Wisdom From the Geek – Reaching For the Nuts – MMA, PGA, and NASCAR Multi-Entry GPP Strategy

Geeks-Vegas-Lines (1)

 

Reaching for the nuts? At this point, you’re probably wondering what on earth I’m talking about. This column is going to discuss  multi-entry GPP strategies for some of the sports I love to play when the NFL is on hiatus with a focus on going for “the nuts” lineup. Before I get to that I’ve got some background to cover first.

It’s the NFL off-season and as a DFS degenerate, I always like to get my fix. I’ve always preferred the once a week sports (like NFL) because it gives me a full week to research and prepare for a slate. I can take my time and craft carefully researched lineups over the course of that week.  The daily nature of NBA and MLB has never been a great fit for me. They require a ton of time and concentration particularly during the time of the day (5 PM to 8 PM EST) I like to spend with my family. I still dabble in those sports but I had to find profitable alternatives during the off-season.

Last season when Draftkings introduced Daily Fantasy MMA I was instantly hooked. I’ve always been a big fan of the sport and just having some action on it made it much more exciting to watch. I played my way up to a top 16 ranking in MMA on Grinders playing almost exclusively in GPP’s. MMA DFS is interesting because the total number of potential lineup combinations are much lower than in any of the other DFS sports. Generally, there are between 10-14 fights going on a particular card and you chose five fighters for your fantasy team. In a mathematical sense there are tens of thousands of possible player combinations depending on the number of fights, but with some research and strategy, you can narrow that down substantially. This brings me to the title of this column.

Reaching For the Nuts

In poker, the nut hand is the strongest possible hand in a given situation. The second-nut hand or third-nut hand (and so on) may refer to the second and third best possible hands. The term applies mostly to community card poker games where the individual holding the strongest possible hand, with the given board of community cards, is capable of knowing that they have the nut hand.

Per Wikipedia, The origins of the term may go back to colonial times when a player betting everything they owned was said to take the nuts from their wagon wheels and put them on the table to ensure that if they lost they could not flee. It was said that a player would only do this if they knew they had no shot of losing. The folklore says that this is how the term for the ultimate hand in poker came to be known as “the nuts”.

When playing in GPP’s what you are trying to do is get as close as possible to the perfect lineup for that evening’s slate. While that feat is rarely accomplished by anyone in GPP’s for sports like NFL, NBA and MLB; in MMA, every GPP pretty much features at least one player that hit The Nuts lineup. I haven’t seen any stats on PGA and NASCAR but I’d venture a guess that it’s very common that those GPP winners hit the nuts lineups as well.

Why Go For The Nuts?

Obviously, one of the things we emphasize at the DFS Army is bankroll management discipline. We approach that discipline by emphasizing cash contests and smaller winable GPP’s as a way to manage your bankroll and in essence pay for the shots we take in larger top heavy GPP’s for that big payday. That is obviously still our recommendation particularly in the sports or slates where the player pools are so large that it is virtually impossible to hit a perfect lineup. Still, winning a GPP is awesome and generally, the only thing that can effectively move a bankroll up into the next stratosphere it makes sense to go for it when you have a somewhat legitimate chance to pull it off. The key is to maintain the balance of bankroll discipline while maximizing upside.

The Approach

I played NASCAR DFS for the first time the other day. One of our DFS Army expert writers, Taco, had shared a spreadsheet about tonight’s NASCAR event in our slack forum. He had gone over some of the basic strategies as far as how to construct a lineup and why. He pointed out a group of drivers he considered top plays and with an explanation around each one as to why they were top targets. Armed with that info and the knowledge that there were only 25 drivers to chose from, I knew there was a high likelihood that  someone was going to find The Nuts lineup in GPP’s. Why not me?

Think of GPP entries like lotto tickets. The more lineups you create the better the odds of hitting the jackpot (a GPP Win). The key difference is that in the lotto you win it all or you get nothing. It’s a zero sum game. In DFS GPP’s you can turn a profit without hitting the nuts lineup as long as your core or ALL IN plays do well.  – Geek

I like these kinds of contests because you can deploy a strategy of attacking the GPP with the intention of finding the nuts lineup while minimizing risk by researching the slate and focusing in on a core group of high odds plays while mixing up the supporting cast. You generally need exposure of a wide array of players from the pool and I’d say a minimum of 25 to as many as 200 or more lineups to properly execute this strategy. I’ve never watched NASCAR and I have no knowledge of the sport. I was still able to take Taco’s recommendations and spin out 50 or so lineups with about an hour of work.

