I recently began my transition into NBA DFS. I have been mixing it up with some GPP and cash-game action each night. I have fared well enough with my cash games to keep my bankroll unharmed by my lack of GPP success. Figuring out how to better build a GPP lineup is going to be my main focus going forward. I have begun to employ a new lineup building process. This process involves adjusting a model while in the process of building a lineup. I thought it would be worth touching upon, in case our users are not taking full advantage of how customizable the tools on FantasyLabs can be. Even more exciting – I am going to use a recent pizza order of mine to help break down the idea.
This is a true story. I accidentally revolutionized my world of pizza and I’m going to tell you how. You may also learn something DFS related as well.
Believe it or not, I have a girlfriend. We live together and we are often faced with a dilemma when deciding on what we are going to eat. On most nights, it breaks down like this: I want pizza, mac and cheese, or nachos; she wants almost anything else. The other night it was a different scenario – we both wanted pizza, but couldn’t come to an agreement on what kind of pizza we would order.
When it comes to food, I am a simple guy who tends to lean more on the safe side. She is a bit more adventurous and willing to try new things. I would say I order my pizzas a little closer to the cash game mindset while she has a more of a GPP taste. Luckily for us, Dominos allows you to customize your pizza online. What we created was what I think is the perfect blend of cash game and GPP strategy; now let’s try to apply that to building an NBA lineup.
Similar to the Dominos pizza builder, FantasyLabs new Lineup Tool has made the process of creating lineups much easier. There have been some questions about the Player Models and which ones are best to use. I suggest that anyone who has any questions take the time to read some of these recent articles:
- The Sophistication of Simplicity: The Simpleton’s NBA Cash Game Model
- One Size Doesn’t Fit All in NBA Model Building
- Accounting For Error in DFS Player Models
My transition into beating NBA has been a learning experience. I have had mild success playing cash games and I am now attempting to figure out a profitable GPP strategy. My goal now is to find the balance between playing the more obvious cash game players and mixing them with some riskier GPP plays. I think one way to do this is going to be by utilizing more than one model when creating a lineup.
When my girlfriend and I were creating our pizza, I went first and selected the always reliable pepperoni. Let’s take the same approach for this lineup. I use the Phan Model when building my cash game lineups, so I think we can start there. I went ahead and selected Andre Drummond, Norris Cole, Devin Booker, Markieff Morris and PJ Tucker. All of these players ranked highly in the Phan Model and they will serve as the not-too-risky pepperoni portion of this lineup.
Now we have some value, security, and great taste. Our lineup doesn’t look too bad either. Let’s say you were comfortable with that core of players and decided you wanted to optimize the rest of the lineup. That would leave us with this:
You may think that this is a fine lineup, but what if we wanted to infuse a little upside? Thankfully, these models are entirely customizable and can be changed or edited to fit our specific needs. We have two options. We could write down the names of our locked players and select them while in a different model, or we can adjust the sliders ourselves to see if we can target a different type of player.
I am going to go ahead and make some adjustments to the sliders. The original settings are shown on the left with the adjustments I made being shown on the right.
Among other adjustments, I eliminated the Projected Floor and Consistency sliders so I could place more emphasis on Upside and Projected Ceiling. These adjustments more resemble my girlfriend’s taste when it comes to building a pizza – sacrificing the sure thing for a chance at something great. She decided that she wanted bacon, banana peppers, and feta. Crazy, right?
Let’s optimize our lineup again and see if we can get some higher upside – more “banana pepper and feta”-type of players. Along with our completed pizza, here is how the optimized lineup comes out:
The two differences in the lineups are Kyle Lowry and DeMar DeRozan replacing Jordan Clarkson and James Harden. It is worth noting that this is being written on Thursday, when there is only four games to choose from. We would most likely see a much larger difference in players on a larger slate.
I use the optimizer as something to help point me in a direction. I would now take the time to look deeper into Kyle Lowry, DeMar DeRozan, Stanley Johnson, and Anthony Davis. Perhaps our research has us feeling uneasy about DeMar DeRozan and we think we are better off fading Anthony Davis instead of James Harden. We could exclude Anthony Davis and lock in Harden, leaving the optimized lineup to now look like this:
We could keep going if we wanted to – adjusting the sliders and excluding players until we feel we have a truly optimal lineup for the contest we plan on entering. The possibilities that are hiding within the Player Models and Lineup Tool are truly endless.
I was a little upset when I received the pizza that night. They had cut it in such a way where two of the slices had all of the toppings. I ate it anyway and was pleasantly surprised. I wouldn’t go as far to say that I will be ordering that pizza again, but I can tell you that mixing and matching two different thought processes can sometimes be a recipe for success.
Just don’t forget the cheesy bread and a few orders of the garlic butter dipping sauce.
