I really try to keep my articles as data-driven as possible. I give anecdotes and thoughts of course, but I always want to ground myself in actionable information. If not, I’m afraid that I’m being more entertaining than useful. And that’s okay, but it has a shelf life, in my opinion.
I say this because this article will have no data. The topic of it is something I think can be researched in the future, but as of now, I don’t have access to ownership data (I know, that’s a touchy subject right now). So this article is going to be more of just spilling my thoughts on the screen and hopefully it will be useful to the both of us.
The topic at large is ownership in tournaments – on my podcast with new Fantasy Labs NBA guy Justin Phan, I mused in a very incoherent way about whether single-player ownership data was really that useful. I mean, of course it is – that combined with outcome probability is obviously the foundation of DFS research. But as you know, tournaments aren’t just about the ownership percentage of a single player; it’s about the entire lineup.
This first crossed my mind in a meaningful, couldn’t-shake-it-off way in the Week 3 DraftKings Milly Maker tournament. Here was the winning lineup.
I’ve talked about this at length before so I won’t rehash it, but obviously user “Hoothead” used a lineup with notable negative correlations – specifically, Devonta Freeman and Joseph Randle (opposing running backs) and Devonta Freeman and Julio Jones (running back and wide receiver from the same team). And it’s not like any of those three players were really low owned and thus gave him an immediate advantage like, say LeGarrette Blount did at his 2.2% ownership.
However, I would imagine that the ownership level of Freeman, Randle, and Jones together was immensely low, if not a unique lineup. And that’s where I come to my question – is multi-player ownership levels more important than single-player’s?
You may be saying, yes of course – everyone knows that having a unique lineup is important. But that’s exactly what I mean: we either think about player ownership by total lineup (so, the odds of having the same nine players/positions as someone else) or by just a single player (the odds of rostering the same guy). However, with stacking and correlations being such an important part of DFS strategy, I think that middle ground – the ownership of two or three players together – should be something on our minds perhaps more than it has been in the past.
Going back to that podcast with Justin, we brought up the idea of stacking in the NBA – it’s largely viewed as a suboptimal play in NBA DFS (probably the only DFS sport where stacking is suboptimal in theory), but we both agreed we liked the strategy, particularly in GPP tournaments. We also came up with an important distinction in finding players with different roles – for example, a Chris Paul-DeAndre Jordan stack has a lot more upside than a Chris Paul-Jamal Crawford stack because of team roles.
The reason that it could be an interesting strategy in tournaments is because of the public’s view that it isn’t optimal, when in fact, that opens up the door of exploiting multi-player ownership issues. The Clippers play at a fast pace and it’s not ridiculous for both Paul and Jordan to have games of 40-plus fantasy points. And in general, in those spots, both of those players are going to be highly owned.
And let’s say that they are in a tournament. We’ll assign them ownership levels of 25% each for our hypothetical tournament. A lot of people will roster either Chris Paul or DeAndre Jordan because they want exposure to that game, but what we also need to be concerned with is how many people will roster both Paul and Jordan. Is it 2%? Because, if so, all of a sudden that becomes a highly optimal tournament play on any night.
In the NFL, you can think about it in the same way. Take Matthew Stafford, Calvin Johnson, and Golden Tate this past week. They all had good games and a lot of smart DFS players were rightfully on all three players in different lineups. However, when picking your stacks, you have to think about not only the ownership of those three individual guys but the ownership of a Stafford-Johnson stack versus the ownership of a Stafford-Tate stack. One may have more fantasy production upside, but the other may have more upside in terms of being lower owned and thus winning a tournament. And to take it even further, what was the percentage of lineups that owned all three? Again, no data, but I’d guess low.
I apologize again if you’re head hurts or you think this is a dumb idea or you’re incredibly offended by my lack of data. However, this is something that I’ve been really thinking about lately and I’ve always said that I would be transparent and open about my thoughts regarding DFS strategy and such – whether those thoughts are good or bad. So thanks for reading them and as always, please let me know your specific thoughts on this topic as well.
