I just finished reading The Logic of Sports Betting, and one of central themes was this concept of finding markets that haven’t been “picked through” or even derivatives (first half ATS, etc). That is, not playing big markets that haven’t been bet up/down early on (ATS NFL at big books, etc), and have been priced to the point you never have a positive EV bet. The author used the analogy of the $99 dollar TVs sold at Best Buy on black Friday (the reality is it’s just a hook, and likely only 10 TV sets are priced as such, the rest come with no discount). Betting on NFL Money lines or spreads isn’t a bad idea just because so much action eats up most of the opening line’s value, but because most of the traditional bets already priced in angles, injuries and anything else you believe (in reality or otherwise) impact the outcome. But, smaller volume markets (relatively speaking) or derivatives, may not be as picked through by bettors and/or as staunchly priced by books.
RULE: Any information you have on a game, that could impact on outcome, ON ITS OWN is worthless, UNLESS, the market/price makers are not aware (at least not aware enough relative to the current price).
With that said, we give you “Empirically proved, year to year consistent First Half NFL Fast/Slow TRUE trends”:
Remember, the goal is always to find information that is predictive, gives you an edge over the market and has not been priced in by the books. Hence, this fall we will be redistributing our typical exposure to NFL money lines and ATS to 1H bets. Below is our playbook, for roughly 10-25% of our weekly exposure.
RULE: Angles, such as Team A is 10-7 ATS after playing a division opponent at home are nonsense. For an angle/trend to be of value, it must be predictive (just because Team A is 10-0 when its a full moon next week means nothing), logical (When winds are above 25 MPH, games go 50-30 Under the total, since scoring may be more difficult) and not baked into the price.
Since we didn’t want to go into this play blindly we came up with a system for 2020, the First Half (1H) Angle for 2020 NFL:
- If you are unfamiliar with DVOA by Football Outsiders, in a nutshell, its an aggregate performance metric that teases out variance, adjusts for strength of opponent and focuses on what truly matters to winning games (first downs and scoring) and puts it in numerical form.
- We wanted to find teams that perform far better in the first half than the second using relative DVOA comparisons. DVOA being a far better predictor of NFL games than even scores over the long run. That is
- Offenses who’s 1H DVOA far outweighed its 2H DVOA
- Offenses that did the opposite
- Defenses that did the same as “a”
- Defenses that did the same as “b”
- Once we found these teams, we wanted to tease out any variance, luck or lack of consistency in this very specific measure (remember, we want to find a predictable trend), so…
- Knowing that year to year averages/consistency will likely be a stronger indicator of next year, we took the average of both 2018 and 2019 seasons
- Then ONLY looked at teams that are consistently in the top 9, bottom 25 each year
- Ignore teams that had a significant QB, OC or DC change this year (as that would impact the likelihood of a team repeating their performance in this metric in 2020)
- And here is what we found (the work): https://docs.google.com/spreadsheets/d/1za7sxmwPmy4Bmo40TstAeMMb39cfJO8c0-dQdoKSi2g/edit?usp=sharing
The findings/plays for 2020 can be summed up here:
|OFF + 1H||OFF – 1H|
|DEF + 1H||DEF – 1H|
- TEN to lose 1H (or ATS)
- MIA to win 1H (or ATS)
- CIN OVER 1H
- GB plays blue, take green 1H
- Red plays purple, take Red 1H ATS
- If ANY of the same color plays each other, OVER 1H
- Green v Green OVER 1H
- Red v Red UNDER 1H
In layman’s terms, I am using a ranking system that teases out variances, and predicts outcomes better than average, OVER THE MARKET. Funnel that into a betting market that is less likely to already price this in (if they are even aware of it themselves), while shielding myself from “early in the week” discounts being erased. Let me know how you do, good luck!