Originally Posted by
livedarklions
But the same VT group says that they recommend any 4 or 5 star helmet on their list, which includes a lot of variation in the relative numbers, and state that cost, comfort and fit are also important factors.
I don't follow your point. The quantitative comparison I highlighted (200%) was the difference between the best and worst helmets that they tested (i.e., those that were given 2 stars), and was a response to Maelochs claim that there's only a few percent difference between the best helmets on the market and the worst. When challenged on this point, he doesn't defend it, but rather returns to his point that in a serious crash, we don't know if a "better" helmet really makes any difference. I didn't say that there are large differences among the better helmets.
Look guys, science isn't perfect and as a laboratory scientist myself (though not in anything close to this field), I am well aware that the lab and what you call the "real world" (a misnomer IMHO, all of it is "real", but the controlled variables are not all the same) have differences. Science is an imperfect approximation that tries to move towards an improved understanding. I'll go with that rather than "we don't really know anything perfectly, so I'll dismiss the insight we might be able to get from what we can measure." To that end, I'll seek out one of the "better" helmets with the understanding that that an incremental improvement of my odds is better than hearsay.