Week 1 NFL Daily Fantasy Football Results - How my Bayesian Inspired Lineups Fared (+$84.00)

in #sports8 years ago (edited)

Week 1 is in the books! And it was profitable!

Despite only being able to predict approximately 21% of the variance of Week 1 with my model, I was able to still cash at approximately a 382% profit with the insights I gained.

How exactly did I accomplish that?

I'm pretty sure this week it was just blind luck. This is because I decided to take a very conservative approach and just dip my toes in the water with my projections to start the year out. This is also especially true knowing my most profitable weeks backtesting my model existed in weeks 12-17 in past seasons.

I also am trying to get more familiar with how to best build my portfolio of lineups. I wanted to experiment with single entry, head to head, and early game only entries to get a feel for what might be best for me.

Therefore, I decided to spread my buy-ins approximately evenly across cash games and Guaranteed Prize Pool (GPP) tournaments. I only entered $22 in fees across seven games and averaged scoring 150.30 points at a 57.14% cash rate across all lineups. It was, however, one specific lineup that won me the most -- a 1:00 PM only lineup where I had all my best hits. I flirted with placing much higher, but had a TE that suffered from the extreme lack of a passing game.

I exclusively played on Draft Kings for every lineup, with my username of twigsta26. Listed below are my results:

So, was the algorithm any good?

I think the box plot below is pretty telling:

I had a median algorithm difference of underestimating a player's performance by 1.3 points, with the majority of my quartile data falling between a 4.9 point underestimate and a 1.3 point underestimate.

In other words, I'm getting pretty good at predicting average performances. I have a lot more work to do when it comes to predicting breakouts for GPP tourneys. I do think I underestimated how good my lineups could be for cash game matchups. This will be an angle I will try to use in weeks to come.

The Detailed List - How the Model Fared for Each Player

Many players were injured, and some didn't play. Regardless, below is the good, the bad, and the ugly. An underestimate for a player is marked in parentheses. I hope to improve on my accuracy here in the weeks to come!

