World 3-1
Accuracy… good
Bullseyes are hard to come by
In the story of William Tell, he aims a bow and arrow at an apple on his willing son’s head to split it (the apple, that is).
An inspiring father-and-son story of trust. Trust ‘games’ aren’t as exciting nowadays.
Most times in life, accuracy doesn’t need to be as good as that.
Close Enough
In real life, graphs don’t look neat like in maths class.
That’s why statistics uses something called a line of best fit. Here’s a rough example to illustrate what I mean:
Here, the meal = 2 bananas.
Each yellow dot is a record of how much my BGL changed a few hours after eating them bananas.
If I inject little insulin, my BGL rises (top-left dot).
If I inject much insulin my BGL falls (bottom-right dot).
After I test a few times, I may see a line that can be drawn, perhaps roughly, between the dots. This gives me an idea of what to expect.
Of course, this is overly simplified; my only point is that graphs are better than tables (for me).
Having a (Tennis) Ball
Another approximation I make is for fractions.
I could use fractions like 13/27, 7/22 etc. – but why overcomplicate? I’m not trying to get to the moon. 1/2 and 1/3 will do me just fine.
For example, say I’ve worked out that playing a morning game of tennis means I inject less insulin than usual for breakfast. So I do a few tests and come up with an exact number: 0.55
My tests give rough averages anyway, so I use a simple fraction instead: 1/2.
2 approximating principles I use are to:
- go as simple as possible, but no simpler.
- stay on the safe side of guesses.
Aiming for Good
I don’t concern myself with perfect accuracy. There are too many variables every day for me to get it perfectly right anyway.
BUT I still try to get good accuracy. I work with my limitations.