Fantasy Football Team
Joshua Peterson
9/21/2016
Player analysis and selection for the week of 09.19.2016
QB Analysis - Pull Player Data
QB Analysis
Cam Newton - Summary Passing Data
Newton Yards Summary
## Location Opp Def.Rank.Yds PassYds PassTD
## Away:42 ATL :11 Min. : 1.00 Min. :114.0 Min. :0.000
## Home:42 NOR :11 1st Qu.:10.00 1st Qu.:194.0 1st Qu.:1.000
## TAM : 8 Median :18.00 Median :231.5 Median :1.000
## MIN : 4 Mean :18.63 Mean :235.4 Mean :1.488
## SEA : 4 3rd Qu.:28.00 3rd Qu.:277.0 3rd Qu.:2.000
## ARI : 3 Max. :32.00 Max. :432.0 Max. :5.000
## (Other):43
## Int RushYds RushTD
## Min. :0.0000 Min. : 4.00 Min. :0.0000
## 1st Qu.:0.0000 1st Qu.: 23.75 1st Qu.:0.0000
## Median :1.0000 Median : 39.00 Median :0.0000
## Mean :0.8333 Mean : 40.67 Mean :0.5476
## 3rd Qu.:1.0000 3rd Qu.: 53.25 3rd Qu.:1.0000
## Max. :4.0000 Max. :116.00 Max. :3.0000
## Exploratory data analysis
Location analysis
par(mfrow = c(5,2), mar = c(4,4,1,1))
plot(CNdf$Location, CNdf$PassYds, ylab = "Passing Yards", col = "red")
plot(CNdf$Location, CNdf$RushYds, ylab = "Rushing Yards", col = "blue")
plot(CNdf$Location, CNdf$PassTD, ylab = "Passing TDs", col = "green")
plot(CNdf$Location, CNdf$RushTD, ylab = "Rushing TDs", col = "purple")
plot(CNdf$Location, CNdf$Int, ylab = "Interceptions", col = "grey")
plot of chunk unnamed-chunk-63
Opponent Analysis
Opponent vs. Pass Yards
par(mfrow = c(5,1), mar = c(4,4,1,1))
plot(CNdf$Opp, CNdf$PassYds, ylab = "Yards Passing", xlab = "Opponent", col = "red")
plot(CNdf$Opp, CNdf$RushYds, ylab = "Yards Rushing", xlab = "Opponent", col = "blue")
plot(CNdf$Opp, CNdf$PassTD, ylab = "Passing TDs", xlab = "Opponent", col = "green")
plot(CNdf$Opp, CNdf$RushTD, ylab = "Rushing TDs", xlab = "Opponent", col = "purple")
plot(CNdf$Opp, CNdf$Int, ylab = "Interceptions", xlab = "Opponent", col = "grey")
plot of chunk unnamed-chunk-64
Defensive Ranking Analysis
Defense rank vs. Pass Yards
library(ggplot2)
par(mfrow = c(5,2), mar = c(4,4,1,1))
qplot(Def.Rank.Yds,PassYds, data = CNdf) + geom_smooth(method = "lm")
plot of chunk unnamed-chunk-65
qplot(Def.Rank.Yds, RushYds, data = CNdf) + geom_smooth(method = "lm")
plot of chunk unnamed-chunk-65
qplot(Def.Rank.Yds, PassTD, data = CNdf) + geom_smooth(method = "lm")
plot of chunk unnamed-chunk-65
qplot(Def.Rank.Yds, RushTD, data = CNdf) + geom_smooth(method = "lm")
plot of chunk unnamed-chunk-65
qplot(Def.Rank.Yds, Int, data = CNdf) + geom_smooth(method = "lm")
plot of chunk unnamed-chunk-65
Newton Yards LSE 
Linear model fitting
Model fitting vs. Location
Pass Yds vs. Location
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 231.476190 10.01434 23.1144679 1.146648e-37
## LocationHome 7.785714 14.16242 0.5497447 5.839887e-01Prediction interval using a linear model
## fit lwr upr
## 1 231.4762 100.8408 362.1116
## 2 231.4762 100.8408 362.1116
## 3 231.4762 100.8408 362.1116
## 4 231.4762 100.8408 362.1116
## 5 231.4762 100.8408 362.1116
## 6 231.4762 100.8408 362.1116Rush Yds. vs. Location
CNfitRushYdsLoc<- lm(RushYds ~ Location, data = CNdf)
