Read the spread, not the mean
The twenty readings in the form average 1.225 in, which sits comfortably under a 1.25 in target and tells you nothing useful. The same twenty run from 0.94 in to 1.50 in — a spread of nine sixteenths — and ten of them are under target. A sheet like that is thin in the places that get the most blade and thick in the places that get the least, which is what every real sheet does.
The two ways to fix it
Add up the shortfall at each low reading and spread it over the sheet and it comes to 0.075 in of missing ice, which is 103 cubic feet, which is 749 gallons of water once the density conversion and a six per cent loss are in. That is under six resurfacer floods and it assumes you can put the water where it is wanted.
Flood the whole sheet deep enough to lift the thinnest reading instead and you need 0.31 in everywhere: 3,077 gallons, twenty-four floods, and the thick end finishes at 1.81 in. It takes more water every time, because it lays the full depth over ice that did not need it. What it buys is that it can be done with the machine instead of by hand at midnight.
Where to drill
A grid you can repeat matters more than a dense one. Mark the hole positions on a plan, use the same ones every month, and the numbers become a trend instead of a snapshot. Add holes where you know the ice suffers — the corners, the creases, the face-off dots, the turn the resurfacer makes at the gate — because those are the readings that will move first.
What this cannot know
It cannot tell you what target to use, and it will not say whether a reading is acceptable. It also assumes each reading covers an equal share of the sheet, which is only true if your grid is even; a survey with six holes clustered at one end and two at the other will produce an average that is honest arithmetic and a poor description of the ice.
Questions people ask
How many holes should a survey have?
Enough that a thin patch cannot hide between two of them, and always the same ones so the readings compare month to month. The page tells you how much sheet each reading is standing in for, which is the figure to look at when you are deciding whether to add holes.
Can I enter millimetres?
Yes. Set the selector at the top and enter both the readings and the target in millimetres. Everything is converted and displayed in inches, at three decimal places with a sixteenths reading beside it, so a thirty-second of an inch does not vanish in the rounding.
Why is the flat flood always more water?
Because it lays its full depth everywhere, including over the parts of the sheet that were already at or above target. Filling only the shortfalls puts water in the low spots and nowhere else. The arithmetic makes the flat flood at least as much water in every case, and usually a great deal more.
Does the page tell me if my ice is too thin?
No. It reports what your readings say against the target you typed. Whether that is acceptable is a matter for whoever governs play on your sheet and for the people responsible for the building, and this page has no standard of its own to hold you to.
What causes the spread in the first place?
The blade takes more where the machine passes most and where it turns; edges and creases get scraped and dug; floods run downhill on a floor that is never perfectly flat; and a sheet that has been cut and reflooded for months carries the history of all of it. That history is what a survey is reading.