Recommend a cheap car for commuting?

[badass]

OK, so I tried having a look on the Which website, but it is behind a paywall so I guess I will never know.
However, I was asked to do a survey of my car ownership experience for which I could win some fuel. I hope they base their results on more than than a survey of the opinions of the sorts of people that buy a Which subscription though (and no, I didn't bother with the survey).

So, based on nothing you assumed it was based on warranty claims and then stated it as fact?

How else could you realistically get meaningful statistics on car reliability other than from surveys from a sample of owners, whether the cars have had failures or not? It's far from perfect but it's certainly better than using warranty company statistics as they only have information on cars for which a successful claim was made compared to the warranty holders.

I was going to paste some screenshots but the only links for pics have to be URL's and I can't be arsed with the faf to be honest so I'll summarise what you see here:

There are up to 5 stars for any car.

Example car:
Brand: Alfa Romeo
Sample Size 300 people
Overall owner satisfaction for this brand: 81%
Dealer Sales Service: 3 Stars
Dealer Servicing & Repair: 3 Stars

Alfa Romeo 159 (2006-2011) Sample size: 40 people
Overall owner satisfaction for this range 84%

Driving enjoyment 5 Stars
Ease of driving 3 Stars
Comfort 4 Stars
Dash layout and controls 4 Stars
Build quality 2 Stars
Practicality 1 Stars
Value for money 3 Stars
Running costs 1 Stars

Up to 3 years old 4-8 years old
All cars Reliability - 1 Stars
Breakdowns - 31%
Average annual repair cost - £193

Note the all cars means they have combined petrol and diesel cars stats as the sample size is too small. Where they can they separate diesel and Pertol cars.

Most common faults
4-8 years old Failure rate
Battery 26%
Suspension components 14%
Exhaust System 9%
Fuel System 6%
Cam Belt 6%
Failure rate: The proportion of owners reporting this problem in the last year

[badass]

A) I didn't I called the survey a load of guff, which even if it isn't based on warranty data, it probably still is. Selection Bias still exists in a survey of which readers. If you cite paywalled surveys people have no choice but to make assumptions about the content.
You won't find any stats on car reliability or anything else car related without selection bias. It's not practical to do anything without some sort of selection bias.

So you took someone elses incorrect assumption to support your argument and stated it as fact.

B) it still doesn't address the point that manufacturers are not reliable or not, models of cars are. Would I trust a 107 more than a 308? Absolutely. Just as I'd trust a punto further than a fiesta ecoboost. Or for that matter a diesel focus further than the focus ecoboost.

There's a degree of truth in that statement, however I could hardly provide a model by model breakdown for that post. However, grouping cars by manufacturers still has relevance. Whilst there may be some great reliable Fiat models out there, there are lot more reliable Toyota models.

Also, on the Aygo point, just because a car is made in the same factory doesn't mean that the same people work on them.
You might find that the quality control processes are different for example. There may be many other ways in which they differ.
However I would expect them to be broadly the same, reliability wise.

I've no interest in who pays attention to who to be honest, I was just pointing out that the assumptions you are making about reliable manufacturers don't hold true, although they may have once.

I didn't make assumptions. I recently bought a second hand car. I did a lot of research on various models and found that what I stated in that post held true. The statements I made were based on researching lots of models from those manufacturers. Notice I even listed some exceptions for specific models or even engines.

[DanceswithUnix]

So, based on nothing you assumed it was based on warranty claims and then stated it as fact?

Not quite, I made a general comment that these things tend to suffer from selection bias and cited warranty claims as an example, partly because I seem to see the figures rehashed and partly because the bias is so bad in such ratings.

So my assumption was that the Which report would probably suffer from selection bias in some form. The only people who make a concerted effort to get a good sample size is JD Power, but their survey is for new cars and says almost nothing to me as a purchaser of second hand vehicles.

Thanks for the pasted information. One star for running costs seems a bit harsh but they aren't the cheapest to run. Practicality one star I just don't get, but never mind.

Common faults seem a bit odd though:

Most common faults
4-8 years old Failure rate
Battery 26%
Suspension components 14%
Exhaust System 9%
Fuel System 6%
Cam Belt 6%
Failure rate: The proportion of owners reporting this problem in the last year

Battery?? If the car doesn't start nice & snappy, get it replaced, it is a consumable in any car. Usually the car gives you fair warning via the dashboard display that you need a new battery long before you get stranded anywhere. Just had this on the wife's car, I think on a cold morning it said "Low battery voltage" after starting. Battery swapped, should be good until at least 2020 now.

Cam belt, well blimey that is a rather scary! If that was belt replacement, then a 5 year schedule should place it at around 20% which would imply that the owners aren't getting their cars correctly serviced. That seems way out, so sadly I suspect that this is cam belt failures (with the resulting £1000 engine rebuild). Why sadly? Because if the owners got their cars serviced properly then the cam belt wouldn't snap, 6% is way higher than I expect. So either scenario, the cars aren't being properly serviced. This is backed up by the "average annual repair cost", because I tell any prospective Alfa owner to budget double that in replacing discs, pads, suspension bushes or whatever else the mechanic can see isn't going to make it thought the year.

