Are you a betting man? Or, if not, a betting woman? Recently, both Doug DeMuro and Tavarish suggested taking advantage of the 996's reputation for catastrophic engine failures to get a cheap Porsche 911 ("cheap" being relative to a 997; you won't be racing it in LeMons). Doug's solution: get the Turbo, which is immune. Tavarish's: spend a grand or two to pre-emptively replace the fatally flawed intermediate shaft (IMS) bearing. Either strategy assumes that the engines in non-turbo 996s are essentially "pin-less grenades." After all, anyone who frequents Porsche forums soon feels that it's not a matter of if, but when. But what if these engines aren't likely to fail? An intelligent, dispassionate gambler seeks situations where the odds assumed by the other players aren't the real odds, and bets the other way. Could the 996 911 and its kleiner Bruder, the 986 Boxster, be such a bet?

Nine years ago I started a car reliability survey because colored dots often lead people to greatly over-estimate the unreliability of "unreliable" cars. Fairly small differences were (and still are) often misperceived as the difference between "never breaks" and "in the shop all the time." In an attempt to fix this, TrueDelta has been reporting repair frequencies—numbers, not just dots.


I first learned of grenading Boxsters and 911s in late 2008 from former buff book EIC Stephan Wilkinson. Forum talk suggested that the failure rate was as high as 20 percent. (Note: when not insinuating near-certain disaster Tavarish rationally suggested half this rate, even further from "all of them.") Maybe TrueDelta's car reliability survey could determine how often these bearings are actually failing, and taking the rest of the engine with them?

We didn't have enough 986 and 996 owners participating at the time. So I contacted some Porsche forums, and both Rennlist and Planet-9 helpfully provided permission to post threads about the survey. Some members were, uh, less than supportive of the project (to be expected, these were forums), but about 200 986 and 996 owners signed up and have been helping. To avoid distorted stats, we don't count any problems that appeared before someone joins, so the next step was to wait as data accumulated.

During 2010, these owners reported four catastrophic IMS failures and two leaking IMS flanges (a precursor to failure). This was more or less in line with forum chatter. Then a curious thing happened. During 2011 we received only a single IMS problem report, for a leak. During 2012…no IMS problems at all. During 2013, these owners reported only two leaks, and no total failures. To recap, with about 200 owners participating, the total number of catastrophic IMS failures reported during 2011, 2012, and 2013 was zero. So far during 2014 three IMS failures have been reported. Two of these involved bearings that had been preventively replaced 12,000 to 16,000 miles earlier. Can it hurt to fix something that might not have been broken?


This suggests that IMS failures, however common they were initially, have not been common recently, probably about one in every 100 cars each year. Maybe those that were going to fail have already failed. (Forum talk has suggested that after 60,000 miles the chances of a failure are much lower.) Some owners might have headed off failures by preventively installing upgraded bearings. Probably some combination of the two. Whatever the reason, the result is the same: in recent years the great majority of the pin-less grenades have proved to be duds.

How is this possible, when common knowledge so strongly suggests otherwise? Well, people tend to be imprecise when factoring in time. Failures that occurred over five years ago get mentioned as if they happened recently. How many people, when disparaging a brand's reliability, bring up their 2002 Jetta, or even their 1982 Olds diesel? Beyond time, active forums have a way of making problems seem far more common than they actually are. If the forum has a few thousand active members, and even a few percent post about a certain problem, you'll have dozens of posts reporting the problem. It will seem like everyone is having it. Paranoia sets in. What's that noise?

Then we have FUD—fear, uncertainty, and doubt. When the cost of a failure is $15,000-plus, and the descriptions of the failure are so graphic, with pictures of the carnage, the odds get magnified. Most people will pay many times the actual risk (potential cost multiplied by the odds it will have to be paid) to avoid this risk if the potential cost is large. No one wants to risk feeling stupid and regretful later: "If only I had changed the IMS...."

The multiplier is even larger (among non-sociopaths) if you're not buying the car yourself, but are recommending it to someone else. Which is why so many car guys who would never buy a Toyota themselves recommend them to friends and family who want "a good car." And which is why I'm not going to flat-out say, "Don't worry about the IMS," either.

Heck, I know the stats, and I've got an RX-8 with failure-prone apex seals in my garage, and yet I'd worry about the IMS in a 986 or 996. One difference: there's no retrofit kit for the Mazda. The only choice is between buying the car and not buying the car. With the Porsches, you have the second choice of whether or not to buy an IMS retrofit. You can do something to prevent catastrophe. Are you still not going to? I'm not a terribly emotional person, yet the emotions might win this one.


But that's me, and not necessarily you. We're back to the original question: are you a betting man / woman? The cumulative odds with IMS roulette over the past four years have been about the same as hitting a specific number on a wheel in a casino, not a specific chamber in a revolver in Saigon. If you can sleep well with these odds, then take advantage of the market's over-reaction by buying a 986 or 996 and not touching the infernal bearing. If, on the other hand, even a small chance of a very expensive failure will make you feel like you're living moment-to-moment with the H6 of Damocles hanging over your head, then do as Tavarish suggested and spend a grand or (if DIY transmission removal isn't your thing) two and replace the bearing with an upgraded aftermarket unit. Just be aware that you're buying peace of mind, not spending a dime today to avoid a near certainty of spending a dollar tomorrow.

And the rest of the car? Also not prone to failure—check the stats. It likely helps that many Boxsters and 911s are not daily drivers. On average, they're only driven about half as many miles per year as the typical car. By choice (love that line), not because their engine grenaded.