There is absolutely no question that the biggest area of growth right now in fantasy football is with the explosion of Daily Fantasy Contests. With several top sites out there already and more entering this space each day, Footballguys is expanding the coverage in this arena to give Footballguys Insiders exclusive coverage of the latest news and information of Daily Fantasy Football.
Over the course of this football season, we will talk about numerous aspects of Daily Fantasy Football. From articles on these sites, strategies to win daily contests, bankroll (money management) strategy, different league and game types, signup and reload bonuses, players to target each week and even contests that look to be undersubscribed ("overlays"), we will cover it all. Welcome to the one-stop shop for Daily coverage.
Please note that even though the NFL has weekly Thursday games and DFS sites have contests that start on Thursdays, in general we will be looking at the majority of the contests and DFS games that do not begin until Sunday. Our weekly coverage does include Thursday DFS information if you are looking to get a jump on the DFS action for the week with the first game on Thursday.
DFS LESSON OF THE WEEK - HEAD-To-HEAD Leagues vs. 50/50 LEAGUES
This week I wanted to take a look at league choice at Daily Fantasy Sites. The most basic method of playing DFS fantasy football is a head-to-head (H2H) matchup where you compete against one other person in a “winner take all” format. This is as simple as it gets – two lineups are entered, and whichever team scores the most points gets the payout minus the league fee. As an example, if I play you in a $10 league, the team with the best score will get $18 and the DFS site collects (usually) 10%, and in this case $2 from our two entry fees of $10.
So this prompts a few questions – how good, on average, must your lineup be to be competitive in a head-to-head league, and is this the best format in which to compete? I will attempt to answer both of these questions, but before I do, I owe you a little personal background. You see, I think it is always helpful to know a little more about the author and where he is coming from when you start to do this type of analysis. I like to think that I am rather uniquely qualified to answer this question from several different angles, and here is why. First, I have a pretty good math background as I have an engineering degree and have been deep in math and statistics for a long, long time – but that is only part of the story. The other part is that I have been watching football for a very long time, and have also coached it (at the youth level) and also played fantasy football for about 20 years. That background in both football and math allow me to analyze games like this from both the numbers angle and also the “gut feel” angle, and look at a problem like this from several different angles. So before I presented all my thoughts, I thought it would be best to give you some idea of how I am drawing my conclusions.
So here we go. First, I am going to do some mathematical analysis on H2H leagues and also compare those numbers to a similar format called the “50/50” league, where 50% of the league (or tournament, as some DFS sites call them) wins that weekly or daily contest. After I get done with the number crunching, I will bring the discussion full circle and offer some different perspectives as to which league is better suited for long term success from both a “gut feel” and personal experience point of view.
Going back to the $10 entry, head-to-head example, the winner of the two person league gets $18 while the site collects 10%. That means for long term success you have to win five times for every four losses, as five wins is a $40 profit ($18-10=8x5=$40) and four losses is a loss of $40 ($10x4=$40). Translating that into a winning percentage, you must win at least five out of every nine contests, or 55.6% of the time, to break even. That number is very important and will be referred to throughout the discussion, so please keep it in mind – you must win at least 55.6% of the time to be competitive in a H2H league, and preferably better than that to earn money over the long run – your ultimate goal.
So what about the 50/50 league? That’s where some heavy math can come in. To the casual observer, the numbers look the same – the top half of a 50/50 league wins, while the bottom half loses, regardless of the 50/50 league size. As another example, let’s take a $10 league of 10 teams – five teams win $18, and five teams lose their $10 entry fee. The site gets 10% once again, in this case $10. So it would seem like the answer is the same – you either win 40% on top of your entry fee ($8 for your $10 entry) or you lose all of your entry – so the winning percentage is the same as the head-to-head league. It appears you need to win 55.6% of the time just to break even.
