GT7 Daily Races, Time Trials and live leaderboard stats

Great to see the Hyundai dominance is over, however the driving quality for me has been sh*t. these people are driving me closer and closer to quitting the game outright, I have made many attempts to drive clean but all I get is crashes and severely dirty driving. A lot of my crashes don't even come under my own error. These B rank drivers are the equivalent of Nikita mazepin. Lost thousands of points to these morons.

Would someone happen to know the mathematical calculation/algorithm behind the DG Edge Score? I know there's a page that describes how it works but it often confuses me from a math standpoint.

Example:

I've completed 99 events with an average delta of 0.68%.

I proceed to set a new event time (the 100th event) with a 0.76% delta and it says my average delta will go up to 0.69%.

Simultaneously, I proceeded to set a new event time (101st event) with a 0.49% delta that said it will lower my average to 0.66%. A day later my delta on this event was 0.53% and now my delta is only projected to go down to 0.67%.

Therefore, these events nullify one another and my average delta will stay at 0.68%.

My question:

How do such small deviations in either direction for the 100th and 101st events affect my average delta by that much? (Forgive me in advance if I'm misunderstanding the statistical formula behind this.)