Large prize pool tournaments are one of the most popular forms of Daily Fantasy Sports. They require very different strategy from other game types, namely cash games, so it’s difficult for someone to be great at both. Most GPP players struggle in cash games, making subpar head to head and 50/50 lineups, and most cash game players make suboptimal GPP lineups. I fall into the latter group: a cash game player who has not had as much GPP success.

Despite my lack of huge tournament scores, I’ve played a great deal of GPPs over the last 2 years and have been profitable playing them. Some of my better results: I won the nightly MLB payoff pitch for $12k, finished 18th in the Draftkings 2015 NBA Live final for $30k, finished 2nd in the week 2 of the 2015 NFL season $1k on fanduel, and finished 2nd in the nightly $1k buyin NBA gpp a few times for $15k. I’m a cash game player at heart, but I have more than enough experience to write an intelligent beginner guide on GPP play.

Let’s go over some basic tournament strategy and approaches for NFL.

Step 1: Identify the Chalk

Cash games require precision: the people who make the best projections and model the games accurately will be the best cash game players. There is no room for error: one player projected incorrectly can cause a domino effect, bringing not just one suboptimal choice but a few others into your lineups.

GPP players don’t need to be nearly as precise. Instead of making accurate projections, GPP players must instead identify who the most popular choices are at each position and decide which ones are worth playing despite their high ownership.

Why does ownership matter?

In tournaments, game theory is often more important than accurate projections. I’ll explain this using an example I came up with inspired by a DFS author and player named Jonathan Bales.

Imagine you were playing a game where you had to guess the gender of a Kardashian-Jenner child at random. For those of you who don’t follow pop culture, there are 8 Kardashian-Jenner children: Kim, Kourtney, Khloe, Rob, Kendall, Kylie, Brody, and Brandon.

Note: there were 2 others from a previous marriage, but let’s ignore them for this example and only consider the last 8 children born.

source: The Kardashian Family

So there are 5 girls and 3 boys. Suppose first someone said if you guess the gender of a random Kardashian-Jenner child correctly, you win $100. Since 5 of the 8 children are girls, you will choose girl every single time because the odds are much better for you than if you chose boy. This is analagous to cash games: you pick the highest percentage players to do well.

Now suppose you are playing a different sort of game. You and 99 other people get together, put in $100 each, and each of you picks a gender for a random Kardashian-Jenner child. If the child is a girl, everyone who chose girl splits up the $10,000 prizepool equally, and if the child is a boy, everyone who chose boy splits up the $10,000 prizepool between them.

In this example, it’s still 62.5% that the random child is a girl and 37.5% that the random child is a boy. However, if you get the feeling that 80% of people will choose girl and 20% will choose boy, then you clearly want to choose boy! The reason is that when you choose boy you and 19 others split up $10,000 37.5% of the time, meaning each person gets $500 37.5% of the time, for an expected return of $187.50 on an investment of $100, while choosing girl loses money. But if more people had this same thinking and 50% chose boy, then you would want to choose girl. The correct choice depends on what everyone else does. This situation is analogous to GPP play.

Being contrarian in GPP play is necessary at times. However, as seen in the example above, if too many people go contrarian, then the straightforward play is better. And sometimes, if the straightforward chalk play is that much better than anything else, it should not be avoided.

So first, GPP players must identify who will be popular.

Their next step is to decide which of those players are such incredibly good plays that they need to be in lineups regardless, and which players should be avoided for less popular choices. For instance in week 3, the player who won the millionaire maker had Shane Vereen in every one of his lineups: he was $3700 and expected to get many touches. It did not matter what his ownership was, he was a must play. But other times, you can avoid a popular choice or only play him on certain teams. Regardless, the first step is identifying who is popular.

Correlation Plays

The most obvious correlation play is quarterback and wide receiver. If my wide receiver catches 12 passes for 300 yards and a touchdown (as Julio Jones did last week), the quarterback who threw him those passes also does well. It was virtually impossible to win a tournament in week 4 without having the Matt Ryan-Julio Jones stack.

However, there are many different correlations someone can seek out.

