Time of Day driving stats are not accurate

The Time of Day statistics have been confusing me.

The 'official' statistics on the Allstate website do not align with my calculations even though I am using the times and distances recorded by Allstate. I have double checked and triple checked my driving stats and then checked my father's stats and found the same problems (errors).

The error is not to my benefit as the error skews towards higher risk driving times.

The easiest check is the one driving event that I had during a "high risk" period between 11:00 pm and midnight.  I drove 9.0 km and at present I have a total distance driven of 1,006.8 km.

Calculation:    9.0 divided by 1,006.8 = 0.0089  or 0.89%
The absolute difference is  0.04%
The percentage different is 0.04%/0.89% = 4.5% error (not to my benefit)

If the discount is based on calculations of all our driving statistics how can we trust the hidden proprietary calculations if simple errors such as this are clearly evident?

Another oddity is that sometimes updates are made in real time and other times changes happen after days of no driving.

The program makes no sense and is confusing.


Same problem on my father's stats

Update:
Either the DriveWise program updates the Time of Day stats only periodically (not all the time after each trip) or there was a problem with the calculations.  I submitted an email to the DriveWise support team pointing out the errors and the next day the online numbers aligned with my own calculations but the anonymous support elf told me there are no problems with the system.

Update 2:
A DriveWise customer support elf told me that some of my calculations are off due to a problem in the FAQs.  Supposedly the FAQs were wrong about what time the high, medium and low risk period start. 

Old FAQ
Under this FAQ driving at 11:00 pm seems to fall under the LOW risk period but the system actually calculates it as HIGH risk.

 

New FAQ
The deleted the :01 from the start time of each risk time
High risk now starts at 11 pm

 

Even these new FAQs are technically wrong because according to this schedule 11:00pm is now both low and high risk according to this FAQ.

If they wanted to be proper it would state that LOW risk is from 12:00:01 pm to 11:59:59 pm