summary(CNfitRushYdsLoc)$coefficients## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 42.809524 3.719746 11.5087210 8.455620e-19
## LocationHome -4.285714 5.260516 -0.8146947 4.176069e-01Prediction interval using a linear model
## fit lwr upr
## 1 42.80952 -5.713951 91.333
## 2 42.80952 -5.713951 91.333
## 3 42.80952 -5.713951 91.333
## 4 42.80952 -5.713951 91.333
## 5 42.80952 -5.713951 91.333
## 6 42.80952 -5.713951 91.333Pass TDs. vs. Location
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.4285714 0.1792766 7.9685335 8.064972e-12
## LocationHome 0.1190476 0.2535354 0.4695503 6.399225e-01Prediction interval using a linear model
## fit lwr upr
## 1 1.428571 -0.9100618 3.767205
## 2 1.428571 -0.9100618 3.767205
## 3 1.428571 -0.9100618 3.767205
## 4 1.428571 -0.9100618 3.767205
## 5 1.428571 -0.9100618 3.767205
## 6 1.428571 -0.9100618 3.767205Rush TDs vs. Location
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.6190476 0.09991424 6.19579 2.207786e-08
## LocationHome -0.1428571 0.14130007 -1.01102 3.149811e-01Prediction interval using a linear model
## fit lwr upr
## 1 0.6190476 -0.684317 1.922412
## 2 0.6190476 -0.684317 1.922412
## 3 0.6190476 -0.684317 1.922412
## 4 0.6190476 -0.684317 1.922412
## 5 0.6190476 -0.684317 1.922412
## 6 0.6190476 -0.684317 1.922412Model fitting vs. opponent
Pass Yards vs. opponent data
CNfitPassYds<- lm(PassYds ~ Opp, data = CNdf)
summary(CNfitPassYds)$coefficients## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 314.0000000 37.57596 8.356406770 3.029459e-11
## OppATL -96.6363636 42.39140 -2.279621941 2.668241e-02
## OppBAL -117.0000000 75.15192 -1.556846484 1.254588e-01
## OppBUF -85.0000000 75.15192 -1.131042318 2.631294e-01
## OppCHI 0.3333333 53.14043 0.006272688 9.950187e-01
## OppCIN -30.0000000 75.15192 -0.399191406 6.913562e-01
## OppCLE -113.0000000 75.15192 -1.503620963 1.386149e-01
## OppDAL -106.0000000 59.41281 -1.784127080 8.012798e-02
## OppDEN -96.5000000 59.41281 -1.624228898 1.102602e-01
## OppDET -33.5000000 59.41281 -0.563851483 5.752335e-01
## OppGNB -2.6666667 53.14043 -0.050181502 9.601664e-01
## OppHOU -142.0000000 59.41281 -2.390057032 2.043314e-02
## OppIND -86.0000000 59.41281 -1.447499329 1.536467e-01
## OppJAX -147.5000000 59.41281 -2.482629663 1.624279e-02
## OppKAN -82.0000000 75.15192 -1.091123177 2.801543e-01
## OppMIA -140.0000000 75.15192 -1.862893229 6.802323e-02
## OppMIN -67.0000000 49.70832 -1.347862847 1.834357e-01
## OppNOR -80.4545455 42.39140 -1.897897853 6.316179e-02
## OppNWE -105.0000000 75.15192 -1.397169922 1.681848e-01
## OppNYG -45.6666667 53.14043 -0.859358227 3.940132e-01
## OppNYJ -41.0000000 75.15192 -0.545561589 