QBR huge correlation to wins

So I have always liked QBR as a way to represent QB production over other methods. I just didn't know that it had such a strong correlation to wins. Looking over the first 3 weeks of the season, a team has at least an 80% chance of victory when their QBR is in the top 10. On the flip side, a team has over a 70% chance at a loss if they are in the bottom 10. 

The outliers are usually super close games. Purdy had the #6 QBR and lost 27-24 this week. In Week 2, Hurts had the #3 QBR and lost 22-21, Burrow lost 26-25, and Jones lost 21-18. 3 losses by 5 pts. In Week 1, Richardson had the #1 QBR and lost 29-27, and Murray had the #4 QBR and lost 34-28. 

On the low end, the wins came only due to the defenses. In Week 1, Caleb had the 4th lowest QBR, but the Bears still won. All the wins by low QBR QBs came when the defense allowed 20 or less except one game, KC's Week 2 win against CIN 26-25 where Mahomes had the 4th lowest QBR and still won due to the refs. 

So for Caleb going forward, he needs to improve himself and the good thing is, he has. He had a 23.4, then 20.6, and jumped to a 36.1 in Week 3. If he can keep that trend going and get over the 50.0 mark (average QB), the Bears will be looking good after the bye. However, he needs to make improvements enough to get out of the bottom 10 which usually is a loss indicator.