case in point.

 

I agree that the numbers are biased, because if you buy a hybrid car, and then join the greenhybrid forum, then start recording and entering fuel consumption tank per tank, for sure, you're not the average moderately milage-concerned driver.
But the milage database and the forum helped me a lot when deciding to buy the car. It also gave me some idea of what to expect. I think I said to myself that an average above 6 L/100 km would disappoint me. The database average is around 5.1 L/100 km, and I have an overall average of 5.23 L/100 km. Pretty close. You can also study the distribution, to get a better idea on what to expect. So the database IS really valuable, and I think most people can study it and apply some critical sense to adjust for the bias.
Most drivers have an optimistic view on their car's fuel consumption, but that goes for the normal cars as well. I was surprised today, when driving around town with my old car for a few km, averaged 16 L/100 km. I remember not long ago being disappointed when making the same kind of trips in the hybrid and getting 7-8 L/100 km.

 

In part I agree, but have to laugh a bit too. My "other car" is a 21 year old F150 pickup, V8, 4.11 gears, automatic. I drafted semi trucks, moving vans, kept my speed under 60mph, timed lights with this truck for years, trying to squeeze out 15 mpg driving to work. If I drove the limit or hit lots of stop and go traffic, I would get 12 to 13 mpg

With my TCH, I sometimes drive a little faster, at times crusing at 65. But overall, I drive the same, and enjoy 42 to 46mpg on the exact same route. I really enjoy going 700 miles on a fillup, and only paying $40! With two tanks, my F150 was lucky to get 400 miles, and often paid over $75 to fill up.

So I guess what I'm saying is "large" all depends on the driver and what the "other car" was.

My drive is rather level, 8 traffic lights, and about 5 miles of surface roads, and 30 miles of freeway (I-15 aka the San Diego parking lot) each way.

 

I don't post data from experiments because they don't represent normal operational use. For example, testing maximum power and use of "B" are outside of the range of normal operation. Letting my wife drive the car would also be outside of normal operation.

Bob Wilson

 

I record EVERY tank, regardless.
As Bob points out, it's possible that entering in an "experimental tank" could skew the averages in a bad way. I think what Bob's doing here is fine, and he is thinking about maintaining the integrity of the database. But, being fair about what tanks to enter and what tanks not to enter NOW becomes a potential issue.
I just enter them all, and can add comments about oddities with a particular tank. Not perfect, for sure, but certainly better (I think) than picking and choosing your tanks. I wish there was a simple way to record a tank (like Bob' experimental ones), where he could put an asterisk in a field, so it would be included or excluded, based on what you want to see out of the database (EVERYTHING, or normal tanks).
This, I bet, is what Livvie (and others) are talking about. I don't think any of us intentionally "fudge the numbers" on a given tank, but some may choose to leave out a poor tank here and there, which calls into question the overall validity of the database as representing a cumulative average.

 

Yes. I agree, it's not that people are doing this on purpose. I just think as a hybrid owner you will be biased. And some hybrid owners are extremely biased.

 

Worse, I could skew the results in a way that is misleading. For example, I can hold 15 mph and burn a tank getting better than 100 MPG. This would make me 'look brilliant' but 15 mph is rarely the normal speed of a Prius.

In contrast, I figured out early that route-planning can improve my daily commute. It has become 'old hat' to drive through my neighborhood until the coolant "blue light" goes out. Then I get on a cross-town, higher speed street. But I document this as 'how I drive' and it is part of my daily practice.

In another case, I needed to find out what happens under maximum power going up a hill and later at different speeds descending using "B". This test was designed to find out at what speeds and what grades "B" works for coming down a mountain. But I don't drive mountains in my normal operation. So I decided before the test that the mileage data would be FUBAR and not logged.

In January, I drove from HSV to Georgetown KY via Knoxville and got a taste of the eastern KY mountains. All of those miles are in the database since that was normal operation.

My rule of thumb is show what the car can do in ordinary service and drop anything that is grossly out of the ordinary. For example, leaving the car idling for a week; loaning it to someone who . . . is usually unskilled . . . or some one-time, unanticipated hardware failure . . . say driving on the donut for a week.

Bob Wilson

 

I would say even loaning the car out should be included.

IMPO, anything that burns fuels should be logged (with the exception of strictly experiments). As long as what was done would be in the relm of a "normal" driver, it should be logged. If you use the car as a generator, then no, but if your wife, girlfriend, boyfriend, husband, daughter, etc.. borrows or drives the car, then those should be included miles, and are probably the most important since they are closer to the average driver's results.

 

I'd probably try to get them a separate, GH account, and record their miles under that entry. But this is not something I feel strongly about. I'd rather see some enhancements to the mileage database, say add an oil testing and tire database too. Shoot, even an insurance database . . . things that relate to the cost of ownership.

Bob Wilson

 

That's how I do it no matter who was driving. I think that this gives the most validity to the database that every mile the car had while owned by the user goes into the database. First tanks, bad tanks, tansk with other drivers, and even extremely high/low tanks under non-normal conditions should go in. The Insight Marathon tank was certainly far from normal conditions, but it was included in the user's car's history. If more cars had a non-resetable lifetime MPG gauge I think this issue would be less important, but putting every tank in brings the average closest to what it would be if every car had a LMPG screen.