Ok, I started filling it on Fuelly on 10-10 this year. To date, I have 6 fill ups. Right now, it says I'm averaging 6 fill ups a month.
If I mathed out the total days, I have I think 42 days. Even at that, I should be showing around 4 per month, but real world statistic should currently be 3 per month.
What I don't know, does Fuelly take into account ALL fill ups then divide by the number of days, or months, or what? Does it drop the first fuel up in this particular calculation like some of the other stuff has to do? Cause either way, it appears off. Does it only track the months you actually report, or does it base its tracking on a per month in real time? For example, the 1994 beretta also appears to be incorrect to an extent. I only drive it during the nicer weather months.
Just trying to understand how Fuelly is arriving at 6 a month, when I have yet to do 6 fill ups in a month. lol
ok, keep in mind that for this statistic a month is 30 days. So I'll show you how we calculated this for your car.
Your first fuel-up was on Oct. 10th. Your last fuel-up was on Nov 21st. That's 42 days. You've filled up six times. So we ask: is 42 closer to 1 month or 2 months? It's closer to 1 month, so you're averaging six fuel-ups per rough month (30 days).
Now, you only have six fuel-ups to work with. So as you add more fuel-ups, this method of calculation will make more sense. If you're dealing with years of data, a few days of rounding here and there won't mean very much. But I understand the confusion when you're looking at a limited set like this. You say, "I've filled up three times in October and three times in November, that's three per month." And you're right. But as you take the long view, a month makes more sense as a span of time.
Maybe the difference you're seeing between the dashboard and vehicle stat page is our use of a running average on the dashboard and total average on vehicle pages.
Just took a closer look at the dashboard trend lines and we were including fuel-ups with 0 mpg there. The zero points weren't mapped, but they were throwing things off a bit. The trend lines should look a little better now.
Well, I had more or less figured that already. I was trying to see how you were getting a month out of 42 day
It sounds like to me, you are going by 30 (or 30. whatever) increments, ignoring a decimal equivalent equation completely. For example:
If I divide 365.25 by 12, I get 30.4375 for a monthly average. Yes, I used the .25 to cover the leap year for simplicity of the math.
Anyway, using that math may not be AS accurate as going by actual calendar month days, but its more accurate than your current 30 day rounding and still is simple. Take the next step:
42 days divided by 30.4375 = approx 1.38 months.
Take your 6 fill ups, divide by 1.38 months, you get just over 4 fill ups a month as it is now. This may still not present correctly to some users with low number of fill ups, but it is more accurate and at least mathematically it makes more sense.
To me its a more realistic value than doing a monthly round off every 30 days. Just cause you have in many cases years of data is immaterial. This would still be more accurate than the way it is now.
Thanks for fixing the graph issues, again! lol Not sure why I keep running into discrepancies. As for being a running vs overall, the graph should still be reflecting the same information. Simple as that.
Let me know if I made any sense with the rest of it.
From first fill up date to last fill up date: 721.5 days.
Divide that by 30.4375 gives you:
Now, divide that into the total number of fill ups:
142 / 23.7 = 5.992 or 6 fill ups a month. Which is what is being listed now.
As for the other vehicle, 1994:
First to last fill up - 705.5 days.
40 / 23.18 = 1.73 or about 2 fill ups a month.
Yes, both of these would match the current averages listed already. But you see that this math does work. Just trying to present a different way of calculating the data for a more accurate representation.