Not so fast, my friend. Here’s where the complicated math comes in to play. When you are competing in a 50/50 league like this, you can look at it as multiple head-to-head leagues being played all at once. In the above example of 10 teams, you are playing nine competitors, and the goal is to beat at least five of them. So the actual question that must be asked is this – what are the odds that I beat at least five of these nine players? For that answer, we need to go to both statistics and combinatorics. If you do not care to know the details, please skip ahead three paragraphs now to the results below, but for those so inclined about the math behind this, we’ll start with defining the problem a little more.
The math problem in front of us is to figure out the winning percentage that breaks even for this 50/50 league of 10 teams. Looking at all the different outcomes of playing nine opponents individually, you can finish anywhere from first (9-0) to last (0-9) to anywhere in between these two extremes. There is only one way to go 9-0, which is to beat every team, and the same can be said about going 0-9. But what about 8-1 or 1-8? Here you have nine different combinations, as you can beat all but any of the nine teams and go 8-1, just like you can lose to all but one of them nine different ways. This is covered in statistics as combinatorics – the different possible combinations of outcomes.
For a 10-team league, there are actually 512 different combinations. Another way to think of this is the different outcomes of flipping a coin nine times in a row, which would simulate playing nine head-to-head matchups. There are two possibilities each time, heads or tails, and these nine combinations multiply out – making 512 (or 2 to the ninth power) combinations. As we just mentioned, there is only one way to get heads or tails nine times in a row, but there are nine different ways to get eight of one and one of the other. The numbers go up quickly from there for getting 7 of 9, 6 of 9, etc. all the way to 2 out of 9 – but the outcomes all add up to 512. There are the same number of possibilities on either side of 5-4 and 4-5, so it looks like the league does have an even amount of outcomes – 256 where you are 5-4 or better and win, 256 where you are 4-5 or worse and lose – exactly 50%.
Not so fast. I actually made an assumption in my example, which is that all the outcomes are equal. Going back to the coin flipping example, there is a 50/50 chance of getting heads or tails – but what if the coin was not fair? That’s really what we are asking – how much above a 50% chance must we have to be successful in a given 50/50 league. If we can figure that out, then we will know whether we should play 50/50 games vs. H2H contests, at least from a math perspective. To figure this out, I had to go back to the combinatorics and figure out what would be the odds of winning all those different outcomes (going 9-0, 8-1, etc.) for a given winning percentage. In other words, the odds of going 9-0 are better than going 0-9 if I am better than 50% at winning on average. I just need to calculate the odds of winning nine, eight, seven and six times and the number of combinations that each one of those represents. Using some algebra I can then figure out what winning percentage gives me at least a 55.6% win expectation, and that tells me how much above 50% I need to be to expect to win more than I lose in 50/50 tournaments.
Table 1 below represents a summary of the winning percentage you need to have to win the different sizes of 50/50 leagues.
# Of Teams |
Win % Needed to be in Top 50% |
1:1 Matchups |
1:1 Matchups |
10 |
52.30% |
29 |
512 |
20 |
51.60% |
219 |
524288 |
30 |
51.29% |
229 |
536870912 |
40 |
51.12% |
239 |
5.49756E+11 |
60 |
50.91% |
259 |
5.76461E+17 |
80 |
50.79% |
279 |
6.04463E+23 |
100 |
50.71% |
299 |
6.33825E+29 |
150 |
50.58% |
2149 |
7.13624E+44 |
170 |
50.54% |
2169 |
7.48289E+50 |
Table 1 – Necessary Winning Percentages and Combinations in 50/50 Leagues
As you can see from the numbers, the bigger the league, the lower the winning percentage you need to have to achieve at least a 55.6% league-wide winning percentage. Even in a 10-team league, you only need to be 52.3% successful on average to win 55.6% of the time in 50/50 leagues, and that required win rate only gets smaller for bigger leagues. For example, in a 100-team 50/50 league, the winning percentage required to win the league at least 55.6% of the time is only 50.71% - or almost 5% less than in a Head-to-Head league. That’s extremely significant. Gaining a 5% win advantage is nothing to sneeze at, and that is why I ran all of these numbers.
