Welcome to our resource for MG Car Information.

MG MGF Technical  HGF statistics
Just been looking at the data on HGF stored at http://www.shame.4mg.com which now has over 260 reports of this expensive problem. Sadly, the data is incomplete, but some interesting trends are there, albeit trends that may have a number of alternative explanations. I noted that there are a total of 4 MGFs registered in 2000 that have suffered from a HGF all seem to be registered in the 1st 3 months of the year which would probably mean that these were 1999 build cars, and therefore built before the switch to steel head locator dowels. There is a HUGE number (proportionally of this population) of 96/97 and 98 model year cars reported. Again, a reporting bias occurs due to the early cars having larger mileages... The median age of registration is 1997 by far and away the worst year. Lower reporting rates in 95/96 model years may reflect the smaller number of cars produced. Interesting that by 98 model year that the numbers of HGF are equivalent to the 96 model year, and 99 there are very, very few reported failures roughly the same as we see in 2000 odd as we shall see below. Multiple HGF: 162 cars registered with one HGF. 15 cars with two. 2 cars with three. (yes, one of those belongs to Dirk ;o) One gets an interesting histogram from all this data, which at present has a marked left squew. This is probably because of a reporting bias for the first failure when the car is brand new, and that leads to p****ed off owners comming here to report the problem! Median mileage of HGF appears to be 28000 miles. A startling and rather morbid predictive equation has emerged from this you can predict the mileages at which each HGF clusters: miles = 22600 + 6790 x N where 'N' is the number of head gasket failures. The regression coefficient is 99%!!! Therefore the first occurs at 29390 miles, and each peak thereafter occurs at 13580 mile intervals... what exactly this means, I have no idea! Lies, damn lies and statistics. The numbers mean nothing, but their interpretation is a minefield. I did it because I was curious and figured it might appeal to the data freaks amongst you! ;o)) It's still early days to be 100% confident that the MY2000 and later models are more resistant to HGF (median age of HGF is 36 months) but given the left shifted bias of the population curves, I think we should be hearing about many more failures than we have done recently. Perhaps more telling is that the numbers of failures of 98/99 cars are much much less than would be predicted from the "bad" registration year of 97. Perhaps Rover made some changes at that time? The future for MY2000 cars/ mgtf? Time will tell, but perhaps we have reason to be optimistic. So to all those who have had a HGF please report it! http://www.shame.4mg.com 
Rob Bell 
You must be bored Rob! :) One more interesting point, in the last four months or so the number of reported HGFs has almost doubled, want to come up with a theory to explain that? :) 
MikeB 
Interesting. But I would imagine that if the 2000 HGF rates are going to start to rocket it won't happen until the sunny weather comes out and all the cars are on the road again. I noticed that most of the HGF's reported here on the board happened when the weather improved. 
Dave Quigley 
Out of interest, what are Elises like for HGF? do they share our problem or do their better ventilation stop this? Nigel 
Nigel Kidd 
I can't wait to see Mega's comment on this thread. :( 
Cecil Kimber 
Hi, I see the germans stopped entering the HOS though. Famous sample is Dot's old MGF (sold to someone in Bavaria. That one had the fourth now resulting in a new engine. From history I know that 1999 all new MGF got checked here in DE before they got sold. Changes of importance where (IMO)  viton sealing (green coloured) to the inlet manifold at all 1.8i in 1998  steel dowels to the reworked gasket in March 2001 The Elise has a few less troubles cause (IMO) the different cooling system. But there are several. The HOS is well known at the german www.elisforum.de :) >So to all those who have had a HGF please report it! >http://www.shame.4mg.com Dieter 
Dieter Koennecke 
>Therefore the first occurs at 29390 miles, and each peak thereafter occurs at >13580 mile intervals... what exactly this means, I have no idea! That you have far to much time on your hands? >:) 
Paul Lathwell 
LOL 
Dieter Koennecke 
<chuckle> Well, I was using the stats programme on my computer anyway for other data analysis, and then Stuart asked his question regarding HGF. So, given that mike has made the table on the web page so that it can be copynpasted into an Excel spreadsheet, I figured I'd have a look at the HGF stats. A case of what's less boring!!! LOL The overwhelming thing that comes through with this data, is that, assuming that sales of MGFs has been constant from '97 on (in fact, more cars have been sold year on year, I suspect) then SIGNIFICANTLY fewer cars than predicted from this year onward are being reported with HGF. What's the explanation? 1. People can't be bothered to report HGFs anymore it's old news. 2. The dealers have woken up to the problem and are sorting things out 3. MGR have made design changes that are effectively reducing the incidence of the problem. 
