DraftKings Week 1 Recap

Every week Draft Kings pays out big prizes to winners of their contests. Most sites focus on what people did, and pick apart the best lineup. I’m not going to do much of that here. Instead I’ll show you what we’re the best lineups possible and analyze how famous players like CONDIA and CSURAM88 constructed their lineups.

Most surprisingly, they don’t use “locks” and employed far less stacks then I’d expect.

Obligatory who won section

For this week it was Bearsfan247 topping out at 198.88 points.

Draft Kings Week 1 Winner

He won the million dollar jackpot and yes we’re all jealous.

How Bearsfan247 Constructed his Lineups

Many top players use lineup optimizers to construct their lineups. They use constraints, look for QB-WR stacks, and limit exposure to players. Looking at Bearsfan247 we see that he entered 22 lineups.

count 22.000000
mean 114.011818
std 28.805900
min 66.920000
25% 97.729993
50% 112.550000
75% 127.395000
max 198.880000

This funky looking chart shows a histogram of point totals and the individual observations. A continuous line approximates the distribution to demonstrate that this is not your typical bell curve. That 1 total at 198.88 is what got Bearsfan247 $1,000,000.

Appallingly, he had the Bears in 36% of his lineups. I’m a Bears fan too but man that’s ridiculous. Thankfully Tarik Cohen saved his day.

Percentage Players
0 40 Todd Gurley II
1 40 Christian McCaffrey
2 36 Bears
3 27 Zach Ertz
4 27 Antonio Brown
5 27 Pierre Garcon
6 27 Brandon Marshall
7 27 DeMarco Murray

As for as his stacks, there really weren’t any.

Percentage Players
0 22 Bears , Todd Gurley II
1 18 Todd Gurley II, Zach Ertz
2 18 Antonio Brown, Christian McCaffrey
3 18 DeMarco Murray, Russell Wilson
4 13 Antonio Brown, Todd Gurley II
5 13 Antonio Brown, Bears
6 13 Carson Wentz, Todd Gurley II
7 13 Antonio Brown, Carson Wentz

I’m going to go out and say it. Even a blind squirrel get’s a nut every once in a while. He may have used a lineup generator grabbing the max expected value players. I’d be very surprised if he wins again.

His lineups don’t look to follow any deliberate pattern or anything. But anyways, well done Bearsfan247. Hopefully you don’t blow your money trying to win another jackpot


CSURAM88 entered 150 lineups in the contest!

count 150.000000
mean 110.411466
std 20.264872
min 64.120000
25% 96.259997
50% 107.710000
75% 123.450000
max 169.580000

Unfortunately, CSURAM’s we’re much more “normal” then you’d want in a large tourney. CSURAM88's Point Totals

He limited his individual player exposure to ~25%. Betting on players like Christian MCCaffrey and Sterling Shepard to have great days. Personally, I think he bought too much into the offseason hype and overlooked a not bad 49er’s defense at home, and a bad offense in the New York Giants. But hey, he’s the winner and not me.

Percentage Player
0 26 Christian McCaffrey
1 26 Jimmy Graham
2 25 Sterling Shepard
3 25 Pierre Garcon
4 24 Rams
5 24 Le’Veon Bell
6 24 Ameer Abdullah
7 23 Delanie Walker
8 22 Amari Cooper
9 22 Doug Baldwin
10 20 Rishard Matthews
11 20 A.J. Green

From looking at these lineups, I’d wager he used some sort of max expected value optimizer with player exposure limits, never face your own defense, and a variety of different stacks at the ~8% range.

  • Russell Wilson, Jimmy Graham 8%
  • Ben Roethlisberger-Antonio Brown 8%
  • Derek Carr-Michael Crabtree 8%
  • Marcus Mariota-Rishard Matthews 9%

He was really hoping one of those stacks would have a monster day but it just didn’t happen this week. In week 2, QB’s should fare much better.

Overall, CSURAM88 wagered $3,000 and lost $1,855 in this contest. Not terrible and he will live to win again.


