Crackpot theory based on the CR Taxi Test

Not open for further replies.
Apr 21, 2003
Happy Thanksgiving;
Wife is working on Turkey so I'm here talking to you.

Recently re-read the "Taxi Experiment" that Consumer Reports did in the 90's. I know, I know it was bad science (no cold starts etc), but I think there are some reasonable conclusions that can be drawn from the fact that their teams were not able to MEASURE significant wear in either the dino or synthetic cars using the 6000 mile interval over 60K miles (granted, the UOAs were probably horrific by our standards):

1) "Stop and Go" driving is not as big a deal as the oil companies would have us believe.

2) The number "3000 miles" - indeed most people's use of mileage as a yardstick - is inadequate when it comes to oil changes. It has no bearing on how many times the car is started cold, or driven like 2 miles and shut off.

3) The carmakers' 7500 mile OCI for "normal" service is not as crazy as people might think. Likewise, people who drive in reasonable traffic are not necessarily in "severe service" as long as they get the oil good and hot. This Quicklube mantra that "Everyone drives under severe conditions" because there are some stoplights on the way to work is BS.

While I'm on that, who ever said that "normal" service automatically meant perfect conditions on a glass-smooth highway, starting the car twice a day and driving 100 miles? The results of this Taxi test suggest that this is more quicklube propaganda. I'm thinking the short trips in cold weather that skank up the oil are the real villian.

This leads me to believe that the best way to set an OCI is to have it more constrained by time than X miles or X months. That way you can get your money's worth out of the oil when you are driving long trips but dump it at a prudent time when you are doing a lot of short trips. I bet most people could set an upper limit of the manufacturer's max (7500 or 10K miles) and a relatively shorter time (like 4 or 6 months), and be perfectly happy with the results even with Dino oil. Or use top of the line PAO synthetic and drive it for the 25K miles / 1 year that it ought to be good for.

What do you think?

Originally posted by Matt89:
Happy Thanksgiving;
Wife is working on Turkey so I'm here talking to you.

Mine I let go to a party, and I'm HOME ALONE - hehe hehe hehe, anyway....

I would not agree with NUMBER 1.
Stop and GO is very important factor.
It does tear the crud-ola out of the oil.

as far as 2; I think for the average Joe it is a good measure... now remember average is in the middle... sure there are some that can go longer, but as a general rule, this is a level that one can change their oil and the oil is still good (way good) and this is what I consider I would want to do If I was changing BLIND.

Number three has some merit, and needs to be explored... I agree that WITH oil testing yes 7-10K miles are a doable deal... On most oils?? nope, on a lot of oil? maybe, On the best oils, SURE-in general.
That's what I do. 6 months for most cars. Either dino or synth depending on miles driven and the value of the car. It is easier to keep track of and fall/spring changes allows a different viscosity for winter and summer. I have to change 3 times a year because I drive over 20,000 miles. I see people changing in the middle of winter etc. I say use an oil that will get you through winter, preferably every 6 months..up to 6000-8000 miles, except LL oil, I say go up to 10k.
I think you are on the right track. For instance my 02 silverado has the "change oil" light system that starts with an upper limit which appears to be around 7500 miles for the 4.3 and then subracts for cold running time, hot running time, starts, etc. I have a long commute and the light usually comes on in the 6k to 7k mile range.
I've mentioned my engine timer or fuel consumed theory. IMO, mileage is overrated.

Get a timer.
Track exact fuel used.
With a couple of UOAs, you could prove that mileage is overrated.
While the test may have been of questionable value I think it was commendable in terms of its scope. It was a considerable undertaking and I don't know of any other test that compares in terms of trying to get some answers. Their previous test in 1987(?) used lab tests and also did some filter work. The notable thing about the 87 test was the variation found in quality control - i.e., wrong vis, high pour points, etc. The worst thing about that test was if an oil did not meet specs, they tested another sample, but if it met specs, they didn't retest. How many of the oils that met specs would not have if they tested another sample? Anyway, I am impressed with what they attempted to do. I also wonder how much they spent doing it.

On one of their tests they rebuilt a bunch of identical chevy engines. Think they said 4.3s but they were in cabs? Maybe the 4.3 V8, yes, GM made an eight-lunger of that displacement for a few years.

Anyhow I wonder if they paid extra for blueprinting to get all the engines exactly alike. They measured bearing and cam clearances before and after, and that's about it IIRC.

CR's methodology is usually to get one sample of a product through regular retail channels. If it performs below manufacurer spec, too bad; they should have had better quality control. Funny how they retested flunking oils. It makes an argument that they use inconsistent science.
Not open for further replies.