The conclusions I draw here are that in bigger leagues, the margin above 50% that you need to be is smaller because of how many different winning outcomes are represented in big leagues. Back to our last example, in a 100-team league, you are playing 99 other players and you need to go 50-49 or better to win in that contest. There are 100 different records you could have (99-0, 98-1, …, 1-98, 0-99), but the odds of winning 99 games are far greater than winning zero, and the same can be said for winning 98 vs. 1, and so on. With 50 outcomes a “win” and 50 a “loss”, it only takes a very small advantage to tilt the odds of the Top 50 outcomes where you win to be at least 55.6% in your favor. That’s what I wanted to prove to both myself and to you, so that we can draw the following conclusions:
Conclusion 1 (from statistics): 50/50 Leagues offer a significant advantage to win in DFS over Head-to-Head (H2H) leagues.
Conclusion 2 (from statistics): While both require winning percentages > 50% for success, the winning percentage required for success (and positive earnings over time) is much smaller in 50/50 leagues than in H2H leagues.
Conclusion 3 (from statistics): The larger the 50/50 league, the lower the winning percentage you need for success and positive earnings over time.
Now, I need to bring this full circle, as I promised. Time for my “gut feel” approach to this problem. My own experience and feeling is that more novices and average players play in 50/50 leagues, and that fact alone will improve your odds of winning because now you have the advantage over those players because you should have good DFS experience and be favored more often than not over casual players. In head-to-head leagues, many sites require someone to start a H2H matchup, and typically more experienced players kick off or set up a H2H league. They do this for “more action”, and they hope that novices and casual players just step in and play them to take their action (and money) in a matchup where the experienced player will have a significant advantage.
Bottom line – stick to 50/50 leagues and avoid H2H matchups in DFS, and if given the choice, enter bigger 50/50 leagues.
INDUSTRY NEWS
DraftKings' Huge Problem Last Sunday
Further Improvements to Footballguys' Tools for DFS Lineup Creation
Footballguys' own Maurile Tremblay has been at it again. This time, he has added informative descriptions and advice to his Interactive Value Charts at Footballguys. They are available as links on each of the pages, but I think everyone can benefit from reading the new information. Here it is:
How to use the value charts
Daily fantasy football contests can be loosely divided into cash games and tournaments. Cash games comprise contests in which roughly half of the field finishes in the money. Tournaments comprise contests in which a much smaller percentage of the field -- generally between 10% and 20% -- finish in the money. Tournaments are often referred to as "GPPs," which stands for Guaranteed Prize Pools. I prefer the term tournaments to minimize ambiguity because a lot of 50/50s and Double-Ups also have guaranteed prize pools.
(Head-to-head contests are cash games. Strictly speaking, they are very small 50/50s. Some contests fall in between cash games and tournaments, such as Triple-Ups, or five-person contests that pay the top two spots.)
Any lineup is capable of scoring a wide range of points. The top end of that range is often referred to as the ceiling, while the bottom end is referred to as the floor. Your primary concerns in putting together either a tournament lineup or a cash-game lineup are (a) to select players who present great value as measured by expected points per dollar, and (b) to spend most or all of the salary cap. Follow those rules, and your team will be expected to score a lot of points, which is how you win both cash games and tournaments (and anything in between).
There are some differences between lineups best suited for cash games and those best suited for tournaments, however. You are invited to click on the "Cash Games" and "Tournaments" links for advice concerning each type of contest, but the basic idea is that, in cash games, you are more concerned with achieving a high floor, while in tournaments you are more concerned with achieving a high ceiling.
In general, a lineup expected to score significantly more points than a competing lineup will have both the higher floor and the higher ceiling; so putting together a lineup with an above-average expected score is the first order of business in either type of contest. That's what we'll focus on here.
