Cheating turns hippie chic into corporate greed!

By now everyone has heard the story of how Volkswagen created software to cheat Federal Air pollution standards. The software was aware when the car was hooked into analysis equipment that measured pollution. When it was tested, the software would run in a special mode that produce far less pollution than it would while on the road. Volkswagen’s stock has been rocked by the confirmed allegations of cheating, and the financial and legal consequences that will follow.  Yes, Volkswagen is in deep trouble Yet, we probably don’t know everything there is to know about the culture of cheating that led to this problem. How do good companies lose their way and become dishonest?

The auto industry’s attempt to cheat on performance standards is hardly a new story. We now know that Volkswagen used “defeat devices” to fool regulators for six years, making their diesel engines appear to be vastly cleaner than they were. SPOILERS! When an industry has come up with a name for devices that allow them to cheat on government tests, it’s not exactly an industry secret.

Today it’s a cheat for diesel, back in the 1970’sVW had a cheat to reduce gas usage when the temperature was low. In the 1990’s, General Motors used another defeat device to show low gas consumption when the A/C was on. Before cars had computers that could be blinded by software, there were stories about car companies that declared the backseat of the car and the spare tire as “accessories” and removed them during mileage testing. Cheating on government tests is more than a car company conspiracy, it is a car industry tradition.

In fact, this type of cheating is more than just something we see in car companies. We see it in food packaging (low sodium or low fat… what exactly does that mean?). We even see it in cigarettes. Remember “low tar” cigarettes? The low tar rating came from devices that measured the tar content of cigarette smoke. However, these special cigarettes only produced less tar when they were smoked by these machines. When a human being smokes a low tar cigarette, the tar level is MUCH higher. More importantly, the rate of cancer and related diseases is identical to that for high tar smokers.

As disturbing as all of this statistical tinkering is, it is nothing compared to the heavy duty cheating in the computer industry. When you buy a new printer, have you noticed that you get a lot less than the printing speed on the box? When you measure your monitor do you notice that it is usually a bit smaller than the size you are told? Still, the biggest lies of all were told by video card manufacturers.

In the late 1980’s, few computers had any noticeable video capabilities. They could put text on a screen, and they could do it in color, but that was about it. As business graphics, predecessors of PowerPoint, started to become commonplace, specialized microchips that improved performance started to appear on the market. Architectural and engineering software needed a video boost to be moved from expensive graphics workstations to much less expensive PCs.

Then, the video game “Doom” arrived, launching the era of graphic video games. As the gaming industry grew, so too did the video card industry. This created a new problem,

A high performance video card… or is it?

with so many different PC cards, built on so many different chip sets, how do you know which one is the best for you? Why, you run a benchmark program, usually created by the video card manufacturers, and you can then see which card is better! But then a funny thing happened. After a while more and more benchmarks were released, and people started to notice that different benchmarks favored cards from different manufacturers.

If we look back to 2003, we can see a great example of video card cheating. It was summed up by a statement from Futuremark, a benchmark manufacturer, “We have now established that NVIDIA’s Detonator FX drivers contain certain detection mechanisms that cause an artificially high score… We have just published a patch 330 for 3DMark03 that defeats the detection mechanisms”.  And with that, the video card wars were truly underway.

Card manufacturers found ways of cheating the tests, or were releasing their own dishonest tests, and independent testers and benchmark manufactures were fighting back with new code and testing methods. After a while, “cheating” evolved into pure fantasy. Video cards had code to takeover the benchmark software and returning wholly fictitious results. The card manufacturers didn’t seem to care what would happen when their customers found out that they had been lied to. They were far more interested in getting their highest (fake) ratings. That’s the big lesson about benchmarks and tests. When any test or performance goal is critical to a purchase decision, the test process had better be accurately described, tightly monitored and heavily penalized when there is cheating. Without these standards, every corporation will, given enough time, cheat. Because there are too many incentives to allow “everyone else” to cheat.

The same applies to not just car and computer companies. It applies to every organization, and individual, that generates and reports performance metrics in a large corporation. If corporations are willing to use illegal means to cheat on government mandated tests, the desire and ability to cheat on internal metrics must be vastly greater. Was that project completed on time and on budget? Did you reduce your staff as requested… including the temporary workers that no one knows you have? We are all motivated to cheat, even if your firm has a long and comprehensive a statement on ethics.

As corporations head into ever more numerically based business, do we really want to know who is cheating? And do we really want to punish cheaters with severe penalties… or are we more likely to give them a big bonus and a great recommendation for their next job… if everyone keeps quiet? Lets keep an eye on Volkswagen over the next few months  and see what turns up. But let’s also think about reports with performance metrics that are just too good to be true. Because, they just might be. And that’s my Niccolls worth for today!