QB-TE or QB-RB

Wide receivers are not the only players on a team who catch passes. Having whoever is most likely to catch touchdown passes from a quarterback paired with that quarterback is a good correlation play for GPPs. In general the most likely player is a wide receiver, but remember the importance of being contrarian: sometimes the unlikely choice is a higher EV option.

QB-WR and RB from opposite team

A quarterback+wide receiver combination will often do best when the game script is optimal for them: if they are down points and need to be passing a lot towards the end of the game in an effort to catch up. When this happens, the opposite team already scored points, and they are trying to run out the clock quicker by rushing the ball more instead of passing. The easiest way to take advantage of this is to pair the QB-WR combination with the running back from the opposing team.

But this combination can be extended to both wide receivers and tight ends from the opposite team: those other players could have contributed to that team scoring many points as well.

RB+defense

The best game script for a running back is having a moderate to big lead and high time of possession. If a running back’s team has a big lead and they have the ball a lot, that means the team’s defense is probably doing pretty well.

The Onslaught: many players from a single team

Last year the Millionairemaker on draftkings was won by an unlikely combination of players: someone chose to play the QB, RB, WR, WR, and defense from the Pittsburgh Steelers. The entire team crushed, Ben Roethlisberger threw for 364 yards and 4 touchdowns, the running back had 134 yards rushing and 5 catches, Antonio Brown had 118 yards receiving on 8 catches with 2 touchdowns, and Martavis Bryant added over 100 yards receiving and a touchdown. The defense added 5 sacks, 2 interceptions, and a forced fumble along with keeping the Colts to just 10 points.

source: Draftkings Playbook Recap

Choosing many players from the same team is not a traditional correlation play: in fact they have a negative correlation overall. Since scoring is so dependent on touchdowns, and for every touchdown one player scores, the others cannot, players on the same team tend to cannibalize each other’s fantasy points. However, having the onslaught is not hoping for simple correlation: it wants a one sided blowout where everyone on the team does lots of stuff. If you think about it from that perspective, when a team wins 42-7 the onslaught did extremely well.

The Game Stack

Last season, the Giants and Saints clashed in an epic battle resulting in over 100 total points. The Saints won 52-49 in a complete shootout where Drew Brees threw for over 500 yards and 7 touchdowns and Eli Manning threw for 350 yards and 6 touchdowns. Having a GPP lineup full of players from this game guaranteed a high score.

Take a look at this box score: if you managed to have Drew Brees, Brandin Cooks, Ben Watson, and Odell Beckham on the same team you were guaranteed a high finish in GPPs regardless of who else you chose.

When you choose to game stack, you basically hope for a shootout with lots of points scored in the game. It’s also not a traditional correlation play, but once the points start rolling in quickly, everyone on your team gets more opportunity to score.

Bankroll Management

If I play mainly head to head matches and set a lineup that finishes in the 40th percentile, I will likely win around 40% of my matches and lose 60%. If I play 50/50s and double ups, I only need to finish in the top 50% to win money. However, GPPs are much more top heavy: usually only 20% of the field cashes and the majority of that money goes to the winner. Even the best players can get unlucky over a long stretch of time doing this, and playing well within your means is necessary to sustain the down stretches.

Multi Enter To Spread Ownership

While retaining good bankroll management, having more than one entry in a contest can help you land on the correct play.

Suppose you are deciding which team to stack in a given week. You really like the Raiders, the Saints, the Chargers, the Packers, and the Panthers. If you were only entering once you would have to make a difficult choice. However, if you enter 10 times, you can stack each team you like a few different ways using different receivers each time, and can even include a couple game stacks or onslaughts alongside them. Sometimes your 8th entry is your most profitable.

Playing GPPs is exciting. You can win huge amounts of money on a small investment.

But it still requires a lot of strategy, and there are many approaches to choosing your team. As long as you practice strong bankroll management, and apply logic to your choices in the form of ownership considerations, mixing in must plays, and correlation, you should have a leg up on the competition. Good luck to everyone in week 5!

---

My name is Ryan Daut and I'd love to have you as a follower. Click here to go to my page, then click in the upper right corner if you would like to see my blogs and articles regularly.

I am a professional gambler, and my interests include poker, fantasy sports, football, basketball, MMA, health and fitness, rock climbing, mathematics, astrophysics, cryptocurrency, and computer gaming.