The injury bug and random variation
Last week was one of the most injury heavy Sundays I can remember. We saw several star players like A.J. Green, Robert Griffin III III, Ryan Mathews, and Jamaal Charles just to name a few all go down with injuries. If your lineups were simply free of injured players you probably had a pretty profitable week. However, if you rode some of these guys heavily you might not be feeling so good. But we don’t want to rely on players getting hurt or staying healthy to have a successful week.
It may seem like I keep hammering on how important bankroll management is but that’s because it is. Read John Lee’s interviews with all the DFS experts and bankroll management was the one thing they kept bringing up. So I am going that route one more time.
In my managing your DFS bankroll article I mentioned that I like to play at least a couple of different cash lineups per week. I also mentioned that if you are playing multiple lineups and you play the same player in that second lineup you really need to decrease the amount of your wager in that second game. Sunday was a perfect example of why that is so key.
The complaint I hear about this type of diversification is that you are selecting suboptimal lineups. This is true but in reality our point projections are only so accurate anyway. The difference between your #1 and even your #6 value isn’t really as big as you think it is. For example, at the WR position this difference might be about 2 points and it gets even smaller as we start to compare our #3 value to our #6. If you are accurate enough to consistently make projections within a point you don’t need a bankroll management system but for the other 99.99% of us we do.
Let’s take a closer look at how often that wide receiver that is projected to score 20 points will actually outscore the WR projected at 18 points. The key number that we need to know is the coefficient of variation (CV) for an elite WR. The CV basically tells us how volatile the position is. It is calculated by taking the standard deviation of the wide receivers scores and dividing it by their average score.
It turns out that top end wide receivers have a CV of approximately 55%. Now, using this CV and each players projected score, we can simulate scores for each of our two players thousands of times and count how often each wins. It turns out that the #1 wide receiver will beat the #6 wide receiver approximately 56% of the time. Is this lower than you expected? The first time I saw the numbers it was pretty surprising to me.
However, in practice our projected number 1 would win even less often than that because in this example we knew exactly how good each player was. When making a projection we don’t really know what the actual mean score would be for each player. So in reality our accuracy is going to be even lower than 56%.
Unfortunately, this analysis doesn’t take into consideration by how much each player will win each matchup by. Sometimes the #1 WR will score 20 and the #6 WR will score 18 just as we predicted and the difference will be 2, which isn’t that big of a deal. If we could continue to play forever that is what our about what our average score would look like if our projection was good. But in some cases the #1 WR will get shut down or injured like AJ Green did last week and score zero while your #6 WR scores 30. It’s these wild variations that we are trying to protect ourselves from and by diversifying our player pool it allows us to do so.
I don’t recommend you straying too far away from your top plays each week but I do recommend setting at least a couple of different lineups to get different player exposures. The best way to do this is by playing at multiple sites. This is the perfect route to go because players are priced differently and have different scoring methodologies at each site. By using the same projections for player A he might be only your 5th best play on Fanduel but your best play on Draftkings. This allows us to play our top tier lineup more often. It’s still going to hurt when Green goes down in the first quarter. It just won’t hurt nearly as much.
Fanduel GPP Plays
Below are my weekly GPP plays at Fanduel. Throughout the season I will be employing strategies mentioned in Week 1 about contrarian plays as well my preseason article about the best approach to allocating money in my lineups. In this article, I showed that in general it was best to spend on studs at wide receiver and to a lesser extent quarterback and save your money at running back and tight end.
Ryan Tannehill ($6,500) 1% owned – Tannehill is coming off a poor performance last week against the Bills which has scared away the majority of daily fantasy players. He is being selected by only 1.4% of the population which makes him a very attractive contrarian play. The Dolphins have picked up the pace this year currently ranking 3rd in plays attempted. This helps a guy like Tannehill who isn’t quite as explosive. Finally, the Chiefs defense made Jake Locker look good allowing 266 passing yards and two touch downs. Tannehill has a legitimate shot at 275 yards and 2 TD if everything goes well and his low salary allows you to spend heavily at the wide receiver positions.
Khiry Robinson ($4,900) 9% - Mark Ingram left last week’s game with an injury that will keep him out of this week’s game. Through two games Ingram has been running like the back we all expected when he was drafted in the first round by Sean Payton. But part of that success has been based on the passing attack opening holes for the running game. Robinson should be able to take advantage of these holes as well to put up some good per carry numbers. The Saints are projected for one of the higher scores of the week at approximately 30 points so Robinson shouldn’t have any trouble finding the end zone. In addition to this the game script should favor Robinson as the Saints are favored by 10.5 points over the Vikings. Robinson should act as the closer and rack up yards as the Saints close out the game. Finally, the Vikings are coming off a game where they allowed Stevan Ridley 25 carries for 101 yards and a touchdown. I wouldn’t be surprised to see a stat line like that from Robinson at all.
Mike Wallace ($6,500) 12% - Mike Wallace has finally started to fit in to the Dolphins offense. Through two games Wallace leads the team in targets with 19 even though one of those games was against Darrel Revis. When you throw in Wallace’s big play ability with all the targets he has been getting it becomes quite easy to imagine him finishing in the top 5 for the week. My base projection shows Wallace at 5-70-0.6. All it will take is for Wallace to break one of his customary big plays to reach value. Pairing him with Tannehill as mentioned above should be a good stack.
Dwayne Allen ($5,400) 1% - Hopefully you were following me on twitter when I mentioned Allen earlier this week but if not I think he makes for a great target this week. Allen has started the season slower than I would have hoped being targeted only 6 times in two games so far. But the key is the Jacksonville Jaguars are making just about every tight end look like a super star so far. During the first two weeks Zach Ertz and Niles Paul who aren’t exactly pro bowlers put up pro bowl level stat lines of 3-77-1 and 8-99-1. The Colts are projected to score 26 points in this game against the Jaguars and I can’t imagine them taking their foot off the pedal after their 0-2 start.
As always if you have any questions or comments email me at buzzard@footballguys.com or follow me on twitter.