The results are in: S6 reliability
Discussion
Haven't had chance to do too much analysis but, given that posts seem to have stopped now, here are the basics:
69 out of 176 cars have had serious engine problems. This is much more than the number of posts in the threads would suggest, as there are many who've had numerous rebuilds. Plus there are those who've put "no engine problems... except for the head gasket failure". Sorry, but to me that's an engine failure (especially given the low mileages), so it's in the thumbs-down category.
For those that don't have a calculator, that's a 39% failure rate.
And for those that like pretty pictures, below's a graph plotting all entered cars. Red is the mileage each car's done when it hits rebuild time. Green is the mileage those that've thus far escaped a rebuild have hit. And the X axis is the date when the car was made.
I'll have a bit of a think about what meaningful information we can pull out of this, but my (simplified) personal view is that I wouldn't even consider a pre-2002 car that hadn't been rebuilt. Truly appalling statistics given that there's been no recall.
>>> Edited by J_S_G on Wednesday 8th September 00:33
69 out of 176 cars have had serious engine problems. This is much more than the number of posts in the threads would suggest, as there are many who've had numerous rebuilds. Plus there are those who've put "no engine problems... except for the head gasket failure". Sorry, but to me that's an engine failure (especially given the low mileages), so it's in the thumbs-down category.
For those that don't have a calculator, that's a 39% failure rate.
And for those that like pretty pictures, below's a graph plotting all entered cars. Red is the mileage each car's done when it hits rebuild time. Green is the mileage those that've thus far escaped a rebuild have hit. And the X axis is the date when the car was made.
I'll have a bit of a think about what meaningful information we can pull out of this, but my (simplified) personal view is that I wouldn't even consider a pre-2002 car that hadn't been rebuilt. Truly appalling statistics given that there's been no recall.
>>> Edited by J_S_G on Wednesday 8th September 00:33
PetrolTed said:
Blimey, picture paints a thousand words eh?
awaits call from TVR...
Doesn't it just! Picture's hosted on my webspace, and won't be going anywhere for quite some time - that's for sure!
Worth making this one sticky, too? (I'll update with new figures/graphs if we get any more posts in either of the other threads, too) Edited to add: Beggars belief that you can withdraw "goodwill rebuilds" with stats like that.
>> Edited by J_S_G on Wednesday 8th September 00:18
HarryW said:
Oct 02 to Aril 03 inclusive looks like the safest period to get one from then
Not sure of the validity of this though, I appreciate well over a 100 responded and that it shows just under 40% have snags.
Out of interest how many Tuscans out there, reasoning being those with an axe to grind (no disrespect to those that have responded btw) tend to shout the loudest.
Slightly OT but very similar IMHO to the JD Power surveys where they solicit people to give their opinions on their cars. Those with an axe to grind do, those that are happy don't and all those with Skodas do too because they want to see it do well, justifying their purchase.
Regulations prevent TVR from selling more than 500 of each model per year, I believe, so, I'd guess at the following:
500 Tuscans per year 2000 onwards = 2250
350 Tamoras per year 2002 onwards = 875
350 T350s per year 2003 onwards = 475
100 S6 Cerbs per year 2000 onwards = 450
Rough estimate ~ 4000 S6 engines.
That gives us a 4.4% representative sample. The exact figures aren't really that important. Having spent many, many days sat in stats lectures in the past, the sample size is plenty big enough, assuming that the sample is representative, as you point out...
This is something that I think owners of good ones would want to post in, though - a bad rep = bad resale value on their car. Whereas those that've had problems might not want to admit to being so foolish as to buy a lemon, etc. Even if the figures are skewed by a factor of two, which is hugely unlikely (in my opinion):
1. That'd still give a 20% failure rate.
2. Just look at all those S6s that've yet to hit 10,000 miles, and look how many red dots there are above that line. Gut feel - Valve guides issue was fixed later than finger follower issue -> Longer before a rebuild, but still necessary for many...
delamars said:
Hang on. I would suggest that most of those rebuilds WERE actually paid for by TVR or warranty companies.
