This site does not make much sense when it says to enter in the odometer reading each time you fill up.. If you enter in the odo reading with each fill up it calculates the mpg from those miles with the current amount of gas you got. That's pointless since you haven't driven anywhere on the gas you just got.. and yes tried using the "partial fill-up" and that didn't correct it. You shouldn't have to wait until you fill up again to input the miles you actually drove on each tank. It should be calculated back, don't get why a site like this wouldn't do that for you
What looks way off to me is the data you have entered. On 5/8 you drove 49 miles but used 14.62 gallons and then on 6/5 drove 263 miles using only 9.39 gallons? IT looks to me like you don't understand what data you should be entering.
I made a spreadsheet many years ago to check my MPG. By tracking my odometer reading + gas filled, it would calculate miles driven (difference in odo reading) then divided by gas I used (my full tank = 13.2 - gas just purchased) and gives my MPG. No fancy algorithms or mystical derivatives here, just simple math.
Bump into Fuelly not long ago and I have to admit, it is better looking than my spreadsheet. However, I continue to use both Fuelly and my spreadsheet as it also keeps up with maintenance and repairs. Yes, I have same MPG from either Fuelly or my spreadsheet.
** When fill up, be sure to use nozzle's preset low/med/high (I use same setting each time I fill up). When nozzle stops, those are the #s (gallon/price/odo reading/date) you need for Fuelly.
** As I understand it, partial fill up is when you stop the nozzle manually before nozzle stops by itself. Such as one only purchased $3 in gas. That results in a partially filled tank instead of a full tank of gas, which will throw off MPG calculation. So, if you want most accurate MPG, please get a full tank of gas each time you fill up.
I'd define partial fill as different from your normal, 'total fill' method. For example, I continue playing with the nozzle when it clicks off to fill the tank as close to the brim as I can.
Some people never fill the tank, though (they feel less painful to pay the same amount in smaller pieces), so the only way for them to get an approximate mileage to average divide the distance and fuel quantity of a lot of these partial fills. But it's harder to assess the effect of driving techniques/weather/whatnot that way.
You can also fill smaller amounts without partial fills by always keeping the tank close to full. That is (of course) some extra weight, and smaller amounts will always give worse accuracy than letting it go near empty - but they can still help when you experiment with new mods or techniques.
Anomalies can still occur. Me and my girlfriend both have near-100 mpg record tanks from shorter fills, while our record full tanks are ~90-93 mpg(US). I can only blame short fills' worse accuracy.