By HCR Staff | Sept. 6, 2018
We introduced the HCR concept last year — it’s essentially analogous to a Passer Rating for head coaches. Our theory is this: Like quarterbacks, head coaches can be graded game-to-game on the strategic soundness of their in-game decision-making.
After all, football is the most highly orchestrated team sport. If the head coach is making poor decisions, even a collection of the best players can underperform (e.g., Philadelphia Eagles circa 2011). Or, to torture another metaphor, the best musicians can sound awful if they have a poor conductor.
Ok, we’re finished with the metaphors. Promise.
As we were saying, we graded every coach in every game according to an algorithm — which has quantitative and qualitative elements — and each week every head coach was given an HCR score. And then we ranked all the head coaches 1-32 each week based on their cumulative score up to that point in the season; again, just like Passer Rating for QBs.
Then we became curious: Could we use our HCR data to making recommendations on Moneyline and Pointspread bets? And not use any other data whatsoever — no analysis of players, injury reports, home and away data, etc. Instead, this would rely solely on our grading data of each head coach’s in-game decision-making.
So we created a (proprietary) methodology and then utilized this methodology in an internal study.
We first collected HCR data for Weeks 1-8 of the 2017 regular season. Why? Because HCR was a brand new concept; we needed what we felt was a reasonably solid baseline set of data.
After we collected eight weeks of data, we then had the data recommend Moneyline and ATS bets for the final weeks of the regular season.
The results were quite astonishing for Weeks 9-17:
• Moneyline: 68 percent winners
• ATS: 66 percent winners
Up to Week 15, the results were more surprising, and very predictive. The results for Weeks 9-15 were:
• Moneyline: 69 percent winners
• ATS: 71 percent winners
Only in Week 17 did our methodology dip below a 50-percent win percentage. We suspect that our win percentage dipped that week because many teams have either clinched playoff spots (and perhaps didn’t play starters in final game) or teams were out of playoff contention (and perhaps coaches who knew they may be terminated had less success regulating player effort).
This was an interesting and fun exercise. Again, we did nothing other than use HCR data to recommend the play to us — we did not inject any opinion into the equation whatsoever.
As we start the 2018 season, we will continue this exercise, and — beginning with Week 3 — we will actually begin making our recommendations here on the site. We’ll indicate the recommendations from strongest to weakest, and let the data — not human opinion — guide us. Keep in mind, however, we’ll be starting this exercise with only two weeks of baseline data for the 2018 season; last year, we had eight weeks of baseline data.
Either way, we hope you’ll enjoy the content and we are curious to see if our methodology holds up similar to 2017.