How to Increase Upside on DraftKings
Exponential growth, which we all know is just so hot right now, isn’t really sought in fantasy football because, in most cases, fantasy points are accumulated in a linear way. Below, the line that has no curve represents linear growth, while the line that starts slowly and ultimately ends up on top represents exponential growth.
If you have a retirement plan, you’re hoping to maintain it long enough that you can continue to see exponential growth through compounded interest. In such a situation, your interest grows very slowly at first and very rapidly as your plan matures. Here’s a very cool visualization of exponential growth.
So can we generate a snowball effect in daily fantasy football? I think there are a few situations that either represent exponential growth or are at least similar in nature.
Red Zone Studs
The most obvious example of exponential growth in fantasy points comes with touchdowns. As offenses approach the goal line, their probability of scoring a touchdown increases in a non-linear way. Per Advanced Football Analytics:
This resembles the typical exponential growth curve, huh? Offenses have very little shot of scoring a touchdown from their own territory; 50-plus-yard touchdowns are very rare. As an offense approaches the opponent’s 30-yard line, though, things start to look up. We see steady growth until the red zone, when exponential growth occurs.
The conclusion here is obvious: target players who excel in the red zone. I’ve written extensively on this subject in regards to every position, and there are so many reasons that you need to be concerned about scoring. For one, touchdowns are clearly important. Too important, actually, as they’re weighted too heavily in fantasy football. Let’s exploit that.
Second, red zone play is consistent. It’s fairly easy to predict both opportunities and efficiency; the same sorts of players excel again and again in the red zone. For pass-catchers, weight is extremely valuable (even more so than height):
Being tall helps receivers in the red zone, but if you think about it, not many passes are jump balls on which being tall and leaping high will be that big of an advantage. But how many times do receivers and tight ends need to use their bodies to fight off defenders? Weight and strength matter for receivers in the red zone.
As a final note on the importance of touchdowns and red zone play, consider fantasy scoring on a typical drive. Fantasy point scoring itself is linear, up to a point. If you have a running back who sees every carry on a 62-yard scoring drive, he might accumulate 6.1 fantasy points for rushing 61 yards, all the way up to the opponent’s one-yard line. If he takes that final play in for the score, we see an explosive growth in scoring that’s completely disproportionate; the running back totals the same 6.1 fantasy points for that one-yard touchdown as he did for the 61 prior rushing yards. We see linear and then explosive growth with fantasy points, and exponential growth in regards to the probability of scoring a touchdown.
Note that, even on a full PPR scoring site like DraftKings, I still want touchdown upside. Yes, high-volume receivers have more value than in standard formats and they’re certainly in play in cash games, but if the name of the game is upside, you’re going to need touchdowns. Even if a player like Wes Welker offers $/point value, I’m very unlikely to use him in a GPP because he doesn’t have much access to non-linear fantasy scoring, i.e. consistently scoring touchdowns.
Pairing Teammates
In my opinion, daily fantasy football is becoming less and less about individual players and their projections and more about position types and roster construction. Since projections are becoming more accurate every year, the biggest advantages for fantasy owners aren’t in making more accurate projections than the next guy, but in selecting the right types of players and pairing them in a way that can increase either upside or floor production.
Take pairing a quarterback and one or more of his wide receivers, for example. This is a popular strategy in large daily fantasy football leagues because you need upside, and pairing teammates together creates an opportunity for explosive jumps in fantasy production. Scoring seven fantasy points from your quarterback on a 75-yard touchdown pass is nice, but totaling 21.5 fantasy points on that same play because you had the receiver who caught the touchdown is even better.
The question is whether or not the upside makes up for the downside. In most cases, I’m not sure that’s the case. I have some data suggesting that stacking a quarterback with his receiver is riskier than you might think.
This graph analyzes the probability of either a No. 1 wide receiver or a No. 2 wide receiver turning in a top four/eight performance on the season (in terms of their individual games) given that their quarterback did the same.
So take a quarterback’s top four fantasy games; what are the chances his wide receiver also had a top four fantasy game? The answer is 40 percent for WR1s and 33 percent for WR2s. That’s above the 25 percent expectation we’d have if things were totally random.
But look at the graphs for the “bottom four” games. When a quarterback turns in one of his worst four performances of the year, there’s a 50 percent chance that his No. 1 wide receiver did the same and a 38 percent chance that his No. 2 wide receiver had one of his worst four games of the year—above the “top four” percentages. Things even out when we analyze “top eight” games, but in terms of real ceiling and floor production—the best and worst games of that player’s season—there’s a low floor on QB-WR pairs (especially for the No. 1 wide receiver).
Note that there are certain quarterback/wide receiver pairs that are probably less susceptible to wild swings in production. WR1s who have elite wide receivers opposite them, for example, typically 1) don’t see as much defensive attention and 2) don’t see as high of a number of targets as the Megatron-esque true WR1s. That means that the quarterback is less dependent on them for his production, and the two aren’t so closely linked.
Shootouts
In early October of 2013, Tony Romo turned in one of the best games of his career—and really one of the top games for any quarterback ever. On just 36 attempts, Romo threw for 506 yards, five touchdowns, and one pick. The fact that he was so efficient with over 500 passing yards is remarkable.
And he lost.
On the other side of the field stood Peyton Manning, whose 412 passing yards, four touchdowns, and one interception weren’t so shabby himself. Together, the duo passed for over 900 yards and nine touchdowns, each feeding off of the other’s performance throughout the game. The Cowboys and Broncos scored on nearly every possession, causing the coaches to continue to dial up pass after pass in an effort to keep up. The fantasy production in the legendary 51-48 contest was unheard of; three Dallas receivers had at least 100 yards and a touchdown, while five Denver receivers caught at least five passes.
One of the ways to enhance the odds of exponential fantasy production is to start players in games that are likely to become high-scoring shootouts with a close final score. The key isn’t only that there’s going to be a lot of points, but also that the game remains close so that each team keeps throwing. You can use the Vegas lines to easily and accurately predict final scores.
If possible, you want to target players facing an opponent that’s going to pass the ball a lot. One of the hidden benefits of seeing the opposition throw is that it increases the total number of plays in a game since the clock stops with incompletions. So if you have a quarterback in a game in which his team is projected to score 34 points and the opposition is projected at 28 points, that’s a whole lot better of a situation than one with a 34-14 projected score. In the latter scenario, there’s a good chance your quarterback will stop passing in the fourth quarter.
Keep in mind that the opponent’s offensive play-calling is important to your player’s production. Very important. All other things equal, you want to start guys facing teams that extend the game as much as possible and increase the total number of plays for both squads. If you’re starting a quarterback playing against a crappy team that runs the ball a lot, he might have good efficiency, but he’s probably not going to rack up a lot of attempts because 1) his team will gain a lead and 2) the other team’s play-calling will prohibit a large number of total plays in the game.
If the goal is exponential growth, we’re seeking lots of passing, lots of scoring, and a close contest.