According to the poll here; 60% haven't had a water pump failure yet. Perhaps the common failure, may not be that common after all?
I don't mean this nearly as harshly as it will probably appear, but that's an absurd conclusion!
First, who'd agree that 40% premature failure is acceptable? If you say that 60% haven't had a water pump failure yet, you're also saying that 40% have -- and you're saying that a 40% failure rate is acceptable.
Aside from that: Look at the numbers in the poll -- which, no offense, has serious flaws of its own that make the results virtually meaningless:
* 25 of the respondents who haven't have a WP failure have <50,000 miles on their car, and I'd bet that many of them have substantially less than 50k
* 17 had a failure at >50k miles -- and there was no provision for indicating that you've had more than one WP failure (as has my car)
* The number of NA & GXP cars totals up to 44 -- while the responses in the "Yes" and "No" options total up to 58. With no provision for one voter to indicate that they've had more than one failure, and presumably the poll software won't allow the same voter to vote more than once, shouldn't the number of responses be the same as the total number of NA & GXP cars combined?
The low number of responses -- either Yes OR No responses -- would probably be considered statistically meaningless when you consider that there are supposedly something like 23-24,000 members of this forum (which seems like an impossibly, unbelievably high number -- but I've seen it somewhere).
Even if the number of forum members was 2300 -- not 23,000 -- 44 cars being reported on or 58 total responses is statistically meaningless.
And finally, here's a thought I'm sure has some merit:
* MANY Solstice owners have sold their car -- and many of them have done so due to the problems they had with it. With water pump failure appearing to be one of the most frequent problems Solstice owner's -- including ones who are current forum members, but certainly also including many who aren't -- have had, doesn't it stand to reason that water pump failures have occurred to FORMER members, and have almost certainly been instrumental in causing the sale of a lot of Solstices?
And isn't it virtually certain that many former owners who sold their car due at least in part to water pump failure are no longer forum members -- and thus didn't participate in the poll?
The anecdotal evidence that many -- too many -- Solstice water pumps fail at an unacceptably premature mileage -- relative to how many miles water pumps typically last in ALL vehicles -- is overwhelming.
Just because it hasn't happened to YOUR car doesn't mean that it hasn't happened to FAR more cars than it should have, based on industry norms.
Your reply doesn't really apply to KellyfromVA's post without extrapolations and assumptions.
FWIW, KellyfromVA never said a 40% failure rate was acceptable. More specifically, he never said anything about what is or is not acceptable for a failure rate so it is incorrect to try to debate something that does not exist in his statements.
KellyfromVA only suggested that based on a 60% non-failure rate, perhaps water pump failures weren't as common as people think. I don't see any major problem in his statement. In fact, it is based on a qualitative judgement that by its very nature is open-ended. Your reply has lots of detail, but it doesn't really refute his general statement. I thought I'd point this out as I've had people do the same thing with some of my statements.
As for your feedback, I generally agree with much of what you say. Of note, for the number of registered forum members, yes, it is a general number based on forum information, which comes with the caveat not considering now former members, active versus inactive members, banned members, etc. Even so, forum numbers are high enough that it appears that a large sample population would still remain after accounting for those member caveats, from which good poll participation would give a reasonable answer to many of these questions. Unfortunately, there hasn't been much member participation in this poll and as such, statistically speaking, there is not enough statistical power to draw any significant conclusions on specific failure rates or mileages.
The use of a positive/negative outcome poll itself is a biased population sampling method that can often skews results, which is very problematic and open to discreditation without a large amount of population participation. If people really want to get an estimate without that bias and at a lower sampling rate, a randomized population sampling technique is needed. One way to do this, for example, is to generate a list of forum members, randomly select a number of them (e.g., 200), PM them and ask them to fill out a poll only open to them (or answer a series of basic questions). If 150 people from the notification participate in full for the poll (or answer all questions), then there will be some good data there. If not enough people participate, select another 200 people in that list randomly and do the same thing (PM and ask for participation), until you finally get at least 150 people who participate. If doing questions, they should be simple and not use language that has inferences (e.g., jinx me, borrowed time, etc).
I said this before and I'll say it again for the thread, anecdotally, everyone I've seen post here (and myself) seem to agree that the solstice has an abnormally high water pump failure rate. People aren't disagreeing with that.
There has been a long debate about relative rate of failure and at what mileage it occurs. It is because some people feel there is lots of data, other people feel that there is not enough data from which to drawn detailed conclusions. Statistics can tell people the power (or lack of power) behind the data, but it isn't easy for people to understand and for some it doesn't really matter to them and their conclusions, so the debate goes on.