In the following article, we will:
- Define “fading the public” as wagering contrary to public consensus
- Prove that fading the public alone has a success rate of 52% (or greater)
- Understand that people LOVE to bet favorites or newsworthy teams
- Conclude that the public consensus metric should be one of many data points a good handicapper looks at
As an avid sports gambler, I’ve always tried to look for trends that give me an edge against the oddsmakers. I’ve spent a great deal of time amassing data from various sources and using that data to determine which way to wager. For me, there is a certain thrill to having knowledge that the typical sports bettor lacks. Throughout the years, I’ve noticed that there are certain tenets that gamblers like to spout off as fact. However, if there is one single axiom that gets repeated over and over again it is none other than the phrase, “fade the public.” If you’ve ever wagered on sports for an extended period of time you’ve heard the phrase uttered ad nauseum. But just what does it mean and how well does it hold up to statistical analysis?
What Exactly Does “Fade the Public” Mean?
“Fade the public” simply put, means betting contrary to the public majority. The logic is that the average bettor is less knowledgeable than the oddsmakers and, as such, will come out a loser more often than not. Look no further than casino profits off of sports wagering in 2017 (upwards of $250 M) as an example.
Fade the Public: By the Numbers
But just how true is the concept of “fade the public?” Let’s dive in and take a look at some statistical analysis from 2017 to see if this concept really holds true. In the NFL, I was able to gather public consensus picks and percentages in the regular season from November 19th, 2017 forward. Now the obvious caveat: this is clearly a limited sample size as it is only taking in the final 6 weeks of the regular season (this is when I began tracking the data), however, it does give us 86 games, a good number of occurrences to test the theory of fading the public. Let’s focus in on the first set of data in the given sample, the games played on November 19th:
Consider the data represented above. For the first game shown (Detroit vs. Chicago), the .67 consensus bet percentage means that 67% of the public bet on Detroit to cover a 3-point spread against Chicago. For this particular game, the result of that wager was a push (Detroit went on to win the game 27-24). If we now take a look at all games played from November 19th through the end of the regular season, we find that wagering on the public consensus would net you a record of 38-44-7 (48.1% hit rate). In this scenario if you just simply took the public consensus and bet against it (faded the public), you would already be coming out with 51.9% success rate. If we start to try and remove outliers, for example, now weeding out any games where the public consensus is under 52%, this success rate increases to 52.4%. If we increase the threshold to 55% (only wagering and fading games where public consensus is greater than 55%), we get even better results, with the success rate going all the way up to 66.7%. Obviously the greater threshold that we use, the smaller the sample size becomes. However, I think we can start to see that using the public consensus percentage for sports wagering can be a powerful metric.
This basic logic doesn’t just apply to the NFL however. Let’s take a look at the same metrics for the NBA. I started pulling the public consensus metric for the NBA on February 10th, 2018 up to March 17th, 2018 (my focus was on NFL/NCAA football prior to these dates). Again the caveat is that this is not a full season’s worth of metrics, but this gives us a whopping 192 games to test out the theory. If I apply the same logic as I did the NFL, the results are as follows:
Merely fading the public picks in the NBA for the time period in question would net me a 55% win percentage and the hit rate only increases as you start to increase the public consensus threshold.
Why Does This Work?
So now that begs the question, “Why does this work?” From a psychological perspective, the casual better typically prefers favorites rather than dogs, especially if there is a lot of news trending around a team. The ESPN network has become a hotbed for sports news that caters to the casual fan and many of these fans get their betting advice solely from these outlets. There is nothing I love more than fluff pieces from ESPN touting some hyperbole like, “greatest team in history” or some ridiculous headline to capture the attention of the inexperienced bettor. The first anecdote that pops into my head is the 2006 NCAA National Championship between the Texas Longhorns and the USC Trojans. Going into the matchup, USC had been hyped up to the point they were calling Matt Leinart, Reggie Bush, Lendale White and company one of the 3 greatest college teams in history. Even the eyeball test could tell you the USC team of 2006 would be hard-pressed to defeat the USC team of 2005. Sure enough, the public consensus followed suit. And the rest, they say, is history.
So wat have we learned? Fading the public and using the public consensus metric is a good start to intelligent betting. Now would I expect the same results if we ran the test against an entire season of data? Absolutely not. The strategy is not perfect (again note the limited sample size embedded in the numbers above) and assuming a 5-15% interest rate that oddsmakers charge for their services, a 55% hit rate probably gets you a net loss overall. And let’s not even get started on public consensus in the Super Bowl (I will save this topic for another day). But any experienced handicapper or professional sports gambler always has his/her eye on that public consensus metric. They use that metric in conjunction with a handful of other statistics to find value in a spread. If you take anything from this article take this, if everybody you know is leaning one way wagering on a game, there’s a good chance it’s the wrong bet.
Sports Data Intelligence has been collecting data since 2017 and has been applying statistical models and advanced metrics to the world of sports gambling. We currently apply and take up positions specifically in the NBA and the NFL, and are in the process of doing back-testing for other arenas such as NCAAF and NCAAM.