The hot tub that feels like a stage

Do not let one good July afternoon approve a permanent object.

This is the annual vote. The same proposed screen, awning, pergola, move or planting change is held fixed while the engine re-runs the property in winter, spring, summer and autumn.

The hour is intentionally held constant. This isolates seasonal geometry instead of letting a different clock time flatter the proposal.

The year, in one line

Everything below is about a change. Before that, it is worth knowing what the place already does — the hours it is open, month by month, on one page you can put on a fridge.

Before a permanent change gets to win

Make January, April, July and October vote on it.

These are the exact same proposed objects re-run through the engine in four seasons. A summer fix that steals winter sun cannot hide inside one July score.

Same geometry · four seasons
ChangeJanAprJulOctAverageYear verdict
Two 8.2 ft parasols instead of one
The same built change does not produce a dependable improvement across the year.
+56
4298
+0
100100
+4
9498
+0
100100
+15weak
Two low shielded lights instead of one high one
The same built change does not produce a dependable improvement across the year.
+12
4254
+0
100100
+0
9494
+0
100100
+3weak
A 9.8 ft umbrella
The same built change does not produce a dependable improvement across the year.
-1
4241
+0
100100
+3
9497
+0
100100
+1weak
Move the table 6.6 ft east
The same built change does not produce a dependable improvement across the year.
+0
4242
+0
100100
+0
9494
+0
100100
+0weak
Change The floodlight to a warmer lamp
The same built change does not produce a dependable improvement across the year.
+0
4242
+0
100100
+0
9494
+0
100100
+0weak
Shield The floodlight
The same built change does not produce a dependable improvement across the year.
+0
4242
+0
100100
+0
9494
+0
100100
+0weak
What changes by season
Two 8.2 ft parasols instead of one

January: no 42works 98; after the change the controlling issue is what it feels like; direct-sun reading +0.

April: works 100works 100; no controlling failure remains; direct-sun reading +0.

July: works 94works 98; after the change the controlling issue is what it feels like; direct-sun reading -94.

October: works 100works 100; no controlling failure remains; direct-sun reading +0.

Two low shielded lights instead of one high one

January: no 42no 54; after the change the controlling issue is light after dark; direct-sun reading +0.

April: works 100works 100; no controlling failure remains; direct-sun reading +0.

July: works 94works 94; after the change the controlling issue is what it feels like; direct-sun reading +0.

October: works 100works 100; no controlling failure remains; direct-sun reading +0.

A 9.8 ft umbrella

January: no 42no 41; after the change the controlling issue is light after dark; direct-sun reading +0.

April: works 100works 100; no controlling failure remains; direct-sun reading -9.

July: works 94works 97; after the change the controlling issue is what it feels like; direct-sun reading -93.

October: works 100works 100; no controlling failure remains; direct-sun reading -9.

Move the table 6.6 ft east

January: no 42no 42; after the change the controlling issue is light after dark; direct-sun reading +0.

April: works 100works 100; no controlling failure remains; direct-sun reading +0.

July: works 94works 94; after the change the controlling issue is what it feels like; direct-sun reading +0.

October: works 100works 100; no controlling failure remains; direct-sun reading +0.

Change The floodlight to a warmer lamp

January: no 42no 42; after the change the controlling issue is light after dark; direct-sun reading +0.

April: works 100works 100; no controlling failure remains; direct-sun reading +0.

July: works 94works 94; after the change the controlling issue is what it feels like; direct-sun reading +0.

October: works 100works 100; no controlling failure remains; direct-sun reading +0.

Shield The floodlight

January: no 42no 42; after the change the controlling issue is light after dark; direct-sun reading +0.

April: works 100works 100; no controlling failure remains; direct-sun reading +0.

July: works 94works 94; after the change the controlling issue is what it feels like; direct-sun reading +0.

October: works 100works 100; no controlling failure remains; direct-sun reading +0.

Annual shortlist

6 exact proposals re-tested

Two 8.2 ft parasols instead of one

weak

The same built change does not produce a dependable improvement across the year.

Best seasonal delta +56; worst +0; four-season average +15.

Two low shielded lights instead of one high one

weak

The same built change does not produce a dependable improvement across the year.

Best seasonal delta +12; worst +0; four-season average +3.

A 9.8 ft umbrella

weak

The same built change does not produce a dependable improvement across the year.

Best seasonal delta +3; worst -1; four-season average +1.

Move the table 6.6 ft east

weak

The same built change does not produce a dependable improvement across the year.

Best seasonal delta +0; worst +0; four-season average +0.

Change The floodlight to a warmer lamp

weak

The same built change does not produce a dependable improvement across the year.

Best seasonal delta +0; worst +0; four-season average +0.

Shield The floodlight

weak

The same built change does not produce a dependable improvement across the year.

Best seasonal delta +0; worst +0; four-season average +0.