Okay, let’s say that for every person that McDonald’s settled with, there were 10 more burns that we never heard of. That would be a total of 7,000 coffee burns, out of 10 billion sales.
Is this a big number? In safety engineering, there are certainly levels of acceptable defects. For defects that do not cause serious injury or death, the standard is often about 3 sigma. For more critical defects, the current standard is around 6 sigma. Six Sigma is the ‘gold standard’ for process control. That’s about the level of defects you’d find in things like GE’s Lightspeed MRI scanner.
Defects that can cause death are held to an even higher standard, but let’s be clear about something: It is perfectly acceptable and reasonable for an engineer to do a safety analysis on something and say, “x% of people may be killed due to defects.” This is not negligent. It is an acceptance of reality. Humans are not perfect, and neither are products or systems. Ford knows that x% of their cars will suffer mechanical failures, and some of those failures will result in death. It can not eliminate them - only reduce them. If Ford tried to build a car that was guaranteed to never kill anyone, no one could afford it and it would weigh 10,000 lbs. If it could be done at all.
So some amount of risk is acceptable. In the case of hot coffee, the worst-case scenario is what happened - 3rd degree burns over a limited part of the body. Not death. And Six Sigma is the ‘gold standard’ of process control and defect management.
7000 burns out of 10 billion servings works out to be 6.33 sigma. If it’s only 700, it’s 6.77 sigma. That’s way better than the industry standard.
So why is McDonalds at fault again?
And let me repeat: If the coffee were as dangerous as the plaintiff’s lawyers and some of you are suggesting, and it’s served to people in moving vehicles, then the number of injuries should be WAY, WAY higher. Lots of people spill beverages in their cars. I myself have spilled coffee and pop on myself.
Let’s assume that there is one coffee spill for every 5000 servings, which sounds low to me. If that were the case, that would be 200 spills per million servings, or 2 MILLION spills per 10 billion servings. And yet, of 2 million spills, only 700 went to court? It would seem that it takes just the right combination of very rare circumstances for McDonald’s coffee to do the damage it did to that lady.
Explain that, please. If a properly-done failure analysis shows a risk that is well within the limits of accepted industry practices, just why should the company be liable for the unlucky ones? Whatever happened to having to prove, you know, things like negligence?
No, that makes them wrong. It’s a failure of lay juries to understand science and engineering. This is a problem. It has been acknowledged as a potential problem in legal journals. Cases like this, where the jury is not competent to weigh scientific and/or engineering matters, usually boils down to who’s expert is more believable. That turns a jury trial into a battle of competing witnesses, not to see who can produce the best set of facts (which the jury isn’t competent to judge), but to see who can sway the jury through faulty analogy, simplification, and appeals to emotion.
The system needs fixing.

Yes, we are talking about 3rd degree burns. I have no idea what YOU are talking about, and I don’t know where you are getting the figures of 3 seconds and 1 second.