5.876562e-01
## OppOAK -144.0000000 75.15192 -1.916118750 6.075031e-02
## OppPHI -44.3333333 53.14043 -0.834267475 4.078749e-01
## OppPIT -64.0000000 75.15192 -0.851608333 3.982631e-01
## OppSDG -83.0000000 75.15192 -1.104429557 2.743958e-01
## OppSEA -137.5000000 49.70832 -2.766136441 7.791241e-03
## OppSFO -53.0000000 59.41281 -0.892063540 3.763912e-01
## OppSTL -110.0000000 75.15192 -1.463701823 1.491815e-01
## OppTAM -85.1250000 44.06172 -1.931949189 5.871932e-02
## OppTEN -99.5000000 59.41281 -1.674723061 9.988019e-02
## OppWAS -79.6666667 53.14043 -1.499172381 1.397623e-01Prediction interval using a linear model
## fit lwr upr
## 1 204 19.38727 388.6127
## 2 204 19.38727 388.6127
## 3 204 19.38727 388.6127
## 4 204 19.38727 388.6127
## 5 204 19.38727 388.6127
## 6 204 19.38727 388.6127Rush Yards vs. opponent data
CNfitRushYds<- lm(RushYds ~ Opp, data = CNdf)
summary(CNfitRushYds)$coefficients## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 28.6666667 13.64632 2.10068854 0.040441916
## OppATL 20.4242424 15.39512 1.32666952 0.190306056
## OppBAL -21.6666667 27.29264 -0.79386486 0.430814227
## OppBUF -13.6666667 27.29264 -0.50074552 0.618623480
## OppCHI -1.6666667 19.29881 -0.08636111 0.931504803
## OppCIN 78.3333333 27.29264 2.87012679 0.005882711
## OppCLE 34.3333333 27.29264 1.25797046 0.213916617
## OppDAL 25.8333333 21.57673 1.19727775 0.236525971
## OppDEN 1.8333333 21.57673 0.08496810 0.932606909
## OppDET -0.6666667 21.57673 -0.03089749 0.975467352
## OppGNB 21.6666667 19.29881 1.12269445 0.266627522
## OppHOU 36.8333333 21.57673 1.70708635 0.093658926
## OppIND 18.3333333 21.57673 0.84968099 0.399324437
## OppJAX 2.3333333 21.57673 0.10814122 0.914292038
## OppKAN 49.3333333 27.29264 1.80756921 0.076350104
## OppMIA 22.3333333 27.29264 0.81829147 0.416854747
## OppMIN 10.8333333 18.05238 0.60010542 0.550993345
## OppNOR 11.7878788 15.39512 0.76568908 0.447257940
## OppNWE 33.3333333 27.29264 1.22133055 0.227365995
## OppNYG 21.6666667 19.29881 1.12269445 0.266627522
## OppNYJ -16.6666667 27.29264 -0.61066527 0.544031201
## OppOAK 31.3333333 27.29264 1.14805071 0.256103198
## OppPHI -2.6666667 19.29881 -0.13817778 0.890623575
## OppPIT -21.6666667 27.29264 -0.79386486 0.430814227
## OppSDG -21.6666667 27.29264 -0.79386486 0.430814227
## OppSEA 4.8333333 18.05238 0.26773934 0.789939195
## OppSFO -2.6666667 21.57673 -0.12358996 0.902107454
## OppSTL -2.6666667 27.29264 -0.09770644 0.922533973
## OppTAM 14.0833333 16.00173 0.88011329 0.382771364
## OppTEN 10.3333333 21.57673 0.47891110 0.633971877
## OppWAS 8.6666667 19.29881 0.44907778 0.655206533Prediction interval using a linear model
## fit lwr upr
## 1 26 -41.04511 93.04511
## 2 26 -41.04511 93.04511
## 3 26 -41.04511 93.04511
## 4 26 -41.04511 93.04511
## 5 26 -41.04511 93.04511
## 6 26 -41.04511 93.04511Pass TDs vs. opponent data
CNfitPassTD<- lm(PassTD ~ Opp, data = CNdf)