Rob Bell 
would anyone like to make an estimate of what percentage 162 cars is with respect to number of poster/readers of this BBS? 
paul weatherill 
One point ref reporting is that newer cars are covered by a 3 year warranty so people are not so bothered if they get HGF during that period, so won't bother to report. My car (already reported) was some 3/4 years old but had only done 10K miles when it had HGF, so you may find the incidence is agerelated rather than mileage, or some complex combination of both. 
T Green 
Interesting facts ROb. Obviously a science man.. I know 5 owners (including me) here in Portugal and NONE had a HGF! 98 and 99' cars... Cheers, Valter. 
Valter 
Paul, it is completely impossible to work out incidence based on this poll of owners, and statistically inappropriate to guess the sample size to which this 260+ reports relate to. Tim, you are probably right to imply that the occurance of HGF has some complex relationship between age and mileage. As I mentioned before, interpretation is fraught with dangers but it is interesting that the median age to failure is 3 years (36 months) and 28000 miles and as your case shows, there is a significant left shift of the data distribution with plenty of people reporting failures in young, low mileage cars hence the observation that few are being reported in 99 on cars is potentially significant. Good point made about the warranty issue a potentially very important confounding factor in the reporting of failures. Valter erm, yup, I confess, I like a bit of science! ;o) 
Rob Bell 
Well the stats on that table are dubious anyway. My car's on there and I've not had an HGF. I asked for it to be removed three times , but to no avail. I think Dirk (didnt he start it?) just added all the names of people on the BBS two years or so ago irrespective of what had (or hadn't) happened to their cars. 
Jim 
I'll paste this in from the MGF General category. MGR have taken lots of action on HGF. The failures are nearly always preceeded by some kind of coolant loss. Intake manifold gasket, dodgy pressure caps, poorly fitted hoses. I dont personally think the steel dowels are that great a factor. Revisions on the gasket have been pretty continuous and all the other sources of leaks have been adressed also. Regular readers of this BBS will know that I ended up at Land Rover in the split. We continue to use K series in Freelander and have noticed that the incidence of HGF has all but been eliminated. I am not sure the incidence in MGF was as high as this BBs would lead people to conclude as only those unlucky enough to experience HGF say anything. I think the problem with MGF is that its difficult to spot the leaks and even more difficult to refill/degas. My daughters Rover 200 had a leak from the bottom hose clip when she picked it up I put it right straight away. 20,000 miles on and 2 years later and still no problems and it hasn't ever used a drop of oil either. She's changing it for a MGTF in May. As someone else said, don't forget you get a 3year waranty now. 
Paul Hollingworth 
Hmmm,interesting stats Rob. Incedentally, as Dirk is one of only two to have reported 3 HGFs, does this mean that i, with 2 engines and no HGFs am his exact opposite? Scarlet is 6 years old @ 12 months = 72 months, divided by the 36 month incedence rate = 2, the same number of engines i have had!!! Also, with 88,000 miles on the clock, divided by the 28,000 mile incedence rate gives 3.142, or PI! Freaky! SF :) 
Scarlet Fever 
And Honda have produced more than 100,000 Type R engines with not ONE problem, not one claim! Mega: TWAT of the year 2002 Award 
Mega 
> as Dirk is one of only two to have reported 3 HGFs. *g*, mate, look up again ;) New calculation please. The 230th is interesting though. >http://www.shame.4mg.com Dieter 
Dieter 
Mega is absolutely correct  I well remember telling him this at one of his informal motor trade evenings when Herr Daimler was telling Bill Morris and Mr Austin how he made trouble free cars. My dear Megasan I am looking forward to your next get together and please can you invite Herr Fishtrousers as I have a job propositionm for him. Honda 
Hondasan 
Muuhhaaahahahiii !! 
Dieter 
Mega is in the running for 2002 STAT awards now as well... He just loves those trophys, well with one exception anyway! ;) SF 
Scarlet Fever 
This thread was discussed between 21/02/2002 and 28/02/2002
MG MGF Technical index
This thread is from the archive. The Live MG MGF Technical BBS is active now.