Google Analytics - Adding More Than 25 New Accounts

I have more than 25 Google Analytics accounts for which I'm an administrator, built up over many years of working with various clients. I have others for which I'm merely a User, rather than an administrator. I wanted to set up a new account today, for the first time in some months. Google won't let me. It says that I must "reach out to my Google Analytics support contact".

I don't have a Google Analytics support contact. I've never knowingly had an email or phone call or letter from Google Analytics telling me who my support contact is or how to contact the team. I don't interact with the Google Analytics team other than infrequently through the Google Analytics Help Forum and the Analytics API Help Forum (where I'm a lurker rather than a contributor).

So, how does one go about adding a new account, when Google prevents you, and gives you a completely useless diagnostic error message. Why is it completely useless? Because not only do I not have a Google Analytics support contact, the Google Analytics Help system doesn't cover this problem and the Help forum is a user to user help forum. It really isn't appropriate to go broadcasting an individual users problems' with the service. But there is no private route visible. So you either have to announce your incompetence in public, or forget about resolving the issue. Neither of these is appropriate for a well designed customer service interface with the business.

Weight Loss, Weighing Scales, Accuracy and Precision

As part of a now six month old health improvement program, I need to lose some weight and keep it off. In the process, I've found several problems connected to measurement, and the psychology of reinforcement - forming effective habits. These are some notes about the technology and psychology of taking measurements.

Measurement Problems

When I step on a weighing machine it gives me a value. Sometimes these machines give very precise values for my weight. The one I have at home now, will give me my weight to tenths of a pound. So you'd think that I could easily tell whether I'm gaining or losing weight, wouldn't you?

Weight before urination

It turns out to be harder than I thought to tell whether I'm losing weight or gaining weight.

How Much Do I Weigh? And Have I lost Any Weight?

Having taken my weight, I can then compare this reading with previous weight readings to see whether I've gone up or down in weight. So surely I can tell whether I'm putting on, or losing weight?

If I've gone down a lot in weight, I can have a celebratory helping of something yummy. I feel happy. My dieting-buddies will congratulate me and all is well with the world.
If I've gone up in weight, I give up on the diet and tell my friends it wasn't working for me.

But... what if the weighing machine is wrong?

Testing The Weighing Machine

Here's a simple test.

1. Strip and take your weight to get the first reading.
2. Drink quite a lot of water - a couple of pints or so - and take a second reading.
3. Dress and go do something - other than eating, drinking or using the loo. Wait until you are busting. Strip and weigh yourself naked to get the third reading.
4. Now use the loo... and weigh yourself again for the fourth reading.

I've done this with some domestic weighing machines and the results can be surprising. Sometimes my weight apparently goes down after drinking pints of water. After what should be a rapid weight loss from using the loo, my weight sometimes goes up, even on very precise weighing machines. And that can't be right. There's no (scientifically plausible) way you can lose a couple of pints of water (2lbs to 2.5lbs - US and UK pints are a different weight) and increase your weight. You can't drink a couple of pints of water and stay the same weight.

Weight after urination - lose water and gain weight? That can't be right!

The problem is that domestic weighing machines can be more precise than they are accurate. You get fooled by seeing a precise number and assume that the accuracy is also high... Not always true.

Precision

The pictured weighing machine might tell me that I'm 12 stone 1.4 pounds (169.4 pounds or 76.83 kilos) - it's giving me a weight that I can use, right down to ounces. It's quite precise. I've used lab weighing machines that will weigh humans at the level of grams (a fraction of an ounce) - they're really precise.

Accuracy

Before changing weight (drinking water or after urination) I might weigh 12 stone 1.4 pounds. After changing weight I might weigh 12 stone 3.1 pounds. An apparent weight increase of 1.7 pounds. But I can put on weight when I've just urinated, and lose weight when I've drunk a few pounds of water. In the pictures above, I've taken my weight before and after urination and increased weight from 12 stone 1.7 pounds up to 12 stone 2.6 pounds - a weight gain of 0.9 pounds after an obvious mass decrease.

The home machine that I used for these numbers, has a high precision, but a low accuracy.

It's a digital machine, and it seems to have some sort of short period memory. So stepping off it, and then on again, shows me the same weight. Stepping off it, waiting a bit and then stepping on it again - with no weight gain or loss activities other than breathing - and my weight may increase or decrease by pounds. These are signs of a weighing machine with low accuracy. Just watch out for that short period memory - it may fool you into thinking that the machine is accurate, when it isn't.

A machine with low accuracy might show me as putting weight on, when I have genuinely lost weight. Or conversely show that I've lost weight when I've really gained weight,

There's one more significant problem that my machine can give me, when trying to diet effectively.

Calibration.

My machine is giving me a precise, yet inaccurate weight. It could also be wrong in another way. If my machine says that I'm 12 stone 1.4 pounds, then a calibrated machine might show my weight differently. If I go to my doctor's office and get my weight taken there, then I come home and weigh myself at home, I might see quite a difference. That's because my home machine hasn't been calibrated - adjusted so that a reference weight shows the same value on the machine at home and at the doctors.

If your scales are badly calibrated, you might think your weight is in the target range, when you've really over- or under- shot the target. You can only really calibrate by comparing different weighing machines, ideally using a standard defined weight. That's what is done to the weighing scales at your doctors, and at the supermarket.

These problems - calibration, precision and accuracy - affect most measurements. Real scientists doing lab experiments spend a lot of time, and money, to make sure that their measurements reflect what is actually happening, to improve accuracy, and precision and to make sure that their calibrated results can be compared with other measurements taken in other laboratories at different times.

Misunderstand the measurements on your weighing machine, and you may be falsely encouraged or discouraged on a diet. And that's before we look at a whole other class of measurement errors - systemic errors.

Quests and Questions

One of my quests is to find a domestic weighing machine at a reasonable price, that is as accurate as it is precise. Ideally down to the tenths of a pound. And preferably has some calibration mechanism. Suggestions?

Another - find an accurate and reasonably priced domestic weighing machine, ideally with a WiFi, BlueTooth or USB connector that I can read from a Linux or Apple computer. Without having to disassemble the weighing machine and rework it significantly, or pay a lot. Suggestions?