This article is part of our NFL Picks series.
If you love sports betting and NFL Odds, VSiN is a great site to look at. One of the features they try to highlight in their subscription products is their "Betting Splits." Specifically, this is data from DraftKings that shows the percentage of both the actual bets on specific games as well as the overall dollar value of the bets -- the "handle" -- on each side.
There are some basic theories on this sort of data. As bettors we want to side with the bigger players -- "sharps" -- and against the small players, or "squares." In a vacuum, the handle is a better reflection of sharp money than the quantity of bets. If there is a noticeable disparity between the two, then going with the handle should work, right?
Well, it is not that simple. First of all, it is important to take in the whole picture. Did a line move "because" of the handle? In other words, is the price on the side of the handle now less favorable than it was when all those bets came in?
Secondly, who is the handle to begin with? In VSiN's Betting Splits case it is DraftKings, perhaps the most "public" book in America. They were the first online sportsbook to open up in New Jersey, and likely among the first in many states. The handle itself may collectively represent something to fade, not tail.
VSiN's Steve Makinen ran numbers based on the DraftKings betting data from 2022, and below is what he found.
These were the overall records of the majority bettors for the 2022 NFL season, using the key terms HANDLE and number of BETS.
- Majority HANDLE on point spreads: 123-153 ATS (44.6%)
- Majority number of BETS on point spreads: 119-155 ATS (43.4%)
- Majority HANDLE on totals: 126-150 (45.7%)
- Majority number of BETS on totals: 132-148 (47.1%)
The "crowd" was generally a fade, whether you looked at the quantity of bets or the handle. Makinen goes on to develop 12 betting systems based on the betting splits, including the following one.
When more than 65 percent of the HANDLE was on a particular side of an ATS wager, this majority group was just 35-53 ATS (39.8 percent). The number for this "supermajority" in college football was closer to 80 percent. In other words, if you see two-thirds of the handle backing one side in an NFL game this season, it's best to fade it.
Right here, right now, six Week 1 games fit this criteria as per the VSiN numbers.
Texans +10 vs 
Ravens (18%)
Colts +3.5 vs 
Jaguars (18%) 
Cardinals +6 vs 
Commanders (28%)
Panthers +3 vs 
Falcons (22%)
Patriots +4 vs 
Eagles (22%)
Jets +1.5 vs 
Bills (32%)
The Jets make the cut now, but as the game nears and betting limits and overall activity goes up, that could change.
One commonality here is that they are all underdogs, which I find attractive in general, and in Week 1 in particular as I laid out in this article I wrote a couple weeks ago. In fact three of these teams overlap -- the Texans, Cards and Panthers. Yes, they all look like bad teams! In the case of the Cardinals, they look like an epically horrible team that on an organizational level is mailing it in for 2023. It is very tough to click on any of them.
Over time, fading the "Joes" tends to work, but not in all circumstances. The public is not always wrong! An interesting finding in the data is that the DraftKings crowd actually does well -- or at least breaks even -- when they buck typical tendencies and side with the under or the road underdog.