Like CSURAM88, CONDIA entered 150 lineups. He spent $3,000 and lost $2050. Topping out at 163.58


CONDIA's Point Totals

His point totals were comparable to CSURAM’s in terms of mean and standard deviation.

count 150.000000
mean 112.647200
std 20.020005
min 70.560000
25% 98.645000
50% 112.480000
75% 127.140000
max 163.580000

He had even less QB-WR stacks the CSURAM with his highest percentage stack at 6% for Matthew Stafford-Marvin Jones Jr.

In terms of individual player exposure, CONDIA capped out at 27%.

Percentage Player
0 27 Le’Veon Bell
1 26 Zach Ertz
2 26 David Johnson
3 26 Carlos Hyde
4 25 Antonio Brown
5 24 Lamar Miller
6 22 Todd Gurley II
7 20 Steelers
8 18 Rob Kelley
9 17 Ezekiel Elliott
10 17 Larry Fitzgerald
11 17 DeMarco Murray

He used a player pool of 97 players, which leads me to believe he is using some form of lineup optimizer, probably by expected value.

Looking back at the Best Possible Lineups

Using the score totals of last week games we can look at the best possible lineups. The highest score possible was 229.98 points.

player_name position score salary team_name
0 LA Rams DST 28.00 3200.0 LAR
1 Jordy Nelson WR 20.90 7600.0 GB
2 Austin Hooper TE 23.80 3000.0 ATL
3 Matthew Stafford QB 28.08 6100.0 DET
4 Golden Tate WR 24.40 6500.0 DET
5 Leonard Fournette RB 24.40 5700.0 JAC
6 Antonio Brown WR 32.20 8800.0 PIT
7 Kenny Golladay FLEX 22.90 3600.0 DET
8 Tarik Cohen RB 25.30 3000.0 CHI

There were at least 956 better lineups than Bearsfan247 and they would have used

Best Lineups

Note the locks for Tarik Cohen and Leonard Fournette. As far as QB-WR stacks go, this week that wasn’t a good pick. Unless you had the Thursday slate. That’s not to say stacks are bad but when they miss they miss, and likewise when they hit, they hit.

This week QB-WR’s didn’t do well together and chances taken on individual players/defenses were better risk reward ratio.

Programming Category (English)468x60

The Herd Picks

No huge stacks this week that I could find.

Percentage Players
0 41 Zach Ertz
1 33 Todd Gurley II
2 25 Le’Veon Bell
3 22 David Johnson
4 22 Doug Baldwin
5 21 Rams
6 20 Larry Fitzgerald
7 19 Christian McCaffrey
8 17 Carlos Hyde
9 16 Delanie Walker
10 15 Amari Cooper
11 14 Michael Crabtree
12 13 Antonio Brown
13 13 Terrelle Pryor Sr.
14 13 Pierre Garcon
15 13 Kelvin Benjamin
16 11 Rishard Matthews
17 11 Brandon Marshall
18 10 A.J. Green
19 10 Texans

Zach Ertz had a surprisingly high utilization, at least I thought so. People really loved Christian McCaffrey this week against a tough matchup. Yes, the 49er’s aren’t terrible at D.

I couldn’t find many big stacks besides Todd Gurley and the Rams D at below 10%.

Closing Thoughts

I think it’s important to pay attention to what the best lineups were rather than study what people did. A lot of players in the best lineups weren’t used at all by the winners. Leonard Fournette had a great day and most people didn’t take a chance on him. 1% utilization. Tarik Cohen was used only by 413 lineups! Or just 0.15%. I’d expect that to change next week.

And sometime this season, the traditional stacks will hit.

Matteo Hoch

Matteo Hoch is the founder of Sports Data Direct and maintains a personal blog called Ergo Sum where he writes about data visualization and coding. You can generally find him in the DraftKings Sunday $3 early only contests under username mhoch2

Leave a Reply

Your email address will not be published. Required fields are marked *