I posted this in a different thread back in February, but I think it is worth revisiting now that books are releasing season win totals-
I have been a profitable recreational sports bettor the past three years, using a quantitative methodology based heavily in analytics and advanced stats. One of my most important tools in betting the regular season win market (and weekly sides for that matter) has been Bill Connelly's SP+ projections, located behind a paywall on ESPN+ or
here (for historical ratings, at least). I use a weighted combination of
different ranking systems for my weekly side projections. Through statistical analysis I've found that SP+ are most predictive for Regular Season Win totals, however.
Generally speaking, SP+ rankings are intended to predictive and forward facing, reflecting opponent-adjusted components of four of what Bill Connelly has deemed the Five Factors of college football: efficiency, explosiveness, field position, and finishing drives. (A fifth factor, turnovers, is informed marginally by sack rates, the only quality-based statistic that has a consistent relationship with turnover margins.)
Preseason SP+ projections are based on three primary factors, weighted by their predictiveness:
- Returning production - accounts for transfers and attrition. The combo of the previous year's SP+ and adjustments based on returning production make up almost half of the projection's formula.
- Recent recruiting, including transfers - makes up for ~ 1/3 of the formula
- Recent history - uses 2-4 years of info from previous seasons to account for overall program health. This is where any adjustments due to coaching changes are made. Makes up about 15% of the formula.
GT finished 2022 with an SP+ of
-11.8. The good news is that with the returning production, transfers, coaching change, etc our preseason rating for '23 is
-2.4. I was expecting our improvement to be around 7/8 points, so this was a pleasant surprise. I've linked the entire Preseason SP+ and Returning Production numbers
here for your reference.
You can then project out Regular Season Win totals by taking the difference in opponents SP+ ratings and factoring in Home Field Advantage to get a projected point spread. I have found that the avg HFA for the past few years is around 1.8; you can get super granular and assign different HFA factors for different teams, but I won't do that here, and GT's factor would be lower than 1.8. Sigh.
Once I get my projected point spread I then convert to the Money Line by using a calculator powered by a ton of historical data. Convert the Money Line to implied probability, then sum up the probability for the projected win totals.
Without any further adieu, here is our 2023 Opponent / Projected Point Spread / Implied Win %:
Louisville (neutral site) /
+12.5 / 18.05%
SC State /
-30 / 97.93%
@ Ole Miss /
+21 / 6.83%
Bowling Green /
-21.5 / 93.75%
UGA /
+32.5 / 0.53%
@ Clemson /
+24 / 4.64%
UNC /
+13.5 / 16.72%
@ Miami /
+13 / 17.39%
@ UVA /
-.5 / 50%
BC /
-2.5 / 54.13%
Syracuse /
+3 / 42.37%
@ Wake /
+8.5 / 25%
Projected Win Total of 4.27 wins.
I would expect the betting market to open up between 4.5 and 5.5 Regular Season Wins for GT. I think our floor is 2 wins and our ceiling is 6 wins.