I’m a little confused about what the “All Sense Homes” number represents.
On my page it shows that to be 1,952 watts over the last 30 days. It also shows my usage at 1,852, 100 watts lower. It also shows I use more than 56% of all Sense users.
Is the 1,952 an average of all?
Here’s my guess. Your 30 day usage was 1852W. That places you higher than 56% of Sense users.
Since 1952W is bigger than your usage, it must be the mean/average, not the median (since your usage is above the median).
[edited based on @samwooly1’s correction]
1 Like
Sam… Here is what I have… Don’t know what to say…Just info…Gerry
Yours makes sense to me. Where the number shown for your home is lower and the comparison to other homes, your home is also lower.
Mine looks strange I think because I’m looking at incorrectly as @kevin1 pointed out.
They have made this a bit too complicated and should be east for anyone looking at it to make sense of. Maybe if they provided more detail or explanation then it would help.
@samwooly1, I think yours looks strange because Sense is using two different metrics. When you see a bar chart like that that breaks values out into multiple ranges, those ranges, or quantiles are based on the breaking statistical distribution of values from all users (or all users in Kentucky) into buckets (in this case 3 buckets) that have the same number of users:
The % with respect to other Sense users is also based on the distribution. If the distribution of users energy usage was nice and symmetrical like the wikipedia example above, then you wouldn’t see the weird results, because your position with respect to all the other users would correspond with where you are with respect to the average/mean. This is because the mean/average for a symmetrical distribution is exactly the same as the median, the user usage where 50% of the other users are on each side of it.
But I’m betting the user usage distribution is fairly asymmetrical, which means that the median , is skewed from the average, like the diagram below.
In this distribution example the you would be bigger than 50% of the users, but still be below the average/mean.
Sense probably should be using the median for “all users” number, or show both clearly labeled.
1 Like
I think they should give explanation of the results and also use the same metric across the board