summary(CNfitPassTD)$coefficients## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.6666667 0.7240525 0.9207436 0.36135643
## OppATL 0.7878788 0.8168413 0.9645432 0.33915419
## OppBAL 0.3333333 1.4481049 0.2301859 0.81883347
## OppBUF 1.3333333 1.4481049 0.9207436 0.36135643
## OppCHI 0.3333333 1.0239648 0.3255320 0.74606017
## OppCIN 1.3333333 1.4481049 0.9207436 0.36135643
## OppCLE 0.3333333 1.4481049 0.2301859 0.81883347
## OppDAL -0.1666667 1.1448274 -0.1455823 0.88480338
## OppDEN 0.8333333 1.1448274 0.7279117 0.46987138
## OppDET 0.3333333 1.1448274 0.2911647 0.77206183
## OppGNB 1.0000000 1.0239648 0.9765961 0.33320583
## OppHOU 1.3333333 1.1448274 1.1646588 0.24937274
## OppIND 0.3333333 1.1448274 0.2911647 0.77206183
## OppJAX 0.3333333 1.1448274 0.2911647 0.77206183
## OppKAN 2.3333333 1.4481049 1.6113013 0.11305415
## OppMIA 0.3333333 1.4481049 0.2301859 0.81883347
## OppMIN 1.0833333 0.9578314 1.1310272 0.26313568
## OppNOR 0.9696970 0.8168413 1.1871301 0.24046994
## OppNWE 2.3333333 1.4481049 1.6113013 0.11305415
## OppNYG 2.0000000 1.0239648 1.9531921 0.05608525
## OppNYJ 0.3333333 1.4481049 0.2301859 0.81883347
## OppOAK 0.3333333 1.4481049 0.2301859 0.81883347
## OppPHI 1.0000000 1.0239648 0.9765961 0.33320583
## OppPIT 0.3333333 1.4481049 0.2301859 0.81883347
## OppSDG 1.3333333 1.4481049 0.9207436 0.36135643
## OppSEA -0.1666667 0.9578314 -0.1740042 0.86252511
## OppSFO 1.3333333 1.1448274 1.1646588 0.24937274
## OppSTL 0.3333333 1.4481049 0.2301859 0.81883347
## OppTAM 1.0833333 0.8490268 1.2759708 0.20752953
## OppTEN -0.1666667 1.1448274 -0.1455823 0.88480338
## OppWAS 1.6666667 1.0239648 1.6276601 0.10952809Prediction interval using a linear model
## fit lwr upr
## 1 1 -2.557309 4.557309
## 2 1 -2.557309 4.557309
## 3 1 -2.557309 4.557309
## 4 1 -2.557309 4.557309
## 5 1 -2.557309 4.557309
## 6 1 -2.557309 4.557309Rush TDs vs. opponent data
CNfitRushTD<- lm(RushTD ~ Opp, data = CNdf)
summary(CNfitRushTD)$coefficients## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.33333333 0.4123706 0.8083343 0.4225117
## OppATL 0.21212121 0.4652168 0.4559620 0.6502801
## OppBAL -0.33333333 0.8247413 -0.4041671 0.6877162
## OppBUF -0.33333333 0.8247413 -0.4041671 0.6877162
## OppCHI 0.33333333 0.5831802 0.5715787 0.5700237
## OppCIN 0.66666667 0.8247413 0.8083343 0.4225117
## OppCLE 0.66666667 0.8247413 0.8083343 0.4225117
## OppDAL 0.16666667 0.6520152 0.2556177 0.7992355
## OppDEN 0.16666667 0.6520152 0.2556177 0.7992355
## OppDET 0.66666667 0.6520152 1.0224710 0.3112025
## OppGNB 0.33333333 0.5831802 0.5715787 0.5700237
## OppHOU 0.16666667 0.6520152 0.2556177 0.7992355
## OppIND 0.16666667 0.6520152 0.2556177 0.7992355
## OppJAX -0.33333333 0.6520152 -0.5112355 0.6113095
## OppKAN -0.33333333 0.8247413 -0.4041671 0.6877162
## OppMIA 0.66666667 