Recently, while watching Antonio Brown’s entire career as a Patriot, someone asked, “How are you so good at FanDuel?”
Game theory!
If you’re new to daily fantasy sports (DFS) there is one key thing you must know. It’s not enough to just build a good lineup, you have to build a unique, good lineup. That’s where game theory comes in.
The original game theory concept addressed zero-sum games, in which each participant’s gains or losses are exactly balanced by those of the other participants. Basically, if someone else is doing bad, then you’re doing good. An example in fantasy football – you have a team’s QB and your opponent has the team’s RB. Every time the QB scores a touchdown, that’s good for you and bad for your opponent, because the QB took away the RB’s opportunity to score points.
But, in DFS you don’t know who your opponents will be playing, so you have to make assumptions. You have to try to figure out who the most popular plays each week will be. Sometimes it’s easy… when the New England defense is playing against the Miami Dolphins, you can reasonably predict that they’ll be the highest-owned defense. Similarly, when the Saints are playing the Rams one week after Christian McCaffrey went for 42 fantasy points against them, you can be sure Alvin Kamara will be the highest-owned RB. 40 to 50% of the people in tournaments may even have him! Most DFS players also like to chase points, meaning if a player does well one week, they’ll be highly owned the next week.
So, what do you do with this information? You want to avoid these players (usually) and instead target good players who just had a bad week. Sometimes the matchup is so good – or the price is so good – that you’ll go right along with everyone else and use that player. But, more often, we need to find a way to differentiate our lineups. So if we use the highest-owned player, or several highly-owned players, we’ll want to add some players that we think will have low ownership. This doesn’t mean we try to hit a homerun by playing a backup TE or 4th string WR; we still need to have players who have the potential to be a top-scoring player, but who just aren’t as popular.
Another way to differentiate your lineup is by not using that highly owned player at all. Instead, look for a player who has a similar projection – or, really, a similar chance of being the top scoring player at the position. Maybe there is a player who is slightly more expensive, so people are bypassing him for the cheaper option. Maybe there is a player who is facing a tough defense so people are shying away, but forgetting that the offensive player is still an elite player capable of scoring a lot of points. Think of an elite WR playing against a top cornerback or an elite RB playing against a good defensive line.
If you choose not to play the highly owned player you can also differentiate and, thinking about game theory, gain leverage on the field by picking a player that is competing with that player for points. In the example above, if half the tournament field will have Alvin Kamara you might instead choose Michael Thomas. If the Saints score their points through the air, then every time Michael Thomas scores that’s good for you and bad for all those people who have Alvin Kamara. Every time Michael Thomas scores you’re gaining a huge advantage over 40 to 50% of the field because you’re moving up the leaderboard while they stay put. You may also decide to avoid the Saints all together, hoping the game is more of a defensive battle with lots of field goals.
When you’re thinking about differentiating your lineups by playing less popular players, you should remember, it’s really the ownership percentage of your nine-player roster that matters, not just the ownership percentage of your individual players. If you have an RB who is owned by 40% of the players in the tournament, but you have a TE who is only owned by 1% of the players, that combination is still a pretty unique combination. But, if you have an RB who is only owned by 10% of the players in the tournament and that same TE, then you really have an advantage over the field. If that 40% RB is a bust and your 10% RB does well, then you’ve leap-frogged 40% of the people in the tournament. That leap-frogging is the secret to tournament success. You can’t expect to win a large tournament when your score is moving up and down right alongside thousands of other people.
Another way to differentiate your lineup is through roster construction. There are basic correlations – for example, pairing a QB and WR, because if the QB has a good game there is a high likelihood that the WR will also have a good game – but there are less popular correlations that you can also take advantage of and even anticorrelations. In our first game theory example, we noted that a QB and RB take opportunities away from each other, so they’re anticorrelated. Normally, you don’t want to pair a QB and RB from the same team, but if that team has a very high projected point total, then there is opportunity for both players to do well, giving you a unique lineup that other people will be afraid to use, because it’s not correlated.
You could also build a lineup that uses multiple correlations. Most people are getting pretty good at using a QB and WR together, but don’t always think about the other team. If that QB and WR pairing has a good game, that means they’re probably out to a big lead and the other team will be passing to try to catch up, so you should also pick a WR from the other team. Or, if the game script is different, maybe your QB and WR have a good game because they were the ones trying to catch up, so the other team will be running the ball more and you can pair that team’s RB with your QB and WR. This is called stacking a team or game.
If you really want to make a stand you can over-stack a team or an entire game. The idea here is that you might have a few highly owned players but you’ll differentiate your lineup by getting more players from one team or game than most people will pick If you think a game will be a shootout, you can start the QB, WR, and TE (or another WR) from one team and also the RB and WR from the other team. That’s just one example, but you can stack a game in many other ways. Try to think about what the game flow will be and what the two teams are likely to do and you can come up with a few different scenarios to build your lineup around. These types of full-game stacks won’t hit very often, but when they do they give you a big advantage over everyone else and leap-frog you to the top of the leaderboard.
When you over-stack a team or game you’re also making daily fantasy sports easier for yourself. Think about it this way… if you pick players from nine different teams, that’s potentially nine different games that you need to correctly guess the game flow for, then you have to hope your specific players are the ones benefiting from the game flow. If you stack a game – or use two or three mini-stacks in your lineup, you only need to get the game flow right for a few games. You still need your specific players to perform well, but you’ve put the odds in your favor. An example of this might be starting a QB and WR from one team and a WR from the opposing team (stack 1) then picking an RB and DEF from the same team (stack 2) and, if there is another really attractive game, you might be able to take your third WR from one team and your TE from the opposing team (stack 3).
You may have noticed this wasn’t really an article about how to pick players. If you use these game theory concepts correctly, you’ll narrow the potential player pool quickly and it will be much easier to build your lineups. If you pick a game and take the QB and WR from one team and a WR from the other team, then you’re just picking from the WRs in that game and then you’ve only got one other WR to pick and, at this point, your available salary will be limited and you’ll be picking from a smaller pool of players. It’s hard to pick the best WR out of 96, but if you’re only trying to pick the best WR out of the 12 you have enough salary for, it’s much easier.
Think about these concepts as you build your lineups and as you look at other people’s lineups and our weekly stack recommendations. You’ll get better at building your own lineups just by learning to notice these correlations, anticorrelations, and leverage plays when you’re looking at other lineups. These ideas will become second nature and you’ll start getting excited to see the ownership percentages of your players. It’s pretty fun to open your lineup at 1pm and see that your favorite player on your lineup is only owned by 2% of the tournament field. It’s even more exciting when you see an RB or WR (that you don’t have) bust when they’re owned by 40 to 50% of the field! That effectively cuts the tournament field in half giving you a much greater chance of cashing, particularly if your own low-owned players do well.
Good luck!