We will want to concentrate on value rather than raw points, so the first step is to click on "Value" from among the options below in blue. Then we'll sort the player pool from highest value to lowest using whichever set of projections you prefer. To do so, click on the appropriate initials in the table header.
Click on the various positional tabs and look for players near the top of the list who stand out as having large salaries. Those are the players we're most interested in because they'll help us achieve both of our goals at once -- finding great values and using up our cap space.
Once you've inserted a few such players into your lineup, let the app fill in the remaining spots. Its suggestions will maximize total expected points, but sometimes you won't be happy with its choices for whatever reason. Remove any players you're not comfortable with, exclude them from consideration (by clicking on the green "o" next to their names in the player pool), and let the app do its thing again. Repeat until you're happy with the results.
If you are entering multiple contests, you will likely want to diversify your lineups rather than choosing the same players over and over again. The multiple sets of projections come in handy here. You may, for example, want to choose some RBs based on Maurile Tremblay's projections, choose a quarterback based on David Dodds' projections, and then fill in the remaining spots using Sigmund Bloom's projections. To find a second lineup with different players, select a QB using Sigmund's projections, some WRs using the average projections, then fill in the remaining spots using David's projections. For your next lineup, try a different combination.
Cash games
We stated in the general strategy section that the first order of business, whether in a cash game or in a tournament, is to construct a roster that is expected to score a lot of points; but that in a cash game, we are more concerned with our team's floor than with its ceiling.
There are a few ways to increase our team's floor.
The most obvious is to select individual players with high floors. That is to say, we will eschew boom-or-bust players and focus instead on more reliable players. This is more of an art than a science, but there are some decent indicators of reliability. First, we will prefer a player if we are confident that we know his health status, and the health statuses of his teammates. (We don't want a player who might be a late scratch or whose role might be limited; we also don't want a WR on a team whose QB is ailing; and finally, we don't want a player who will get significant playing time only if his teammate, who is currently listed as questionable, sits out.) Second, we will prefer a player if his involvement in the offense is durable. A one-dimensional runner, for example, can be taken out of the game plan if his team gets behind early; while a running back who is also heavily involved in the passing game should get touches no matter the score. We will prefer a player whose fantasy production depends more on yards than on touchdowns, and we will prefer a player who steadily gets yards in small chunks over a home-run hitter who depends on the big play.
Another way to reduce a team's volatility (and thus increase its floor) is to avoid QB-receiver combinations from the same NFL team. If a team's QB has a poor game, most likely his receivers will, too. Avoiding such QB-receiver combos is a form of insurance: if our fantasy QB has a poor game, our fantasy receivers may bail us out by having good games. That possibility is slim if our QB and receivers come from the same NFL offense.
While pairing players who score fantasy points together (such as a QB and receiver from the same NFL team) will increase volatility, pairing players who generally compete for fantasy points (such as a QB and RB from the same NFL team) can reduce volatility, and therefore makes sense in a cash game. If we have the quarterback and the primary running back from a given NFL team, we are very likely to score points one way or another whenever that NFL team scores. If it scores a passing touchdown, we get points from the QB; if it scores a rushing touchdown, we're likely covered there as well. Add that team's kicker to the mix so that we're covered even when the offense stalls in the red zone and has to settle for a field goal. Feel free to add the opposing team's defense as well, so that if your QB, RB, and PK have bad games, at least your defense should get you points.
Other combinations of players who generally compete for points (and whose collective output is therefore relatively stable) include RB-RB, RB-WR, RB-TE and WR-TE combos from the same team. (It is difficult, however, to find two RBBC-members who are both good values, so the RB-RB combo will be rare.)
All of these factors should be considered tie-breakers more than anything else. When you're choosing among players with significantly different values (as measured in expected points per dollar), the player with the highest value is almost always the best choice. That's true in any type of contest, but especially in cash games. Only when choosing among players with similar values should these additional considerations come into play.
