No real weather feed or ZIP lookup, no cooling season, no smart-thermostat import, no humidity or solar gain, and no share URL. One house, one typical week per month.
Limitations
- Outdoor temperatures are approximate monthly normals with a smooth daily swing, not a real week. Cold snaps matter a lot for heat pumps and aren't modeled.
- Heat loss is a single UA number per square foot (0.45 leaky, 0.30 average, 0.18 tight, in BTU/hr per degree), and the house coasts down at most 1.5 degrees an hour. Real houses vary a lot, so the savings percentages run higher than typical published estimates of around 10%. Treat the numbers as a comparison between schedules, not a bill.
- Heat pump efficiency follows a simple curve from about COP 3.5 at 47°F to about 1.8 at 5°F, with strip heat below about 25°F or on recoveries of more than 2°F in an hour. Many modern cold-climate heat pumps do better than that.
Suggested next steps
- Pull an actual hourly forecast for a ZIP code and price the coming week.
- Add an "optimize for me" pass that finds the cheapest schedule which stays above your comfort floor whenever you're home.
- Calibrate UA from a past utility bill and heating degree days.