MyRankPlayerPosTeamOppMyModelActualDifference
1Cam NewtonQBCARDEN23.822.16(1.7)
2Antonio BrownWRPITWAS22.835.612.8
3Drew BreesQBNOOAK22.735.4212.7
4Andrew LuckQBINDDET22.538.516.0
5Russell WilsonQBSEAMIA21.813.92(7.9)
6Julio JonesWRATLTB20.316.6(3.7)
7Ben RoethlisbergerQBPITWAS20.225.85.6
8Eli ManningQBNYGDAL20.219.28(0.9)
9Matthew StaffordQBDETIND19.929.19.2
10Blake BortlesQBJAXGB19.719(0.7)
11Odell Beckham JrWRNYGDAL19.511.3(8.2)
12Joe FlaccoQBBALBUF19.513.02(6.5)
13Brandon MarshallWRNYJCIN19.46.2(13.2)
14Aaron RodgersQBGBJAX19.123.564.5
15Arian FosterRBMIASEA18.913(5.9)
16Ryan FitzpatrickQBNYJCIN18.816.06(2.8)
17Philip RiversQBSDKC18.814.62(4.2)
18Kirk CousinsQBWASPIT18.714.96(3.8)
19Ryan TannehillQBMIASEA18.515.14(3.4)
20Jameis WinstonQBTBATL18.526.548.1
21Jordy NelsonWRGBJAX17.815.2(2.6)
22Derek CarrQBOAKNO17.625.367.8
23Tyrod TaylorQBBUFBAL17.35.54(11.8)
24Sam BradfordQBMINTEN17.10(17.1)
25Matt RyanQBATLTB17.027.3610.4
26Carson PalmerQBARINE16.718.942.3
27Demaryius ThomasWRDENCAR16.38.8(7.5)
28Jay CutlerQBCHIHOU16.312.84(3.4)
29Alex SmithQBKCSD16.132.0216.0
30Sammy WatkinsWRBUFBAL16.08.3(7.7)
31Matt ForteRBNYJCIN16.020.54.5
32Adrian PetersonRBMINTEN15.43.1(12.3)
33Jamaal CharlesRBKCSD15.40(15.4)
34Mark SanchezQBDALNYG15.10(15.1)
35DeAndre HopkinsWRHOUCHI15.116.41.3
36Jordan MatthewsWRPHICLE15.027.512.5
37Andy DaltonQBCINNYJ15.021.346.4
38Rashad JenningsRBNYGDAL14.98.8(6.1)
39Robert Griffin IIIQBCLEPHI14.810.3(4.5)
40Emmanuel SandersWRDENCAR14.49.9(4.5)
41Zach ErtzTEPHICLE14.411.8(2.6)
42Doug BaldwinWRSEAMIA14.224.210.0
43Mike EvansWRTBATL14.020.96.9
44AJ GreenWRCINNYJ14.03925.0
45Marcus MariotaQBTENMIN13.718.745.0
46Devonta FreemanRBATLTB13.78(5.7)
47Golden TateWRDETIND13.611.3(2.3)
48Brandin CooksWRNOOAK13.436.423.0
49Eric DeckerWRNYJCIN13.311.7(1.6)
50Keenan AllenWRSDKC13.312.3(1.0)
51Jeremy MaclinWRKCSD13.317.23.9
52Mark IngramRBNOOAK13.210.7(2.5)
53Randall CobbWRGBJAX13.212.8(0.4)
54Jordan ReedTEWASPIT13.113.40.3
55Todd GurleyRBLASF13.05.2(7.8)
56Delanie WalkerTETENMIN12.87.2(5.6)
57LeSean McCoyRBBUFBAL12.7174.3
58Lamar MillerRBHOUCHI12.718.76.0
59David JohnsonRBARINE12.623.210.6
60Tim HightowerRBNOOAK12.52(10.5)
61TY HiltonWRINDDET12.513.91.4
62Case KeenumQBLASF12.54.2(8.3)
63Allen RobinsonWRJAXGB12.513.20.7
64Blaine GabbertQBSFLA12.514.52.0
65Jarvis LandryWRMIASEA12.312.90.6
66Rob GronkowskiTENEARI12.30(12.3)
67Dez BryantWRDALNYG12.11.8(10.3)
68Steve SmithWRBALBUF12.16.9(5.2)
69Amari CooperWROAKNO12.124.712.6
70Brock OsweilerQBHOUCHI12.017.645.6
71Alshon JefferyWRCHIHOU12.017.55.5
72Allen HurnsWRJAXGB11.811.5(0.3)
73Eddie LacyRBGBJAX11.88.8(3.0)
74Frank GoreRBINDDET11.711.80.1
75Danny WoodheadRBSDKC11.62311.4
76Anquan BoldinWRDETIND11.56.5(5.0)
77Michael CrabtreeWROAKNO11.517.76.2
78Julian EdelmanWRNEARI11.415.23.8
79Darren McFaddenRBDALNYG11.30(11.3)
80Terrance WilliamsWRDALNYG11.26.4(4.8)
81Larry FitzgeraldWRARINE11.228.116.9
82Gary BarnidgeTECLEPHI11.20(11.2)
83Alfred MorrisRBDALNYG11.23.5(7.7)
84DeMarco MurrayRBTENMIN11.023.712.7
85Jason WittenTEDALNYG11.015.64.6
86Michael FloydWRARINE11.09.1(1.9)
87DeAngelo WilliamsRBPITWAS10.938.127.2
88Latavius MurrayRBOAKNO10.914.23.3
89Isaiah CrowellRBCLEPHI10.815.85.0
90Jimmy GrahamTESEAMIA10.82.1(8.7)
91Doug MartinRBTBATL10.714.63.9
92Pierre GarconWRWASPIT10.611.10.5
93DeSean JacksonWRWASPIT10.619.28.6
94Greg OlsenTECARDEN10.614.33.7
95John BrownWRARINE10.51.8(8.7)
96Kamar AikenWRBALBUF10.53.4(7.1)
97Thomas RawlsRBSEAMIA10.48.8(1.6)
98Antonio GatesTESDKC10.35(5.3)
99Dorial Green-BeckhamWRPHICLE10.33.4(6.9)
100TJ YeldonRBJAXGB10.316.96.6
101Markus WheatonWRPITWAS10.10(10.1)
102Tavon AustinWRLASF10.15.5(4.6)
103Cecil ShortsWRHOUCHI10.00(10.0)
104Jermaine KearseWRSEAMIA9.910.70.8
105Jeremy HillRBCINNYJ9.99.1(0.8)
106Charles SimsRBTBATL9.913.13.2
107Bilal PowellRBNYJCIN9.86.8(3.0)
108Tyler LockettWRSEAMIA9.84.7(5.1)
109Cole BeasleyWRDALNYG9.814.54.7
110Darren SprolesRBPHICLE9.75.6(4.1)
111Robert WoodsWRBUFBAL9.66(3.6)
112Willie SneadWRNOOAK9.635.225.6
113Andrew HawkinsWRCLEPHI9.50(9.5)
114Ronnie HillmanRBDENCAR9.50(9.5)
115Ted GinnWRCARDEN9.43.5(5.9)
116Shane VereenRBNYGDAL9.49.1(0.3)
