Topic: '05 Accord V-6 Vs '09 G-8 V6

[mkaresh]

The products of the new GM (CTS, Outlook, Vue, Cobalt, Aura) have generally had much worse than average reliability in the first year, going by both CR and True Delta.

I have seen the True Delta numbers and the above quote is indeed correct. I know because I participate in True Delta. The new GM cars are not reliable no matter what GM faithful and apologists might want to spin.

For example, the 2008 CTS has a problem rate of 97 per 100 cars. For a 2005 it goes to 141.

In comparison, a 2008 Lexus ES350 has a rate of 27 problems per 100 cars.

The 2009 Corolla, a new model comes it at 31. A 2009 Cobalt (a five year old model) comes in at 55.

The 2008 Fit rates at 13 problems per 100 cars.

I could go on but there is a reason the General is going bust.

Two things:

1. The sample size for the vast majority of True Delta ratings is statistically too small to give any sort of accurate information.

2. Notwithstanding point 1, the rating for the G8 (47) puts it between the Hyundai Genesis (43) and the Infiniti G35/G37 (50). Not bad territory.

Point one goes a bit far. With the smaller sample sizes, the results are less precise, but are still accurate enough to be useful when comparing cars with widely differing repair rates. I wouldn't assume that a car with 40 repair trips per year is more reliable than one with a reported repair rate of 50, if the sample sizes are near the minimum. But 40 vs. 80? Then the former is very likely more reliable.

Point two might add that the Malibu also appears to be doing well in its first year. The Enclave started out strong, and remains about average. And similar to the Honda Pilot.

With the 2008 Lambdas, our sample size is well over 200. Would I like all sample sizes to be this large? Absolutely. Working on it.

If everyone who has said, "I'm not going to participate because the sample sizes are small" actually participated, then the sample sizes wouldn't be small.

TrueDelta will have updated results for all of these cars in May, with a preview of preliminary results next month for participants. Current results here:

TrueDelta Vehicle Reliability Survey results

[Sir Osis of Liver]

The products of the new GM (CTS, Outlook, Vue, Cobalt, Aura) have generally had much worse than average reliability in the first year, going by both CR and True Delta.

I have seen the True Delta numbers and the above quote is indeed correct. I know because I participate in True Delta. The new GM cars are not reliable no matter what GM faithful and apologists might want to spin.

For example, the 2008 CTS has a problem rate of 97 per 100 cars. For a 2005 it goes to 141.

In comparison, a 2008 Lexus ES350 has a rate of 27 problems per 100 cars.

The 2009 Corolla, a new model comes it at 31. A 2009 Cobalt (a five year old model) comes in at 55.

The 2008 Fit rates at 13 problems per 100 cars.

I could go on but there is a reason the General is going bust.

Two things:

1. The sample size for the vast majority of True Delta ratings is statistically too small to give any sort of accurate information.

2. Notwithstanding point 1, the rating for the G8 (47) puts it between the Hyundai Genesis (43) and the Infiniti G35/G37 (50). Not bad territory.

Point one goes a bit far. With the smaller sample sizes, the results are less precise, but are still accurate enough to be useful when comparing cars with widely differing repair rates. I wouldn't assume that a car with 40 repair trips per year is more reliable than one with a reported repair rate of 50, if the sample sizes are near the minimum. But 40 vs. 80? Then the former is very likely more reliable.

Point two might add that the Malibu also appears to be doing well in its first year. The Enclave started out strong, and remains about average. And similar to the Honda Pilot.

With the 2008 Lambdas, our sample size is well over 200. Would I like all sample sizes to be this large? Absolutely. Working on it.

If everyone who has said, "I'm not going to participate because the sample sizes are small" actually participated, then the sample sizes wouldn't be small.

TrueDelta will have updated results for all of these cars in May, with a preview of preliminary results next month for participants. Current results here:

TrueDelta Vehicle Reliability Survey results

I appreciate your efforts, and hope you keep fighting the good fight. Hopefully more people will respond. But currently, the sample size for a lot of models is really too small to form any reasonable conclusion.

An example: in 2005, Ford sold 901,463 F150s, the sample size you have for that year is 11. There is no way that 11 vehicles are going to be in any way representative of the better part of 1 million vehicles.

I don't want to get into a pissing match. I just want people to think about just what the numbers represent.

[mkaresh]

I'll readily grant that the #1 weakness of TrueDelta's results are the sample sizes. But I'd rather have the best process and need more participants than have plenty of participants but have a flawed process. In the long run it'll prove easier to "fix" the sample sizes.

Anything with a sample size below 25 is a broad indicator at best. That's why those results are asterisked and visible only to logged-in members. It's also why the results page explicitly states that those results are not for citation elsewhere. These results seem to be accurate more often than not, but the potential for an inaccurate result is well above ten percent.

FWIW, though, the size of the population is not relevant to the necessary sample size. It's not even part of such calculations. What matters: the variance, and how many variables you want to include.

[Cord]

But I'd rather have the best process and need more participants than have plenty of participants but have a flawed process. In the long run it'll prove easier to "fix" the sample sizes.

Excellent point.

[toolatecrew]

I saw a G8 GT V8 yesterday up close its its one nice looking car both inside and out. The price they are selling at is also very attractive.

What would tunr me off is the reportedly dismal snow performance of this car. The Canaddriver tests showed the combination of RWD and a very poor traction control system really hampered this car. There are a lot of places in Canada where it snows. If this was your only car or main family seden that would be a huge negative.

[mkaresh]

I saw a G8 GT V8 yesterday up close its its one nice looking car both inside and out. The price they are selling at is also very attractive.

What would tunr me off is the reportedly dismal snow performance of this car. The Canaddriver tests showed the combination of RWD and a very poor traction control system really hampered this car. There are a lot of places in Canada where it snows. If this was your only car or main family seden that would be a huge negative.

I live in Michigan, so was very interested in the details. Link here:

http://www.canadiandriver.com/ctc/blog/sedans/2009-pontiac-g8.htm

Apparently even winter tires weren't a huge help. This was with the V6, and the 2009 (so second model year). There's a possibility that the V8's system works better, but I wouldn't count on it. This car was primarily developed in Australia--and might have undergone little or even no snow testing. Not much snow in Australia.

Now I wish I'd taken one for a test drive this winter, to see if it really is this bad.

[rrocket]

I think it would really depend on the tires...and rim size.  A set of top notch winter tires on 16" rims as opposed to the stock sized rims on the G8 would make a dramatic difference.

Maybe Wing can chime in on the tire type and size used....

[toolatecrew]

I think it would really depend on the tires...and rim size.  A set of top notch winter tires on 16" rims as opposed to the stock sized rims on the G8 would make a dramatic difference.

Maybe Wing can chime in on the tire type and size used....

Not saying right or wrong but it sounded to me that the problem was the traction control system. The car did have winters (although it would be nice to know the details)

but the traction control is a little too aggressive. Stopping fully at stop signs meant I could not get going again at all; I had to switch the system off, get some wheel spin and then turn the system back on. At least this is easy, with just a touch of a button.

I got tired of getting stuck, though, so I just left the system off, which meant I was sawing at the steering wheel my entire drive, as the G8’s rear end wanted to step out on me even with the slightest of throttle inputs. So, unfortunately, as fun as it may sound, the G8 is not a great performer in the snow at all.

If the car can move while spinning its wheels from a stop but won't move with traction control system on its a system issue not a tire issue.