any evidence on the RATE of bore scoring? If not, I'll compute it

The thread title is pretty to the point. And assuming no one has computed the rate, has someone put together a single good dataset on "clinically" significant bore scoring cases, with which we can estimate a crude rate, when divided by some reasonable denominator? As a medical researcher who trades in data and evidence-based studies each day, to me, there has been an elephant in the room on bore scoring. The first thing we do in a medical paper is review (or, if unknown, determine) the statistics on the burden of the problem - any discussion of a problem should lead with this piece of information. The scope of a problem dictates the level of interest and concern about it - e.g., while pancreatic cancer has a generally poor prognosis, it doesn't dominate medical discussions because it affects <<1% of the people. I don't see that kind of fundamental context so far. For example, the informative videos on bore scoring jump right into discussion on the problem's seriousness and causes, with a general tone that might lead the viewer into thinking bore scoring is a certainty, or at least a very high-probability issue. The first thing I would like to hear is how common it is, or at least a declaration that we don't know. Without that knowledge, you get people feeling like... https://rennlist.com/forums/996-foru...ng-me-off.html On gathering data, so far, I find items like: https://rennlist.com/forums/997-foru...-2-engine.html https://rennlist.com/forums/997-foru...s-issue-2.html https://rennlist.com/forums/996-foru...96-owners.html https://www.pistonheads.com/gassing/...=1052310&i=520 I see some rough estimates of the rate in some of these discussions. It also seems to me that some of the cases across websites/sources are likely the same guys using multiple sites...so the rate perception may be inflated. If we can amass some good data, I'll volunteer to do the analysis....I have statistical software at my command, I just need data. I can picture controlling for climate, mileage, stratifying by model and year, etc. to produce a set of adjusted rates, in addition to the crude rate. The worthwhile product might be at least a somewhat evidence-based estimate of the risk of a bore-scoring-related failure for a given car. It would be crude, but a start. To understand how I do this sort of thing, see the methods used in my regular work - you can see some of it at: https://profiles.stanford.edu/alan-n...b=publications Note: I made one edit to delete a caution about info from shops that work on our cars. While we would normally see that general type of info as conflicted in medical studies (think a shareholder in a pharma company running a study on a drug they sell), I believe their input could be valuable if properly used. I believe the plurality of data sources and the independent nature of the analysis would make it OK.

Best of luck and i applaud your willingness to attempt to tackle this issue from a purely statistical angle. Not sure how data will be gathered, or exactly what data would be needed/helpful, but i will be following this thread with much interest.

Not sure how you even begin to calculate this. Forums are usually perused by "enthusiasts". There may be lots of folks out there driving around that have no idea what a bore is, let alone if it is scored. Hope you find your answers.

You'll never come up with this rate with any sort of reasonable accuracy. There's several reasons why: - No one knows how many cases were mis- diagnosed -No one knows how many engines were covered under warranty, with an unknown diagnosis, and sent back to Porsche - No one knows exactly how many engines were built -No one knows how many people sold the car like it was, and was parted out These are just a few reasons why these failure rates can't ever be calculated with a level of accuracy that makes them worth doing.

Porsche company has the data and has done the analysis, and to get it from them you file a lawsuit that has merit and after you win a motion to compel discovery they have to provide it to you. But I don't think it's necessary to have the actual data; rough estimates from a Porsche engine expert such as Mr. Raby or Mr. Hart, along with a camera taking pictures inside each cylinder of your engine, should give you all the info you need for when you must address your engine block. Does bore scoring come from cylinder ovalization, or from something else?

Quotes source post: 16240705

No, they don't either.. They only have the data for the engines that came back to them. I know several hundred that never did. And, these cars are old, a lawsuit now would be laughed out of court.

Quotes source post: 16240705

What this, and the rest of them... You Tube

It sounds to me like the rate of bore scoring is 100% - the missing variable is time. Oh or if another mode of failure doesn't get to the finish line first.

Ask Rain Man.