0.8247413 0.8083343 0.4225117
## OppMIN 0.16666667 0.5455151 0.3055216 0.7611657
## OppNOR 0.21212121 0.4652168 0.4559620 0.6502801
## OppNWE -0.33333333 0.8247413 -0.4041671 0.6877162
## OppNYG 0.33333333 0.5831802 0.5715787 0.5700237
## OppNYJ -0.33333333 0.8247413 -0.4041671 0.6877162
## OppOAK 0.66666667 0.8247413 0.8083343 0.4225117
## OppPHI 0.66666667 0.5831802 1.1431573 0.2581108
## OppPIT -0.33333333 0.8247413 -0.4041671 0.6877162
## OppSDG -0.33333333 0.8247413 -0.4041671 0.6877162
## OppSEA -0.08333333 0.5455151 -0.1527608 0.8791670
## OppSFO -0.33333333 0.6520152 -0.5112355 0.6113095
## OppSTL -0.33333333 0.8247413 -0.4041671 0.6877162
## OppTAM 0.66666667 0.4835474 1.3786996 0.1737790
## OppTEN 0.16666667 0.6520152 0.2556177 0.7992355
## OppWAS 0.33333333 0.5831802 0.5715787 0.5700237Prediction interval using a linear model
## fit lwr upr
## 1 9.992007e-16 -2.025999 2.025999
## 2 9.992007e-16 -2.025999 2.025999
## 3 9.992007e-16 -2.025999 2.025999
## 4 9.992007e-16 -2.025999 2.025999
## 5 9.992007e-16 -2.025999 2.025999
## 6 9.992007e-16 -2.025999 2.025999Interceptions vs. opponent data
CNfitInt<- lm(Int ~ Opp, data = CNdf)
summary(CNfitInt)$coefficients## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.333333e+00 0.6136355 2.172843e+00 0.0342835
## OppATL -4.242424e-01 0.6922742 -6.128243e-01 0.5426133
## OppBAL -1.333333e+00 1.2272710 -1.086421e+00 0.2822092
## OppBUF -3.333333e-01 1.2272710 -2.716053e-01 0.7869807
## OppCHI -2.195188e-15 0.8678116 -2.529567e-15 1.0000000
## OppCIN -3.333333e-01 1.2272710 -2.716053e-01 0.7869807
## OppCLE -3.333333e-01 1.2272710 -2.716053e-01 0.7869807
## OppDAL -8.333333e-01 0.9702429 -8.588915e-01 0.3942684
## OppDEN 1.666667e-01 0.9702429 1.717783e-01 0.8642661
## OppDET 6.666667e-01 0.9702429 6.871132e-01 0.4950072
## OppGNB 3.333333e-01 0.8678116 3.841079e-01 0.7024352
## OppHOU -8.333333e-01 0.9702429 -8.588915e-01 0.3942684
## OppIND -8.333333e-01 0.9702429 -8.588915e-01 0.3942684
## OppJAX -8.333333e-01 0.9702429 -8.588915e-01 0.3942684
## OppKAN -1.333333e+00 1.2272710 -1.086421e+00 0.2822092
## OppMIA -3.333333e-01 1.2272710 -2.716053e-01 0.7869807
## OppMIN -3.333333e-01 0.8117634 -4.106287e-01 0.6830003
## OppNOR -6.969697e-01 0.6922742 -1.006783e+00 0.3186137
## OppNWE -1.333333e+00 1.2272710 -1.086421e+00 0.2822092
## OppNYG -9.243760e-17 0.8678116 -1.065180e-16 1.0000000
## OppNYJ -1.333333e+00 1.2272710 -1.086421e+00 0.2822092
## OppOAK -3.333333e-01 1.2272710 -2.716053e-01 0.7869807
## OppPHI 6.666667e-01 0.8678116 7.682159e-01 0.4457684
## OppPIT -1.333333e+00 1.2272710 -1.086421e+00 0.2822092
## OppSDG -1.333333e+00 1.2272710 -1.086421e+00 0.2822092
## OppSEA -5.833333e-01 0.8117634 -7.186002e-01 0.4755435
## OppSFO -3.333333e-01 0.9702429 -3.435566e-01 0.7325390
## OppSTL -1.333333e+00 1.2272710 -1.086421e+00 0.2822092
## OppTAM -8.333333e-01 