117Kenny BrittWRLASF9.410.71.3
118Giovani BernardRBCINNYJ9.35(4.3)
119Kelvin BenjaminWRCARDEN9.321.111.8
120Duke JohnsonRBCLEPHI9.38(1.3)
121Vincent JacksonWRTBATL9.23.8(5.4)
122Marvin JonesWRDETIND9.112.53.4
123Javorius AllenRBBALBUF9.00(9.0)
124Jonathan StewartRBCARDEN8.96.4(2.5)
125Torrey SmithWRSFLA8.83.3(5.5)
126Stefon DiggsWRMINTEN8.720.211.5
127Victor CruzWRNYGDAL8.613.44.8
128Brandon LaFellWRCINNYJ8.613.14.5
129Julius ThomasTEJAXGB8.617.48.8
130Martellus BennettTENEARI8.54.4(4.1)
131Andre JohnsonWRTENMIN8.56(2.5)
132Mike WallaceWRBALBUF8.419.210.8
133Charcandrick WestRBKCSD8.48.3(0.1)
134James StarksRBGBJAX8.32.3(6.0)
135Travis BenjaminWRSDKC8.310.21.9
136Theo RiddickRBDETIND8.227.819.6
137Devante ParkerWRMIASEA8.20(8.2)
138Seth RobertsWROAKNO8.19.91.8
139Jamison CrowderWRWASPIT8.111.83.7
140Coby FleenerTENOOAK8.11.6(6.5)
141Travis KelceTEKCSD8.013.45.4
142Kyle RudolphTEMINTEN8.010.52.5
143CJ AndersonRBDENCAR8.029.921.9
144Eric EbronTEDETIND8.015.67.6
145Davante AdamsWRGBJAX7.9146.1
146Brandon ColemanWRNOOAK7.80(7.8)
147Dontrelle InmanWRSDKC7.81.6(6.2)
148Chris IvoryRBJAXGB7.80(7.8)
149Eddie RoyalWRCHIHOU7.615.78.1
150Rishard MatthewsWRTENMIN7.65.6(2.0)
151Chris JohnsonRBARINE7.50.2(7.3)
152Alfred BlueRBHOUCHI7.50.4(7.1)
153Mohamed SanuWRATLTB7.42113.6
154Richard RodgersTEGBJAX7.33.2(4.1)
155Brice ButlerWRDALNYG7.32.6(4.7)
156Devin FunchessWRCARDEN7.31.9(5.4)
157Charles ClayTEBUFBAL7.36(1.3)
158Will TyeTENYGDAL7.14.6(2.5)
159Ameer AbdullahRBDETIND7.02316.0
160Andre HolmesWROAKNO7.00(7.0)
161Andre EllingtonRBARINE6.92.8(4.1)
162Nelson AgholorWRPHICLE6.915.78.8
163Jerick McKinnonRBMINTEN6.81.9(4.9)
164Ryan MathewsRBPHICLE6.813.76.9
165Spencer WareRBKCSD6.835.929.1
166Marquess WilsonWRCHIHOU6.80(6.8)
167Shaun HillQBMINTEN6.810.243.4
168Jared CookTEGBJAX6.71.7(5.0)
169Lance KendricksTELASF6.73.5(3.2)
170Mychal RiveraTEOAKNO6.70(6.7)
171James WhiteRBNEARI6.69.42.8
172Melvin GordonRBSDKC6.617.711.1
173Zach MillerTECHIHOU6.54.4(2.1)
174JJ NelsonWRARINE6.42.1(4.3)
175Larry DonnellTENYGDAL6.48.52.1
176Matt JonesRBWASPIT6.44.3(2.1)
177Donte MoncriefWRINDDET6.318.412.1
178Albert WilsonWRKCSD6.33.1(3.2)
179Kendall WrightWRTENMIN6.20(6.2)
180Danny AmendolaWRNEARI6.27.81.6
181Austin Seferian-JenkinsTETBATL6.2103.8
182CJ SpillerRBNOOAK6.10(6.1)
183Jaelen StrongWRHOUCHI6.00(6.0)
184Benjamin CunninghamRBLASF6.01.8(4.2)
185Corey BrownWRCARDEN6.02.1(3.9)
186Cameron Artis-PayneRBCARDEN5.90(5.9)
187Justin HardyWRATLTB5.91(4.9)
188Crockett GillmoreTEBALBUF5.81.5(4.3)
189Dennis PittaTEBALBUF5.76.91.2
190Marqise LeeWRJAXGB5.64.2(1.4)
191Shaun DraughnRBSFLA5.510.65.1
192Jarius WrightWRMINTEN5.40(5.4)
193Kenny StillsWRMIASEA5.32.6(2.7)
194Tevin ColemanRBATLTB5.316.711.4
195Dwayne AllenTEINDDET5.217.312.1
196Carlos HydeRBSFLA5.223.318.1
197Terrance WestRBBALBUF5.25.80.6
198Paul RichardsonWRSEAMIA5.22.1(3.1)
199Maxx WilliamsTEBALBUF5.20(5.2)
200Brent CelekTEPHICLE5.22.1(3.1)
201Vance McDonaldTESFLA5.19.44.3
202Jacob TammeTEATLTB5.111.16.0
203Quinton PattonWRSFLA5.0116.0
204Clive WalfordTEOAKNO4.95.50.6
205Chris ThompsonRBWASPIT4.911.97.0
206Jeremy LangfordRBCHIHOU4.814.39.5
207Quincy EnunwaWRNYJCIN4.819.314.5
208Phillip DorsettWRINDDET4.813.48.6
209Tyler KroftTECINNYJ4.70(4.7)
210LeGarrette BlountRBNEARI4.7127.3
211Darren FellsTEARINE4.60(4.6)
212Jeremy KerleyWRSFLA4.613.18.5
213Chris HoganWRNEARI4.61510.4
214Dwayne HarrisWRNYGDAL4.50(4.5)
215Chris ConleyWRKCSD4.58.33.8
216Luke WillsonTESEAMIA4.44.60.2
217Jordan CameronTEMIASEA4.42.6(1.8)
218TJ JonesWRDETIND4.30(4.3)
219Christine MichaelRBSEAMIA4.19.15.0
220Rashad GreeneWRJAXGB4.01.9(2.1)
221Jack DoyleTEINDDET3.918.514.6
222Charles JohnsonWRMINTEN3.81.5(2.3)
223Garrett CelekTESFLA3.82.5(1.3)
224Jimmy GaroppoloQBNEARI3.816.0612.2
225Ryan GriffinTEHOUCHI3.83.7(0.1)
226Jace AmaroTETENMIN3.70(3.7)
227Cameron BrateTETBATL3.662.4
228Josh HillTENOOAK3.50(3.5)
229Dion SimsTEMIASEA3.40(3.4)
230Robert TurbinRBINDDET3.23.90.7
231Jay AjayiRBMIASEA3.00(3.0)
232Jesse JamesTEPITWAS2.58.15.6
233Virgil GreenTEDENCAR2.36.84.5
234Ryan HewittTECINNYJ2.20(2.2)
235Bennie FowlerWRDENCAR2.10(2.1)