You also haven't differentiated between 4.0 litre engines and 3.6 litre engines. From recollection I think that you'll find most S6s built AFTER Apr 02 are sorted and MOST failures after that date were 3.6 engines which also appear to be sorted post Jun 03. Since then your figures show ZERO failures.
My guess is that the issues are pretty much sorted and TVR DID actually pay for most of them !
1. I've got a hole in my bank balance for several £k from a SECOND rebuild just a couple of months ago. Warranty company wouldn't touch it as they say it's an inherent flaw, and it's actually premature wear of the parts. I'm sure I'm not the only one...
2. TVR has made the statement that they're withdrawing goodwill rebuilds, even though there are clearly many early engines that've only done a couple of thousand miles.
3. Yep - could well be totally fixed post Jun 2003. Unfortunately, with only 1 car since then that's done over 8000 miles since then, it's very difficult to tell. Does look like a vast improvement however you look at it, though.
4. When I get a bit of time, I'll see if I can do a 4.0 vs. 3.6 breakdown.
swilly said:
JSG, if you want your research to be given credibility I would suggest you need to present it in an objective unbiased way.
Let those reading it come to their own conclusions and no doubt a thread(s) will arise to discuss it.
Your comments thus far appear to be loaded, and may effect peoples view of the results.
I'm not after credibility, etc. etc. The raw stats are there for anyone to do what they want with should this proceed further in any way. And, as with all stats, people will read into them whtatever they want to see. I doubt anyone here's weak minded enough to think "That's JS's opinion, so it must be true"!
I have stated them accurately, and I've given my opinion separately, beneath all of the facts, just as every man and his dog with an S6 will also do. Sorry, but I'm not going to be an apologist for the fact that at midnight I can't be arsed sanitising my views on what's clearly been an endemic problem.
Got the awful feeling this's going to polarise the views of those that've suffered a rebuild and those that've never had a problem...
>> Edited by J_S_G on Wednesday 8th September 09:03
True, but not a single engine over 30,000 miles full stop. Shame there aren't more high-milers (especially the newer ones).swilly said:
X-reg Dec 2000
30,000 miles, 16000 inlast year or so.
Previous owner had unspecified pre-emptive work done on engine.
Uh-huh.
You KNOW it's had engine work done on it, which means I should probably have put it in the thumbs down category. However, I had to remove it from the stats as the history's unknown to an extent - it'd be wrong to put it in either column... Maybe it was preemptive cos he/she didn't want to take any risk at all. Maybe preemptive means it was burning oil at the same rate as petrol and he/she didn't want to wait the 20 extra miles it'd take for it to fail completely & start chewing up spark plugs, etc.
Along with the 2 that I put from "thumbs up" into "thumbs down" for head gasket failures, I think those are the only judgement calls I made. I'll check this again tonight to flag up any other calls I had to make. But feel free to question any of those.
If it makes you feel better that TVR have had no engine problems, you can download the graph and stick a big green cross on it for yours.
Edited to add: I've been through the "no rebuilds", and the only other usage of judgement that I had to discount was apprentice's two "other" tuscans with "in excess of 10,000 miles" and no other stats on mileage, build date, etc.
Stuck butterfly valves, strange noises currently coming from the engine, replaced throttle bodies, etc. I've all left as "no problem".
>> Edited by J_S_G on Wednesday 8th September 09:47
andyvdg said:
P.S. Very difficult to conclude much from this graph. Maybe try visually comparing the 6 month period started Jan '00 with the period starting Jul '02. Similar spread and number of samples, much fewer red ?
A large proportion of people haven't given the reason for rebuild. Would be great if we could show that, but I'm not sure if there's the quantity of data needed. I'll take a look tonight, though, and see if there's anything around this I can pull into a graph.
swilly said:
I thought the issue here was to get stats on the 'engine failure' issue and not the 'engine problem of any sorts' issue.