Tournaments
Whenever your team is expected to score more points than its target, volatility is your enemy. But when your team is expected to score fewer points than its target, volatility is your friend.
In a cash game, your target is the score earned by the median entrant in your contest -- and if you can't beat that more often than not, you shouldn't be playing. But in a tournament, your target will often be somewhere in the top 10% to 20% of the field. Nobody can expect to finish there most of the time. The goal is to finish there more than your fair share of the time, and a boost from volatility (which amplifies both your good results and your bad ones) can be rather helpful.
One way to increase the volatility of your roster is to load it with high-risk, high-reward players -- players with great upside potential relative to their projections. Look for players whose number of rushes or targets is inconsistent from week to week. Look for home-run hitters who can score bunches of points on relatively few opportunities. Look for players, other than quarterbacks, on offenses that are expected to score a lot of points. (The qualifier "other than quarterbacks" may strike you as odd. Quarterbacks on high-scoring offenses will be projected to score a lot of points, and will be valuable in any format. But they are not particularly likely to greatly outscore their lofty projections, so they are not especially valuable in tournaments compared to cash games. On the other hand, moderate differences in how looks are distributed among runners and receivers in a given week will have proportionately larger effects on their fantasy output if they are in higher-scoring offenses, making those players more volatile.)
Another common way to increase volatility is by "stacking" your lineup with a quarterback and receiver from the same NFL team. By doing so, you are increasing your chances of having a very high or very low score from that combination of players instead of a medium score. This makes sense in tournaments because a very high score is the goal, while a very low score is generally no worse than a medium score: you're out of the money either way.
Let's say you've got Kelvin Benjamin at WR, and now you're choosing between Cam Newton or Matt Ryan at QB. If you select Ryan, the QB-WR combination could be great-great, medium-medium, horrible-horrible, great-horrible, horrible-great, medium-great, and so on. All possible combinations are on the table because the QB's performance and the WR's performance are uncorrelated.
But if you select Newton, now certain possibilities that lead to an overall medium score (like horrible-great) are much less likely. The odds of great-great and horrible-horrible are increased, while the odds of great-horrible and horrible-great are decreased, and the overall effect is to make the combined score more volatile. Compared to the Ryan-Benjamin combo, the Newton-Benjamin combo is more likely to be very high or very low, and less likely to be medium. That's an advantage in tournaments, where medium scores are useless.
A further strategy that can be helpful in tournaments is to seek out uncommonly owned players. Your primary goal at each roster position is to earn more points per dollar than your opponents. You can't win a tournament without being successful in that respect at most positions. Taking it for granted that you'll have to outscore the bulk of your opponents at a given position, you'd prefer to do so with a player owned by 5% of the field rather than with a player owned by 40% of the field. To win cash in a tournament, you may need to beat 90% of your opponents. If you get a great performance from a player who is 40% owned, you just passed 60% of the field. That's nice, but if you get a great performance from a player who is only 5% owned, you just passed 95% of the field. That brings you significantly closer to your ultimate goal.
None of this is absolutely essential. You can have a successful tournament lineup without boom-or-bust players, without stacking, and without sparsely owned players. But at the margin, among many factors to consider, those things are generally points in favor of a tournament lineup.
Couple all of this with his "Top 20 Stacks" that are also available, and these pages are a gold mine. Well done, Maurile.
ON THE DAILY AND INTERACTIVE VALUE CHARTS
Footballguys' own Maurile Tremblay does not just do projections for Footballguys - oh no, far from it. One of the best innovations that he has provided since last year is his "On the Daily" column, where he takes his projections, David Dodds' projections, and Sigmund Bloom's projections and provides recommended optimal lineups for five major DFS sites (FanDuel, DraftKings, DraftDay, FantasyAces, FantasyFeud) based on those projections. How cool is that?
