The only problem with this chart is that it was prepared either by or for an idiot...

1. It is meaningless as "head injury" defined.

2. There is no guarantee that the studies used the same definition. This allows the results to be anything the author likes according to the studies that he selects.

3. A broad definition of injury treats all injuries as the same. If a helmet offers you a 50% chance of reducing minor injury at the cost of a 50% chance of making a severe injury worse, would you wear it? Such a helmet would look great in these charts - because trivial injuries are much common. But do 10 prevented minor concussions really outweigh 1 death? And this isn't just an academic point: there is a lot of evidence that helmets make rotational damage, the bigger killer of cyclists, worse.