Sort:  

This is awesome. I'm no good at daily fantasy... tried a few times, with no success. I don't have any real strategy or analysis to fall back on.

I do love the game, though! I try to play a traditional fantasy football league every year. I've been playing with the same group of guys for a few years now... it would be really nice to win it one year!

Fantasy football is the best! So much is luck though, strategy helps a ton, but you can't project injuries. Just putting yourself in the best spot possible is all you can do!

WOW awesome stuff here! DFS is back to legal status in NY, so I am dabbling again in it. I enjoyed this article a bunch!!! FOLLOWED and UPVOTED WHOOHOO!

-bigedude

Just curious, what is the adjusted R squared? Or is that irrelevant with the number of variables in the algorithm? Sorry, I'm not a very math-y person just wondering.

If you get a 300% profit at .21 R^2 can't help but wonder what the profit percentage will be at .65 R^2.

Good question! For this post, I only applied regression testing to predicted v. actual. I am away from my database currently to see the actual coefficient, but it was still around the .21 mark, slightly under. My skew this week happened mainly with inactives and injuries. All those dots on the x-axis really didn't help with my overall accuracy. I am going to work on squishing those out in weeks to come.

In backtesting, my highest tested r^2 was about 0.40, but this has only happened after I've had a half season or more of data to work through.

Barnidge!!! He does a lot better with McCown FWIW (so don't be afraid to try him again), but if he scored a TD I think you might have won the $3 gpp.

Congrats on the win

Thanks! I couldn't explain the ways I was screaming at my TV about Barnidge. I had him as marginally worse with RGIII so I ran with it. Just got a bad break. I was pleased overall though with my lineup structuring. Going to focus more on cash games this week and see how that plays out.

I had all the barnidge :(

Awesome to see all the growing dfs content here though!

Nice. You should do a blog post on the practical applications of Bayesian probability.. I always thought it would be really useful when learning about years ago, but I've never used it and forgotten all about it.

Coin Marketplace

STEEM 0.16
TRX 0.13
JST 0.027
BTC 60589.35
ETH 2628.62
USDT 1.00
SBD 2.53