At this point you may be thinking that this type of strategy is for big bankroll players that can afford to churn out tons of lineups. It’s a common excuse I hear from whiners that complain about multi-entry GPP contests in general. That of course is bullshit. Draftkings offers multi-entry GPP’s starting as low as $0.25. Lack of bankroll is not the issue. Lack of confidence is generally what holds players back from deploying the strategies I’m about to get in to. I think of my Multi-GPP entries in the same way portfolio manager looks at a stock portfolio. Portfolio managers are trying to bead the market as a whole. In multi-entry GPP’s we are trying to beat the field. There is some safety in numbers and there is power in knowing that when you spin a large number of lineups chances are some of them will be in the money and some will not. In an average GPP on Draftkings the top 20% of finishers generally get a piece of the prize pool. Payout structures change from one GPP to another (I prefer flatter payout structures) but one constant is that they are quite top heavy. There is a massive premium built in for finishers in the top 10 or top .1%. I tend to look at the payouts in terms of what multiple of the entry fee they are or in simpler terms, how many entries did that prize pay for. In a typical $3 entry GPP for example if a 100th place finish paid out $30 I’d look at it as having covered the cost of 10 lineups entered.

Controlling Exposure

Armed with the knowledge that creating a diversified portfolio of lineups actually reduces risk the next step is to begin the process of building lineups. It’s important to remember that we aren’t just randomly selecting a bunch of teams and crossing our fingers.

I always take a bottom-up approach to lineup construction. The starting point is about identifying value across the spectrum of players in all the different salary ranges. Obviously there are a number of ways to identify value but researching the Vegas odds and comparing them to player salaries is always a solid start. It is inevitable that you will find a number of players in every slate that stand out as under-priced. Occasionally there are players which for whatever reason are wildly under-priced compared to their odds.  Those become the all in value plays. The next step is to research the studs to get a feel for who stands out in the elite salary group as the best play. The research is key because in GPP’s there can be a benefit to fading the chalk high ownership players to get some differentiation going in your lineups.

The next step in the process is to look over the entire player pool and eliminate players that you plan to fade completely. It’s difficult generalize this process across the different sports because each has its own flavor. In PGA I look at recent form as well as course history. If a player has been missing cuts recently or has a terrible history at a particular course I write them off completely. If in the process of comparing the Vegas Odds to player salaries I find players that stick out as overvalued compared to their odds I scratch them off the list. These are obviously some examples but there are plenty of ways to narrow down your player list not the least of which is following the picks of the experts we have writing our DFS Army columns.

Armed with a list of value plays, stud plays, and pivots, its time to start constructing lineups. When creating a portfolio of lineups I have three basic types of lineups that I like to construct – Studs and Duds, Balanced and Contrarian.

Studs and Duds

Regardless of the sport you are playing one constant in DFS is that you have to pay up for studs. Fitting more then one in a lineup often requires a studs and duds strategy. In this strategy we basically load up on some of the favorites and fill out the lineups with duds (low priced players you’ve identified as solid values).  It’s important to hweavily diversify the duds in this type of lineup. The cheaper salary players are there for good reason. They are going to be more volatile and more likely to sink a lineup then a top dollar stud. One of the ways to ensure you are getting a diverse mix of players in your portfolio is to limit exposure to the players with the longest odds. This type of lineup works best when the chalk (favorites) do what they are expected to do.

Balanced

Another lineup construction method I like to use is the balanced approach. Rather then grabbing the high dollar favorites and being forced to fill out the rest of your lineup with duds, you go for the safer plays in the middle tier of salary. As an example in Draftkings PGA you need to choose 6 golfers within a 50K salary cap. That leaves an average of $8333 in salary for each golfer. In a balanced lineup you may limit yourself to golfers with salaries between $7500 and $9200. Normally the chalk favorites are priced between $10,000 to $13,000. Using a balanced approach to lineup construction forces you to fade the chalk plays which adds differentiation to your lineup. A balanced approach is also a useful cash strategy.