As far as I am aware my engine has had no rebuild, any pre-emptive work done was done by the previous owner on the basis of other Speed 6 engines going bang.
OK, you win... your d
ck's the biggest. Who's to say what other work was done when the headgasket was replaced? Who's to say that your engine wouldn't have gone bang if it hadn't been rebuilt.
If the head gasket goes, the engine has failed. It's not a 200mph car at that point, it's a lump of plastic and metal. It's off the road. In the garage. Needing repair. It's failed.
Maybe I'll just p
ss off and do my day job rather than trying to help out and stop the endless "it's only a small percentage"/"it's nearly all of them" debates. andyvdg said:
Another useful graph would show engines with multiple rebuilds, or the mileage since last rebuild. In other words something to show the probability of a rebuilt engine going again based on rebuild month or something.
Didn't capture the data in quite that way, but was going to add it in. Will definitely do that.
swilly said:
Now calm down.
You've gotten together a good set of data, which after years of 'The Speed6 is crap-no its great' type threads on PH, could actually give people a better view of the issue.
It has already been said on this thread that this info could polarise the two sides of the arguement. On the contrary I would suggest this info SHOULD be able to present a pretty reliable picture of the whole debate.
But you are beginning to cloud things. I am only suggesting you dont turn this data into a "30% of road deaths are due to speeding" type of spin.
I get my spark plugs changed at each service. That is pre-emptive. Does that count as an engine failure too.
edited to say I agree my dick is the biggests
Think there's a difference between a few pound bill for pre-emptive work on a wearable item, and what I'm assuming is a bill for several £k on a supposedly lifetime part. Can we agree to disagree on that and give the stats an accuracy of +/- 1?
Agree that we should be careful not to read too much into something as simplistic as a graph like that - it doesn't include how many owners thrap their engines from cold, etc. (Then again, I wonder how many Honda owners warm their engines up...)
To me, there's a simple statistic: about 40% of all S6 engines have needed serious engine-related attention.
Here's a new stat for everyone:
Only taking into account engines upto Dec-2001, 56 rebuilds, 23 no rebuilds. 67%.
MEMSDesign said:
several people said:
Another useful graph would show ...XYZ
It might be best to dish out your raw Excel data or you're going to get run ragged here I reckon
The raw data is available to anyone who can be bothered to make the effort to read the threads. If anyone wants to take issue with the analysis, why don't they do their own, and post the results here?
As soon as I've put everything into the spreadsheet (reason for failure, etc.) I'll make the Excel file available for download. No problems there.
daftlad said:
Whilst I acknowledge there was a fault of the earlier engines, I think that this kind of pole will only ever distort the actual situation.
Certainly of 4000 ish engines 1600 have not been rebuilt.
Must be careful with statistics - something about damed lies?
The situation remains the same. All we've done is add a bit of information to what was spurious guesswork before. Can only help, however little. But yes, stats can be interpreted any way you like.
As for rebuilds, who knows... You could do a bit of guesswork (and this is PURE guesswork)...
1600 = 400 at TVR Power, 1200 at Blackpool.
2 days labour for a rebuild = 800 days labour at Power -> 1 guy full time for 3 years. Blackpool -> 2400 days labour = 3 guys full time over the last 3 years. Technically feasible? I'm sure. Accurate? No clue.
We can't go saying "they must have only had X go wrong overall"... You can only extrapolate from the data that's available, not work back from what you think sounds right.
Ted - get some more S6 owners to sign up quick sharp so we can get some better stats!!! joospeed said:
LOL.. when the next person phones me and asks why i don't work on SP6 engined cars (engine related work anyway) .. I'll show them this set of figures, should make them realise!
good work JSG .. frightening set of numbers ..
Ah, Joolz, that reminds me - I think I need to pop up to see you with the Tusc, actually - having sorted everything else out (tyres, tracking, etc.) I think I might need some softer springs on the car.
