Leverage: How to increase your own odds in DFS

By now, you have to wonder why my entire operation is built on finding leverage. Every single post I make on twitter (@dfsedric) has leverage in mind actually.

Recently, I saw a movie on netflix called 21, which is based on blackjack. I feel that this one clip answers everything you need to know about DFS leverage.

How can we relate this to DFS? Let's look at a very recent example of a Nascar Draftkings contest. Going into the Nascar QuikTrip 500 at Atlanta, all the research that you could have done points to playing the cars in the back of the field. We expected wrecks and Daytona style conditions - and we got all of that. Because we want to find drivers who were starting in the back of the race and collecting on place differential points, all roads pointed to playing Ricky Stenhouse Jr before the race. Infact here are a few data points for those drivers starting from the back, before the race (paired with the actual results)

He was so popular that 45% of the 117,000 lineups entered were projected to play RIcky. It makes sense why he would be popular, as you would earn points as long as he finishes the race higher than 28th. 40-1 odds would indicate he finishes higher than 28th as well. But as the video concludes - but this is where variable change should kick in.

Prior to the contest, it was estimated that Ricky had about 10% chance of winning and rougly 45% projected ownership. Without naming him Chalky, he was the artist formally known as "Wrecky Stenhouse". At the point in which you know that 40% of all lineups are going to be on one player, you should re-evaluate. That is why the following chart matters from my twitter posts:

https://twitter.com/dfsedric/status/1505605501949812738/photo/1

While Chalky is on the bottom left of the 7k bubble, some other good plays were within reach. Bubba Wallace was a direct pivot point. He had better odds of success and was half the ownership of Stenhouse. He scored over double the points of Wrecky so that was definitely a better play. Ross Chastain was also out there and he finished 2nd . You can see his opportunity based on the green in his dot.

To further push the point on variable change, we can look at the same data sorted by salary, in order to see how many actual points were scored by those around Wrecky:

I mentioned before that if you initially picked Stenhouse, he had roughly 10% odds of success going into this race. Thats door number 1. Then we learn that he's 40% owned and as a gameshow host, we remove Wrecky and pick anyone else in that range. All but 3 drivers within his salary radius appear to have done better than Wrecky, which means getting off Wrecky would increase your chances of improving by 78% in this case.

As fate would have it, I chose the wrong door as well and took Cindric to pivot off of Wrecky. Wrecky blows a tire and takes out Cindric, so I missed the cash line by 1 point (-6 scored by Cindric in the process so that burns). THAT is karma and bad luck but as you can how many other chances I had to get 1 point and more by playing the leverage.

Ill end this by showing the overall winner of the large field GPP. This guy clearly was going for leverage and avoided pretty much all chalk as only 1 player is owned above 20%. I don't feel like he just randomly picked these guys. Instead he most likely thought to himself that many unknown variables are likely to happen today with this new track. Therefore I need to be unlike everyone else. Thats what we should strive for when building lineups. Thats what we are consistantly searching for with DFSed.