Fuelly - Again

[asb]

I've searched around trying to find an answer to this, but often the answers are hidden in other threads.  So, sorry if this is a repeat.

 

When entering information into Fuelly, which number should I use for the gallons used?  What's on the gas receipt or what is showing on the trip screen?  

Thanks.

[RaPieR]

I've searched around trying to find an answer to this, but often the answers are hidden in other threads.  So, sorry if this is a repeat.

 

When entering information into Fuelly, which number should I use for the gallons used?  What's on the gas receipt or what is showing on the trip screen?  

Thanks.

I'm not sure what the consensus is, but I use the actual gallons used based on what I paid at the pump.

[ScubaDadMiami]

I go with what the C-MAX states, not the pump.

[salsaguy]

actual gallons used shown on pump.

the dash is known to be inflated by as much as 2 mpg.

[asb]

I've been using what's listed at the pump, and my numbers are always less than expected.  This last fill-up looked like this:

 

From the pump - mpg average for the tank = 44.7

From Trip 2 - mpg average for the tank = 45.7

 

That's a huge difference.

[plus 3 golfer]

So, apparently the fuelly numbers are inflated by those that use the dash gallons. :)  There's more than one on this site that have indicated they use the dash gallons in fuelly. My Lifetime Summary is 41.4 mpg but my FE based on pump readings is 40.3 mpg.  That's about a 2.7% difference. 

 

Also, I believe ptjones said he adjusts his fuelly numbers up for his odometer error of IIRC of about 1.5% (his actual distance traveled is greater than the odometer indicates).  But as tires wear such odometer error will decrease and one should then continually adjust as tires wear and then replaced.  Over the life of the tire such error should decrease by about 1%.  My odometer when I checked after I first bought the car was indicating an error of about 2% but I do run 44 psi which could increase the diameter slightly and contribute to my odometer reading being less than actual distance traveled.  

 

So, if I adjusted for my odometer error when my tires were new and used dash gallons, my lifetime FE number would be 42.2 mpg.  Bottom line: report whatever you feel comfortable with to fuelly.  I feel fine reporting data that gives me 40.3 mpg as my overall FE.

[HannahWCU]

Pump reading.

[darrelld]

I use the pump reading, its the only accurate way to report. As Plus3 stated there are too many variables with the dash readout.

[asb]

Pump, it is!  

[ScubaDadMiami]

Speaking of Fuelly, why is my signature not updating past fill number four?  I see new stats on Fuelly, and I even re-did my signature here, just in case.  :headscratch:

[Andy from Fuelly]

ScubaDad, just a quick heads up that we are looking into this issue about your signature image.  Hopefully we will have this resolved soon, I don't think any update will be needed on your part.  What you have in place should work.  Sorry for the inconvenience.

[Andy from Fuelly]

ScubaDad, how did you add your last fill up?  Was it using the GasTracker+ app?

[Jus-A-CMax]

ScubaDad, how did you add your last fill up?  Was it using the GasTracker+ app?

 

Andy, thanks for signing up and the assisting with our members, its much appreciated by all. And welcome :)

 

PS Since you've mentioned shill hunting, have a look at "P.O.S"   the data input does not appear credible especially tanks 4 & 8

[salsaguy]

plus, the pump is 2 digits whereas the dash is only one digit after the decimal right?

 

I use the pump reading, its the only accurate way to report. As Plus3 stated there are too many variables with the dash readout.

[salsaguy]

that was one of the very first ones i flagged as bogus even i was analyzing the data, so i removed it.

their name and numbers are definitely suspect.

 

 
Andy, thanks for signing up and the assisting with our members, its much appreciated by all. And welcome :)
 
PS Since you've mentioned shill hunting, have a look at "P.O.S"   the data input does not appear credible especially tanks 4 & 8

[Jus-A-CMax]

that was one of the very first ones i flagged as bogus even i was analyzing the data, so i removed it.

their name and numbers are definitely suspect.

