What driving data reveals about your most serious crashes
Not every crash carries the same cost. Severe and major crashes cause the greatest harm – to people, to society, and to organizations. Independent validation found that our AI is particularly effective at identifying the drivers involved in these crashes. In fact, the highest-risk group is 91 times more likely to be involved in a crash linked to a severe or major claim than the lowest-risk group. Here, Anders Lindelöf explains exactly what that means.
Can the drivers most likely to be involved in serious crashes be identified?
That’s exactly what this figure confirms. When we compare the group of drivers identified by our AI as highest risk against the group identified as lowest risk, the highest-risk group is 91 times more likely to be involved in a crash linked to a severe or major claim.
What does it mean that the highest-risk drivers are 91x more likely to be involved in a severe crash?
I’ve used an archery analogy before, and I’ll use it again here. An archer’s aim tells you less than their pattern does. If a person’s shots are mainly off target, yet landing around the same place every time, that’s likely to be a simple thing to explain and fix. If a person has shots all over, it’s more difficult to understand why.
The same logic applies when we look at crashes – particularly at severity. When you isolate the crashes with the most serious outcomes, the drivers with the least predictable, riskiest behavioral patterns stand out even more sharply. This is why the crash probability gap between the highest-risk drivers and lowest-risk drivers widens further, to 91 times.
Can you explain the difference between Greater Than’s two key AI validation findings: 25x and 91x?
When looking at crashes of all severity, the group of drivers identified as highest risk by our AI experiences crashes at 25 times the rate of those identified as lowest risk.
When looking only at the most severe crashes, the gap between those identified as highest risk and lowest risk widens. This is because the drivers linked to severe and major claims specifically make up a smaller, more concentrated group than the drivers linked to crashes overall. The group of drivers identified by our AI as highest risk are 91 times more likely to be involved in a severe crash than the group identified by the AI as lowest risk.
Does the 91x number matter more from a road safety perspective?
Severe and major claims carry the greatest cost – in terms of harm as well as from a financial perspective. If an organization wants to focus resources where they will have the greatest impact, being aware of the group of drivers most likely to be involved in the most serious crashes is more likely to make a difference.
How was the 91x finding calculated?
Drivers within a population were ranked by their Crash Probability Score, and the highest-ranked 10% was compared against the crash rate of the lowest-ranked 10% – looking at crashes associated with severe and major claims only. That comparison is where the 91x figure comes from, independently reviewed as part of our AI validation.
Does 91x mean these drivers are guaranteed to be in a severe crash?
No. The 91x figure describes a group, not single drivers. It doesn’t guarantee that any individual driver in the highest-risk group will be involved in a severe crash, and it doesn’t guarantee that any individual driver in the lowest-risk group won’t be.
Should organizations only focus on preventing severe crashes?
All crash prevention matters. But training is most effective when it is targeted. Also, when resources are limited, these insights show where prevention will have the greatest impact.
FAQs
In the context of Greater Than’s AI validation, what does the 91x finding mean?
The group of drivers identified as highest risk by our AI experiences crashes linked to a severe or major claim at 91 times the rate of those identified as lowest risk.
Why is 91x different from the 25x figure?
25x reflects crashes of any severity. 91x looks specifically at the most serious crashes where the gap between highest-risk and lowest-risk drivers is even wider.
Who validated this finding?
The 91x figure was confirmed through an independent technical validation of our AI, conducted by Anders Arpteg, Ph.D., an AI expert with no prior affiliation with Greater Than.
Read more about the validation.