But what if you want to play at other sites or build your own lineup (or multiple lineups)? Once again, Maurile has you covered. He has provided interactive value chart pages where you can click on each player, sort by position, salary and value (points per $10K) to craft your own custom team. All of the links for these are provided below in our Daily Content on Footballguys table plus they appear weekly on Footballguys' weekly home page.
But wait - there's more! Maurile went even further by adding the ability to partially fill out a lineup and have the tool fill in the rest of your roster. From Maurile himself:
How cool is that?
DAILY CONTENT AROUND FOOTBALLGUYS.COM
NFL WEEK 12
Halfway into the NFL season, and we have had some big, big contests in DFS so far - and the fun continues. Let's dig into some of the big contests out there. If there are any that I miss, please drop me an email (pasquino@footballguys.com) and I will update this article.
FANDUEL - WEEK 12 BIG CONTESTS
All of these contests can be found in the lobby at FanDuel.
FanDuel Presents the Playboy Championship
This year our friends at FanDuel are offering you the chance to go on the trip of a lifetime to the world famous Playboy Mansion in Los Angeles! You and 75 finalists will compete for $1,000,000 in prizes, including a $150,000 first place prize as you sit poolside with Hef at this year’s hottest fantasy football event.
Not only that, they’ll be running contests for 25 additional trip-only packages to maximize your chances of getting in on this exclusive party. That’s 100 chances to win!
Qualifiers start at only $5 to enter, seats to this epic event will be going fast, so enter one of the qualifiers and punch your ticket now!
Click to enter PFC $5 Qualifier
Click to enter PFC $50 Qualifier
Click to enter PFC $535 Qualifier
DRAFTKINGS - WEEK 12 BIG CONTESTS
Are you interested in winning $1 Million Dollars in a single week? DraftKings hopes that you are.
NFL $2.2M Millionaire Maker - $1,000,000 to First Place. Another week, another guaranteed Millionaire Maker, where you can turn $27 into $1 Million! Over the past three weeks, the Millionaire Maker has paid out $1 Million to first place, and DraftKings is doing it again this week. First place will become an instant millionaire, and over 15,000 spots are paid out in total. Multiple lineups per player are permitted. Enter now!
King of the Beach - Only two weeks left to qualify! Just $12 to Enter. This King of the Beach Qualifier awards first place one of only 50 spots in the $2.5M King of the Beach and a VIP trip to Atlantis (Dec 4th - 8th) for the live final on December 7th! Enter now for your shot at the $1M top prize! See King of the Beach promotions page for full terms & conditions.
NFL Sunday 300 Grand. $300,000 in total prizes. $200 to enter. Up to 1,667 entries with the top 350 getting a payout. $50,000 to first.
World's Biggest $50 50/50 - This 2,224-player contest boasts a $100,080 prize pool and pays out the top 1,112 finishing positions. All winners earn $90. Good luck!
World's Biggest $20 50/50 - This 6,668-player contest boasts a $100,008 prize pool and pays out the top 5,556 finishing positions. All winners earn $36. Good luck!
World's Biggest $10 50/50 - This 11,112-player contest boasts a $100,008 prize pool and pays out the top 5,556 finishing positions. All winners earn $18. Good luck!
Note from Jeff Pasquino: As of Thursday afternoon, all of these contests were less than 40% full. Keep in mind that this is a guaranteed prize, so there is definitely an opportunity here for a large overlay situation. Watch these closely, all the way to Sunday morning.
Future Big Contests:
"Four months, TEN millionaires": Nine contests where the top winner will win at least $1 Million. Weeks 5, 6, 7, 8, 9, 10, 11, 12, 15 and 16 of the NFL season. Direct entry is $27, but you can earn your way in for as little as a quarter - yes, $0.25. (Odds are certainly much steeper at $0.25 vs. $27, but hard to get more action for a quarter!).
Questions, suggestions and comments are always welcome to pasquino@footballguys.com.