I'll start my ratio was 1:1, but never again! I like the Rain Man idea... You Tube

I think it’s due to the materials in the cheaper pistons used back when Porsche was trying to make a buck and sacrificing quality a bit when the 996 came out. It has to do with metal expansion and silicon levels. Apparently the bore scoring cars were from places with harsh weather mostly. People on forums are usually reporting bad news and complaining. While the guys who make their living working on these cars will hear a small a noise and tell you your car needs a 25k rebuild. They will also tell you your ims needs to be replaced every 5-6 years with no real reasoning but fear mongering. Don’t get me wrong, these guys have done tons for the m96 crowd. I live in California and have never seen or heard of a friend or a m96 of my own having issues at all. This is about 15 cars. I know issues exist and I follow strict warm up routines and maintenance. Who knows? It gets tiresome when people hear a tick and almost every m96 has some kind noise is automatically told bore scoring. Sell your soul and house to get a rebuild. I don’t believe it. These engines are 20 years old and act how they want too sometimes....

Quotes source post: 16240705

Mr. Raby is NOT an expert! He�s still learning!

I would suggest you contact some of the Power Train Warranty Companies for data on particular failure percentages on various vehicles. They have the most to loose and have the most accurate data on failure outside the warranty period, which is the data that you probably want for this study. Automotive repair facilities are not REQURED to report failures to reporting agencies like is done in Health Care Systems. So the data is not collected anywhere, except from businesses who make their living off the vehicles such as trucking fleets or Taxi Companies, ect. I'm sure the Porsche Driving Schools had data on their down time for repairs/failures but were limited to 20k 30k miles on newer model cars, and some race teams who had Rent-A Seat programs for certain class track cars had data to track costs of repairs, but again that would be limited data. Again, Power Train Warranty Companies would be the best bet for data on Porsche engine failures and their causes, since there livelihood depends on knowing the risk of "covered failures". But whether or not they will be willing to give out that data is very questionable, they probably won't.

Quotes source post: 16241211

Exactly. You can only become an Expert when you know it all... I never will get there. Every single day, is a school day.

Quotes source post: 16240950

^This is the impression I think many carry. I'm agnostic - you have to start (and stay) that way to do a good analysis. But to follow the theory: The theory might hold that there is a global flaw or set of flaws in the 996 design that merely requires a given set of environmental conditions to manifest as failure - all would eventually fail if conditions were aligned under this view. I believe this is the general message, or at least the impression I get, of Mr. Raby's videos. (Here, "failure" might be defined as bore scoring of sufficient severity to create noticeable symptoms - this needs discussion, too). The theory might then also hold that some (maybe many) cars get lucky. That part of the theory would say that with the right set of conditions, failure never happens within the typical observation time (or life of the car, depending on the mileage attained). We have 996s with very high mileage and multiple owners. So if the theory is true, in such cases, perhaps each owner was a maintenance fanatic, happened to drive the car just right, the car was always in the right climate, etc. There are many unknowns and statistical challenges here, but as originally stated, pursuing this study would be recognized right up front as a crude first swing at independently analyzing the topic. But, I'm cautiously optimistic - based on experience in many statistical analyses, with a surprisingly small amount of data, I find you can start to see the truth emerge. In years of conducting all sorts of crazy analyses for the DoD and others, I've done OK with murkier situations than this. I can picture the data collection spreadsheet that the software would turn into the working file; columns would represent variables that we can actually capture on most cars, even just with crude internet scrapes. I can think of a number of plausible variables, here is a start. It is important to recognize up front that many information items will be imprecise, but with enough data, a pattern typically starts to emerge. The data can also be augmented in time using direct contact with owners: data source (website on which the problem was reported, or other source) if source is a website, user name - in combination with careful selection to avoid duplication of the same car, these first two variables might help rule out other duplicates mileage at diagnosis - recognizing the potential imprecision here model year model - we could bring in the whole M96/97-using family body type - sounds silly, but controlling for unexpected factors often proves useful climate at last ownership - while recognizing the car could have spent most of its life elsewhere - with enough data, this should wash out number of owners - will help with the issue just above - interaction terms may help here perhaps nature of failure or symptoms - need to think about it more Given the ability to PM owners on web forums for info confirmation, I think the POTENTIAL for a good study is here. If someone can assist with data gathering, I remain happy to explore this. PM me and I will provide contact info, and let's talk. If we have multiple interested parties, this could be a team effort (and maybe a cool paper for an actual journal could even come out of it). We might task-organize the data gathering so that each of us takes a given website or set of threads. I'm open to ideas.

By the way, I tried to contact CamsPorsche (who started a poll of overall cylinder-related failures) to assess his interest, but his mailbox is full...