I read a really interesting article over at Football Perspective on game scripts—the nature by which a football game unfolds. I think predicting game scripts is extremely important and a massively underutilized way to project players. To accurately predict fantasy production, we need to understand how the flow of a particular game will affect the total number of passing and rushing attempts for each team/player.
Game flow is a difficult thing to predict because different sorts of teams match up differently; there are a lot of variables involved. Maybe Team X is run-oriented, but they become extremely pass-heavy when they get down in games. How should we project their players if they enter a contest as a heavy underdog?
Chase Stuart at Football Perspective has a formula to calculate each team’s true “pass identity,” but I think there’s probably a simpler way to predict the effects of game flow: using Vegas lines and prop bets.
Here at Fantasy Labs, we obviously believe Vegas data can be of massive importance to daily fantasy players. Vegas’s lines and prop bets are simply their best guess as to the number of points, attempts, yards, touchdowns, catches, or whatever that a team/player will accumulate during a game.
We are working on implementing prop bets into our NFL product. There are three reasons I like using prop bets, in particular, when projecting players.
- They’re accurate.
Vegas has millions of dollars on the line. If they post a shitty line, it will get exploited. They have a strong incentive to make sure they’re right.
- They’re accurate.
Now wait, what? Number two is the same as number one, but for a different reason. Prop bets make for accurate predictions not only because the sharps and algorithms used to set them are some of the best in the world, but also because the props get molded—and, for the most part, more accurate—as people bet on them.
If Vegas puts out a prop bet of 280.5 passing yards for Robert Griffin III in a given game and the majority of people view that number as being much too high, they’ll bet the under and Vegas will move the line, bringing it down to reflect those bets.
Although public bets aren’t always smart, there’s still a “wisdom of the crowd” effect by which lines generally (but not always) become more accurate as time proceeds. That’s why so many professional bettors try to bet lines as soon as they come out—before they’re altered by public betting trends.
- They’re efficient.
Let’s say that you’ve studied the accuracy of prop bets and you think you can beat them by a couple of yards per game by importing past game data and creating a projection model. Well, chances are you can’t actually do that, but even if you could, is that really the best use of your time? I’m not asking that in some dick-ish why-are-you-playing-fantasy-football-so-much-you-loser sort of way because that sort of stuff is literally all I do, but I just mean there are more effective ways than trying to beat Vegas for you to become a better daily fantasy player.
In the end, using game props is so simple, so efficient, and so accurate.
Prop Bets and Game Flow
Prop bets generally come out on Sunday mornings, and they look something like this:
So you check the prop bets and see that Russell Wilson is projected with a total of 26.5 passing attempts for the week. Now, how should you alter that projection to account for things like game flow, opponent strength, recent efficiency, and so on?
Answer: you shouldn’t. Prop bets already account for all of that stuff and more, so you don’t really need to change them at all—hence why I said they’re an incredibly efficient way to do fantasy football business.
If there’s one situation in which altering the prop bet is recommended, it’s when there’s a late shift in weather. If Wilson is at 26.5 attempts and there are reports of heavy snow where he’s playing, then obviously that projection needs to be altered.
Otherwise, we’ll just continue to ride Vegas’s coattails to daily fantasy football dominance.