Contrarian

Chasing the nuts is all about taking down 1st place in a GPP. The promise of the big payday that comes with a 1st place finish is enough to offset the risk of upping the ante as far as massively increasing the number of lineups you create for a single event. In sports like MMA and PGA and NASCAR, where the total number of lineup combinations is low, there tends to be a ton of lineup overlap which can lead to large numbers of identical lineups in the larger GPP’s. One way to reduce the chances of a tie is to leave some salary on the table.  This feels a bit counter intuitive but if you think about it analytically what are the chances that the nuts lineup also happens to be one that leaves zero salary on the table? Very slim. I wouldn’t go crazy with this one and start chosing teams with all duds or anything like that bit don’t be afraid to leave a few hundred dollars in salary on the table (more in MMA) to increase the odds that the lineup you are entering will be unique in your content.

Differentiation and Game Theory

Another way to differentiate your lineup is by purposefully choosing players that you normally would avoid for various reasons or purposefully fading the chalk plays. Each sport is unique with regards to how chalky a player can be but I’ve seen smaller slate contests in the major sports where a player was 80% or higher drafted. Game theory says that one way to get an edge on the field is to fade that heavily owned player. Lets assume for a moment that Player X that is heavily owned has a 45% chance of delivering value for the slate with a 30% shot of exceeding value by one standard deviation and a 10% chance to exceeding value by 2 standard deviations or more. That player is the chalk play because they have a 85% chance to meet or exceed value for the slate. Obviously, in cash games you want to use that player early and often. It is a +EV situation that you should exploit to the fullest. In GPP’s however there is another possibility to consider. Going back to the example lets assume that the chalk play mentioned above has a 10% chance of missing value by one standard deviation unit and a 5% chance of missing value by 2 or more standard deviation units. In live sports there is always a downside. Players get injured, ejected or just have an off night. It’s impossible to account for injuries when constructing lineups but being aware that shit happens is important. Now lets assume that the chalky player with 80% ownership tanks. That player completely destroys 80% of the lineups in that GPP. If you faded him you are looking at a likely cash and a much easier and higher odds chance of a GPP win.

Recently there was an MMA slate where Rhonda Rousey took on (and eventually lost to) Holly Holm. Holm was a massive underdog and most people gave her no shot to even make it competitive. Using her and her extremely cheap price tag, plus avoiding Rhonda and her hefty salary opened up some incredible lineup construction possibilities in that slate. Even when I feel really good about a fighter or golfer or driver it makes sense to hedge your bets, particularly when the field is all about one particular player. Upsets happen and by their very definition, it is difficult to predict when.

Using an Optimizer

One great way to multi-enter a GPP is to use an optimizer that can quickly generate as many lineups as you want. We are currently working on an multi-entry optimization product for DFS Army members. There are some acceptable free ones floating around the web. The trick with an optimizer is to manipulate them into spitting out the types of lineups you want to see. Here is a brief breakdown of how I would set up the optimizer for a PGA contest:

  1. Run through the player list and remove all of the golfers you have determined through your research that you will be fading that week.
  2. Modify the projections for all of the golfers so that they are neutral values (we want to be able to set exposure % manually as opposed to allowing the optimizer to decide how often to use a particular player
  3. Set up your exposure % for each golfer. This part is tricky and needs to be handled with nuance. The cheaper salaried value plays are more volatile by nature and carry lower odds of making the cut then the more expensive “favorites”. Still when going for the nuts you want at least minimal exposure to all but the worst plays. Assuming you plan to make 100 lineups, I’d set exposure max on some of the cheaper salary duds as low as 2%. Work your way up from there. Set up your high value dud plays and your core value and chalk plays between 15-50% exposure. I rarely will go over 50% on any one player.
  4. Do some test runs by spinning lineups and watching to see how the player spread works out. I tend to add a parameter that requires a minimum of two distinct players in each lineup to increase the player pool diversity over the 100 lineups.

Conclusion

One of the benefits of the NFL off-season is the opportunity to dabble in some DFS sports you normally wouldn’t be interested in. Here at the DFS Army we have added expert writers for PGA and NASCAR. I’m in the DFS Army slack forums discussing MMA and giving out my player picks as well. Give one of these sports a try by executing a multi-entry GPP strategy in the Draftkings quarter arcade or at whatever $ entry level you are comfortable with. You may even get a chance to grab the nuts and squeeze out a big payday!