IP check and its a good chance its a Prii owner.... ;)

[Andy from Fuelly]

ScubaDad, it appears your sig image not updating was a one off issue.  We were able to manually request the image be rebuilt and it's showing the proper number of fuel-ups now (5).  In the future, if any of you all have this problem, please follow the instructions here which basically tell you to click edit and then save a fuel-up to tell the system to rebuild your sig image.

 

As for the P.O.S. vehicle, I did some preliminary research and could not find anything blatantly obvious that it's spam so there is nothing I can do at this time.  Our current system only allows me to ignore an entire vehicle, not individual entries which look suspicious.  Thanks for pointing out that entry though, it's helpful to review and take notes.  We'll find ways to promote the value of trusted users and demote the value of potentially spammy contributions but that won't be for a little while.  For example, if someone signs up, loads in a bunch of back dated data on their first visit and never shows up again, that follow the normal/natural patters for fuel logging.  I'm not going to get into details about our plans on how to identify unnatural behavior because I don't want anyone to research my posts and find a way to game the system.  Just rest assured it's on our To Do list and we'll find some clever ways to identify spammy contributions.

 

The POS example is a great way to also showcase why we should allow users to view data in various ways.  For example, If you have a new vehicle and we have a small set of cars in the database (for this example lets say there are just two vehicles in the database).  If one of them has 4 fuel-ups and is averaging 20 MPG but another has 16 fuel-ups and is averaging 40 mpg the current setup would simply take these two and divide by two giving 30 mpg.  Another way to analyze this data would be to take all 20 fuel-ups and average those which would give the car a 36 MPG.  Now there are issues to this latter formula as well as it would weigh the vehicles with more fuel-ups with more weight then a vehicle with less fuel-ups and in many cases a person who logs for a ongoing period of time might be a more conscious driver then an average Joe and thus their great MPG would skew the stats.

 

My guess is that we'll come up with an algorithm that will identify statistically significant outliers and remove them from the formula.  From there well do some math on fuel-ups from all vehicles combined along with the math for each vehicle and their respective MPG and then blend these into our official estimate.  We would provide distribution graphs, scatter plots and other visual/raw data values so it's easy to see more then one way to look at the data.

[ScubaDadMiami]

Just as an FYI, I added data from a mobile phone (Android), which I believe is how I also loaded at least some of the prior fills.

[Andy from Fuelly]

Thanks for the info Dan.  We actually tested, web, mobile, apps and SMS last night they were all working.  So this was just some kind of fluke issue.  If this ever happens again, just edit a feul-up and click save (with the original data) and that should trigger the image update.

[fotomoto]

Welcome Andy!

 

I just noticed that one can now add "mini-stats" and there are many selections; I added gallons per 100 miles.  And now when you view other members pages, their values for those same mini-stats will also display on their pages so you can compare.  Nice.

[salsaguy]

Andy

Id like % hwy/city to be understood by more folks since most dont or just guess at it when entering at fuelly

if this table was shown somewhere to the right when entering your % hwy/city info for ref it might help others get closer to the real #

 

see here:

http://www.fuelly.com/tip/555/Find-your-true-cityhighway-driving-splits

 

calculations are:

Percentage of Highway Driving = 100*(Your AVG SPD for the tank - 21)/(48 - 21)

 

Percentage of City Driving = 100-(Calculated HWY%)   note see value obtained above for % Hwy

 

 

I created a table to show your %city/hwy for each change in your avg mph over your tank

if you have a trip computer on your car that shows your avg mph driven over the tank full,  you can use this to your advantage as i do on my 2012 Passat. When i fill up, I check my avg mph for the tank and then know if I drove more city or hwy miles.

 

See attached pix/table

     

here is the chart broken down at every 1 mpg avg change instead:

 

hope this helps

Edited August 31, 2013 by salsaguy

[DaveofDurham]

Andy -

 

Thank you for Fuelly and your attention to the C-Max data.

 

I believe unconscious driving is severely frowned upon and results in really poor mpg or worse.

 

 

 

. . . Now there are issues to this latter formula as well as it would weigh the vehicles with more fuel-ups with more weight then a vehicle with less fuel-ups and in many cases a person who logs for a ongoing period of time might be a more conscious driver then an average Joe and thus their great MPG would skew the stats. . .