Promise I won't make you look at the engine
jimmyt said:
How about:
Only taking into account engines built from Jul-2002, 6 rebuilds, 55 no rebuilds. <10%
I don't think anyone can argue that that's not encouraging especially as some of those cars have plenty of miles on? Agreed 10% is still too high though.
Absolutely... Including July 2002, the failure rate of the sample is about 9%. A huge, huge improvement.
Still, think this would be a dire number resulting in recall if it were any major manufacturer
millyad said:
sorry to be an idiot!
who do i contact to find out whether my tuscans had a rebuild?
Many options:
* Ring the factory and ask them to go over the books (they'll need your engine/chassis number, probably)
* Ask the dealer/previous owner
* If it's a 2000/2001 Tuscan, drive it for 10,000 miles and see what happens.
daftlad said:
JSG,
For ammusment what you've put together is fun. Its not representative, and if you've sat in many a STATS lecture, you'll know why.
Maybe I should start a thread along the lines of
"i wonder what % has been wiped of the value of my car by inaccurate distorted polls being run on a website"
Taken out of context, and many stats are, this has all the potential to turn into a quite damaging piece of work.
It's representative of certain things, not of others. It doesn't matter if the statistics of pre 2002 cars are actually 10% or 100%. Even if the accuracy of the survey is only 25% or so (unlikely in my opinion), it's double figure percentages, and hundreds of actual failures - whatever way you look at it or bias you decide upon. And that's not a cost that the owner should have to bear whilst being mocked with cries of "T.A.D.T.S.".
It's as good a sample as has been done thus far, that I know of. May it cause damage to residuals? Absolutely. Maybe those with unbuilt 2000 spec engines with 1500 miles on them out there will lose a grand or two. Maybe those with 2004 spec engines will make a grand. It doesn't matter - the damage/repair will have to be done at sometime, by someone. There's been years of idle speculation on this site that's done untold damage.
Maybe by tomorrow this thread'll have slipped down into obscurity. Maybe someone'll decide to get a group of people together and claim for damages. No matter how small the step forwards, I personally see this as progress from the "you're an isolated incident", "no, you're just yet to have it happen" arguments that fill the S6 forums.
The only blame for anything in this lies at TVR's door, I'm afraid... if there'd been a recall 2 years ago, these threads would've died a death a long, LONG time ago. As it is, this is causing a great deal of interest because there's been a breakdown in the relationship and communication between TVR/dealers/owners - a triangle that needs to have a level of trust. Until I go on Watchdog saying that noone should buy a TVR because they all blow up and you'll be funding Soviet terrorism, please don't blame me for trying to do the only things I can to break the current deadlock.
>> Edited by J_S_G on Wednesday 8th September 14:55
daftlad said:
Point well made, and i guess I agree with one notable exception - TVR aren't the only peolpe here to balme. 
Think owners and mechanical sympathy.
Agree - that is, no doubt, a consideration. And a point that dealers should probably be much more vocal about to potential owners. In all the years of the family's garage, noone ever warned a customer about not thrashing a Honda/Toyota/whatever engine from cold, though, with no unhappy clientelle... And we're not just talking 21st century engines here.
BCA said:
Thanks J_S_G for trying to find out exactly what everyone was having a go at me for asking previously. On the plus side, that 30k car is a new mileage record, before I had heard of no better than 24k
No probs. But, unfortunately, none on the graph has done 30k before a problem occured. That car is bandit's...
bandit said:
1999 Speed Six Cerbera.
14,000 miles - finger followers
23,000 miles - valve guides/seals - lost compression
30,000 miles - head gasket
So, by 30,000 it was having its fourth engine dropped in.
I'll do another graph when I get a chance (been a bit busy tonight) that subtracts these mileages, so would put these at rebulds of 14, 9, and 7 respectively.
Looks like mty4000 holds the record at the moment with 27,000. What a record. [shakes head and wonders if this really is the 21st century]
Edited for shocking grammar, etc.
>> Edited by J_S_G on Thursday 9th September 00:37
Gassing Station | Speed Six Engine | Top of Page | What's New | My Stuff