Fuelly computing the wrong MPG... and it's intentional...?
 

I normally fuel up with 91 octane fuel. The other day I bought a tank of 93 octane fuel for the first time (because the station didn't have 91).

When I looked at my app, I was surprised to see that it had already computed my fuel efficiency for 93 octane despite the fact that I had not yet driven a tank of 93 octane through the car.

i.e. The app is wrong. The mathematically correct behaviour would be to associate the MPG computed during the fill-up with the octane of the previous tank.

I put in an issue ticket, and support told me that there "are advantages and disadvantages to both methods of which tank to assign the fuel efficiency to. Years ago, Fuelly decided to assign it to the current tank."

The disadvantage is obvious: the graph is unreliable. (attached) Anyone who wants to know the correlation of MPG with octane is being misled.

I am struggling to see the advantage of their choice. Can someone help me understand?

Fuelly takes no regard of the octane of the fuel used in calculating the mpg. The octane is only recorded for your benefit. I regularly chop and change between octanes and seeing what octane I used allows me to decide if "richer" fuel gives me any monetary advantage.

@JockoT

If you "regularly chop and change between octanes" to see if what octane you used affects mileage per gallon, and you're using the chart in the app... then you're looking at a flawed analysis.

When I grabbed the bar chart above, my car had not yet burned any 93 octane fuel. It is provably impossible for Fuelly (or anything else) to have the data to accurately generate that bar chart.

==========================

Thought experiment:

If, with a brand new car, you were to consistently alternate between 87 and 91 octane on each fuel-up to see how the octane rating correlates with fuel efficiency, and then used the bar chart in the app to view the results 1 year later...

Then you would find the two bars were completely reversed!

I hear you, what you're saying is, when you fuel up on fuelly, you're inputting the octane and price of the fuel you just put in, but the volume is essentially the fuel you just used.

If I fill up with 97 RON I don't know what mpg I get until the next time I top up. The mpg calculated is always the fuel used the top up before. As Draigflag says, only the volume of fuel is used in the calculation.

What I'm saying is that it's attributed to the wrong octane. It's great that the volume is used and therefore the correct division is computed, but the correlation with octane is completely wrong.

In the bar chart on the first thread post, Fuelly shows a computed MPG for 93 octane fuel even though no 93 octane fuel had ever been driven through the engine.

Octane <==> MPG correlations are not computed correctly.

No.

The issue is not one of arithmetic. It's that the attribution to octane rating is wrong.

I realize now that I put an inaccurate title on the thread. The MPG computation is correct. The MPG/Octane correlation is wrong.

My bad on the poor thread naming.

That doesn't change the existence of the problem.

I see what you mean. I use Fuelly on a laptop and it doesn't produce the bar chart you refer to. The second bar should be blue. It is the next one which should be red. have you made an error in Settings or some such?

I don't think that it's an error in the settings because (as quoted in my first post) support's email back to me says that there "are advantages and disadvantages to both methods of which tank to assign the fuel efficiency to. Years ago, Fuelly decided to assign it to the current tank."

So... Fuelly claims to have intentionally decided to mis-attributed the octane/MPG correlation.

Hence my confused post. I was hoping someone could share with me the "advantage" that they claim drives this intentional decision.

It seems to me the advantage of doing it the way it's being done now is that the computation is instant and simple to track without having to refer to the previous entry.

Of course, I use only RUG. If I were interested in keeping up with MPG differences as related to fuel grades, I'd put a notation in the area that allows such, and there I would note the added information regarding fuel grade and MPG differentials for each fill-up using the previous tank for the present miles traveled.

I'm not aware that Fuelly has data showing the differential fuel grade MPG results. All I see is the MPG results without such differentiation. For Fuelly to do otherwise would be much more cumbersome with no benefit except for those few who might be experimenting with fuel grades. I doubt Fuelly is a good vehicle for doing that for the general user. And besides, as noted, anyone who wants such a study can easily do it himself/herself/itself.

The same argument can be used for other influences on MPGs that folks might track, such as tire pressures, etc.