For a wiki reference for any terminology that is unfamiliar, refer to this article.

3 profitable weeks in a row, the model is on a roll as expected, with the first truly profitable (as in significant return since the other two weeks were only +~.75U) week. The better part of that story is that the ROI continues to increase every week, also as expected.

As a sidenote, I wanted to discuss the “luck factor” briefly. I think it’s important to note that luck finally went the model’s way this week. The NFL is and always will be incredibly unpredictable in a sense, which is why it draws such high crowds every damn Sunday; but when modeling these games it is important to remember that luck will always even out. There was a lot of unlucky picks going against the model over the first few weeks (IND, CIN kickers missing game-winning field goals, BAL collapse vs MIA, CLE collapse vs NYJ, etc.), but this week the AZ backup kicker missing a game-tyer and Carson Wentz throwing an interception at the goal-line really helped carry the model into it’s first truly profitable week. So, one can say it was lucky this week, but then you need to consider the previous weekly unlucky.

Week 5 Results

total units bet: 20.37

Total units won/(lost): +2.67500

ROIv1: 13.12%

ROIv2: 2.67%

YTD Results

total units bet: 96.51

Total units won/(lost): -2.33

ROIv1: -2.4%

ROIv2: -2.1%

Remember, this YTD is really irrelevant since I mentioned many times the model should only be used after Week 4/5. This is just for transparency purposes.

YTD Since Week 4 Results

total units bet: 64.48

total units won/(lost): 3.35

ROIv1: 5.2%

ROIv2: 3.38%