There’s clearly a whole lot more strategy involved with setting a daily fantasy lineup than working on your season-long teams simply because of the increased pool of available players. Even with restrictions on which players you can start in season-long leagues, though, there are little tricks you can use to get the most out of your lineup.
The most important step I believe people skip when setting their season-long fantasy football lineups is to simply ask “What do I want to get out of this lineup?” The obvious answer is “Uh, yeah, hi Jon, I’d like lots of points.”
Yes, we all want lots of points. But that doesn’t mean that your sole aim should always be point-maximization. What if the result of increasing your lineup’s ceiling is to dramatically decrease its floor? Would you rather have a team that can score 140 points 70 percent of the time, or one with more upside (greater access to lots of points) that reaches 140 points just half of the time?
It’s kind of like how teams should be calling plays in the NFL. Yeah, most of the time they just want to score as many points as possible. But in some situations (say, up 14 points with five minutes to go), the goal is no longer point-maximization; the ultimate purpose for both NFL teams and fantasy owners isn’t to score points at all costs, but rather to win games.
Sometimes we should shoot for the moon. Sometimes we should take the sure thing. It all depends on your goals. Your opponent will have a lot to do with your lineup decisions. Playing your league’s clown who might not even start a defense this week? It’s all about risk-minimization, baby. The guys who might not have elite upside (but will provide a set floor of production) are of increased value to you.
Much of this sort of approach to fantasy football is about thinking in terms of probabilities. Maybe you have Aaron Rodgers and Colin Kaepernick projected equally this week. Which one is the right guy to start? Well, maybe you have reason to believe Kaepernick is a more high-variance play—someone with a wider range of possible outcomes than Rodgers—in which case he’d be the right man for the job in certain situations (when you believe you’re an underdog) and the wrong choice other times (if you’re the favorite or otherwise want to minimize variance).
So what we’re trying to do is figure out our goals, then make sure every possible decision is a reflection of those goals.
Projecting Players
The second thing we need to do when making lineup decisions is to project the players in one way or another. I personally start with median projections—not a floor, not a ceiling, but just an idea of how many points I believe each player will score in the majority of cases.
I go about this a few different ways, but the foundation is aggregating expert opinions. I take projections from a number of fantasy sources that I trust (including my own personal projections), then average them together.
This concept is known as the “wisdom of the crowd,” and it’s based on the premise that the average of independently generated expert opinions is going to beat out each individual opinion in isolation over the long run.
So let’s say that you project a player for 15.0 points this week. Chances are it isn’t a perfect prediction and it’s off in one direction or the other. Now say that Rotoworld has the same player projected at 20.0 points. With that new information, how you should alter your own projection? You could potentially weight each source based on past accuracy (which is something I do), but if you trust Rotoworld as much as you trust your own projection, your new projection should be 17.5 points. Over the long run, players in comparable situations are likely to score closer to 17.5 points than either 15.0 points or 20.0 points.
Once I have a foundation of aggregate projections, I tweak them based on areas where I believe there might be market inefficiencies, i.e. where everyone could be wrong. I believe most people don’t include Vegas opinions enough (or at all) in their projections, so that’s a big source of knowledge for me. If Vegas likes the Chiefs to score 31 points and places Jamaal Charles at -200 to score a touchdown, that’s important information to me.
In most situations, the aggregate projections are all I use for my season-long lineups. I’m not always trying to go super high-variance—in most cases, I just want players who have access to a certain level of points a high percentage of the time—so I’m good to go.
Again, though, there are times when either risk-minimization or seeking upside are the name of the game. I use things like player types to sort players into buckets based on variance—so I know a player like http://www.nbcsports.com/edge/player/NFL/5698/Antonio-BrownAntonio Brown who sees a lot of short targets is more consistent than someone like Vincent Jackson, who is super high-variance—but I also use the rotoViz GLSP apps to visualize potential results.
The apps look at comparable players versus similar defenses from past games, giving you a range of outcomes for each player. The app helps to visualize ceiling and floor projections for players in a given week, and thinking about player projections in this probabilistic manner is valuable in and of itself.
Creating Relationships
For the most part, you’ll be starting the best players—how you define “best” might be different one week to the next based on your goals—but there are other times when that might not be the best course of action. Specifically, there are situations when you should be looking to create dependent relationships within your lineup that can help you achieve your goals.
Dependent relationships occur when players are participating in the same game. A running back’s production on Team X is negatively correlated to the production of the running back on the opposite team, for example; because of game flow, they’re less likely to both have two huge games than as a pair of comparable backs who are playing in different games.
Again, you should generally be starting the best options, but definitely consider specific relationships whenever things are close. Some of these include:
- Starting a quarterback on the same team as your opponent’s best wide receiver (hedge)
- Starting a quarterback with one of his own wide receivers (risk-seeking)
- Starting a running back on the same team as your defense (risk-seeking)
- Starting a returner on the same team as your defense (upside)
When you start a running back and the same team defense, for example, you’re generally going to have production that is connected; if the defense is playing well, the team is probably winning, which means the running back should see heavier usage. If the defense stinks it up, the team is probably losing, so the running back’s workload is probably down. Thus, a running back/defense combination is risk-seeking, increasing your lineup’s ceiling and lowering the floor (good when you’re an underdog).
One of the underrated pairings out there—one that doesn’t really even reduce your floor, but creates a higher ceiling “for free”—is using a returner with his defense. If you start Cordarrelle Patterson and the Vikings D/ST, for example, you can “double up” on points with a return touchdown.
Look out for these sorts of relationships and consider how they might affect your lineup’s access to certain ranges of outcomes.
Variance of the Flex
Finally, I want to mention that the way you approach your flex position(s) should also be a reflection of your goals. That’s because different positions have much different levels of consistency from week to week.
When you’re a heavy favorite, it makes sense to play a running back in the flex. That actually won’t maximize your projected points in PPR leagues—normally wide receivers will do that—but running backs are more consistent than receivers on a weekly basis, so it will provide you with a higher floor. If the only way you believe you can lose is if your team tanks, you want to do everything possible to avoid that, and using a running back in the flex can help you achieve that goal.
Certain player types are naturally more volatile on a weekly basis, too. Like I mentioned, the average depth of a receiver’s targets has a lot to do with his consistency. At the running back position, those who catch passes have higher floors because they aren’t so dependent on specific game scripts.
The point is to consider what you need out of your fantasy team in a given week, then structure your lineup in such a way that you maximize the probability of achieving that goal.