Note that although I am saying the cars aren't being looked after, it could be well be the fault of the garage they took it to. Most don't see enough Alfas to know the car properly, so as well as expensive parts that does add to Alfa ownership.

Suspension components is a fair cop Alfa thing, they seem to like using bushes made of cheese. Some people get the bushes replaced with aftermarket ones and never see the problem again.

The rest could be a mixture of things, so hard to read much into those.

So thanks for that, it was very interesting, but I haven't changed my opinion of the survey. You have also reinforced my usual line that I won't buy a second hand car unless the cam belt is recently changed with better proof than "well it looks quite new" because people just don't look after cars properly, even an Alfa where you would hope the buyer to be at least a bit of a petrol head.

[herulach]

So you took someone elses incorrect assumption to support your argument and stated it as fact.

There's a degree of truth in that statement, however I could hardly provide a model by model breakdown for that post. However, grouping cars by manufacturers still has relevance. Whilst there may be some great reliable Fiat models out there, there are lot more reliable Toyota models.

Also, on the Aygo point, just because a car is made in the same factory doesn't mean that the same people work on them.
You might find that the quality control processes are different for example. There may be many other ways in which they differ.
However I would expect them to be broadly the same, reliability wise.

I didn't make assumptions. I recently bought a second hand car. I did a lot of research on various models and found that what I stated in that post held true. The statements I made were based on researching lots of models from those manufacturers. Notice I even listed some exceptions for specific models or even engines.

If DancesWithUnix hadn't posted I'd have made the same assumption as thats generally what those types of data survey are based on.

The stats you've posted are useful, but really just reiterate the point. Appreciate that the 159 is a bit of a rare car, but making comparative assessments based on those sample sizes is an exercise in futility.

Its not clear how the stats are calculated i.e. were there 12 breakdowns, 3 of which were battery related, or did 10 cars have battery related failures in the last year? Either way they're tiny numbers relative to the total numbers sold and could be a result of poor maintainence. They also don't speak for engine types.

By way of example I used to run an 02 Hyundai Coupe 2.0, the engines are very strong, although things like window winders fail regularly due to them getting water in. I don't know how that would look in those stats, but if people are self reporting electrical failures I'd suspect they have a 100% unreliability rating, even though the engine never missed a beat. The 2.7 on the other hand is tosh, one of the bearings (IIRC big end) is pretty much guaranteed to go before 60k.

Once it gets beyond a certain age (Say its 5 years although i've no data on it) reliabitly is as much a result of it being properly maintained than anything innate to the car. I'd trust an impeccably maintained 159 (or even a 147!) much further than a sketchily looked after Mondeo.

[Rave]

I do my 8-mile each way suburban commute on a scooter and have done for a couple of years. Bought it brand new for £1600, a Honda Vision 110. I've never got less than 110mpg out of it, I once got 125mpg in the middle of summer. Insurance this year with 2 years no claims was 86 quid, and tax 17.

However if you don't fancy taking to two wheels I'd be inclined to just buy an old petrol ford of some description. A Mk4/Mk5 Fiesta 1.25 would be just the ticket for a little commuute, just check carefully for rust in all the usual places. Alternatively a 1.4 or 1.6 Mk1 Focus would probably cost about the same. If you can find an old Honda or Toyota that's been owned by an old person rather than a barry boy then that'll likely do just as well, though be less interesting to drive.

To be perfectly honest, I think buying secondhand cars is largely down to luck. I bought my lovely 05 reg 3.0 V6 Mondeo Ghia X, with leather and all the toys, 120k on the clock 2 years ago for £1600 with a year's ticket. In 10,000 miles all it's needed is an oil and filter service and a 70 quid coil pack. I changed the plugs as well but the old ones looked mint when I got them out. I'm even still on the tyres it came with having rotated them once (will need new ones fairly soon now).

My brother OTOH spent £5k about a year ago on a 57 reg C-Max 2.0 (petrol) automatic with only abut 30k on the clock. Last week, after him doing about 14k miles, the auto box crapped itself and he's having to lump out 1100 quid to get a refurbished one fitted.

That's why I favour buying big cars over small ones, and bangers over cars 3-5 years old. If you've got the space and can afford the tax and insurance there's a great deal to be said for having 2 cheap old cars over one nice new one. My MX-5 is sorned at the moment awaiting a new roof but I could in theory tax it tonight and drag it out of the garage for some reason if I needed it! As it is my brother's been driving my car on his "driving other cars with the owner's permission" entitlement, and I've just borrowed my father-in-law's as and when I needed to!

Edit: come to think of it a Mk1 MX-5 would be a leftfield but surprisingy sensible choice of commuter vehicle! This time of year they're going cheap. £600 gets you into one that needs a little work, and £1000-1200 should get you one with a watertight roof and solid sills- the sills are the only bit that really rusts on a Mk1 AFAIK, Mk2s surprisingly can be more problematic, something about the front subframe mounts rotting out I think. The mechanicals are typically Japanese bulletproof. They're not amazing on fuel but bimbling down your country lanes to work with the roof down should put you in a good mood for the day!