0.7195514 -1.158129e+00 0.2520036
## OppTEN -8.333333e-01 0.9702429 -8.588915e-01 0.3942684
## OppWAS -1.333333e+00 0.8678116 -1.536432e+00 0.1303815Prediction interval using a linear model
## fit lwr upr
## 1 6.661338e-16 -3.014824 3.014824
## 2 6.661338e-16 -3.014824 3.014824
## 3 6.661338e-16 -3.014824 3.014824
## 4 6.661338e-16 -3.014824 3.014824
## 5 6.661338e-16 -3.014824 3.014824
## 6 6.661338e-16 -3.014824 3.014824Model fitting vs. defensive rank
Pass Yards vs. Defensive Rank
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 209.684496 14.6329907 14.329572 4.863700e-24
## Def.Rank.Yds 1.378596 0.6917944 1.992783 4.961161e-02Prediction interval using a linear model
## fit lwr upr
## 1 223.4705 95.84432 351.0966
## 2 223.4705 95.84432 351.0966
## 3 223.4705 95.84432 351.0966
## 4 223.4705 95.84432 351.0966
## 5 223.4705 95.84432 351.0966
## 6 223.4705 95.84432 351.0966Rush Yards vs. Defensive Rank
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 31.636691 5.461463 5.792714 1.236875e-07
## Def.Rank.Yds 0.484676 0.258198 1.877149 6.405391e-02Prediction interval using a linear model
## fit lwr upr
## 1 36.48345 -11.15038 84.11728
## 2 36.48345 -11.15038 84.11728
## 3 36.48345 -11.15038 84.11728
## 4 36.48345 -11.15038 84.11728
## 5 36.48345 -11.15038 84.11728
## 6 36.48345 -11.15038 84.11728Pass TDs vs. Defensive Rank
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.76030700 0.25208857 3.016031 0.003409254
## Def.Rank.Yds 0.03906339 0.01191783 3.277728 0.001535963Prediction interval using a linear model
## fit lwr upr
## 1 1.150941 -1.047727 3.349609
## 2 1.150941 -1.047727 3.349609
## 3 1.150941 -1.047727 3.349609
## 4 1.150941 -1.047727 3.349609
## 5 1.150941 -1.047727 3.349609
## 6 1.150941 -1.047727 3.349609Rush TDs vs. Defensive Rank
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.28453388 0.1464728 1.942571 0.05550046
## Def.Rank.Yds 0.01412087 0.0069247 2.039202 0.04465064Prediction interval using a linear model
## fit lwr upr
## 1 0.4257425 -0.8517653 1.70325
## 2 0.4257425 -0.8517653 1.70325
## 3 0.4257425 -0.8517653 1.70325
## 4 0.4257425 -0.8517653 1.70325
## 5 0.4257425 -0.8517653 1.70325
## 6 0.4257425 -0.8517653 1.70325Interceptions vs. Defensive Rank
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.782843304 0.22988339 3.4053930 0.001025241
## Def.Rank.Yds 0.002710008 0.01086805 0.2493556 0.803709321Prediction interval using a linear model
## fit lwr upr
## 1 0.8099434 -1.195055 2.814942
## 2 0.8099434 -1.195055 2.814942
## 3 0.8099434 -1.195055 2.814942
## 4 0.8099434 -1.195055 2.814942
## 5 0.8099434 -1.195055 2.814942
## 6 0.8099434 -1.195055 2.814942Newton Pass Yards Plot of Data Yards against opponents
## Warning in qt((1 - level)/2, df): NaNs produced
## Warning in qt((1 - level)/2, df): NaNs produced
## Warning in qt((1 - level)/2, df): NaNs produced
## Warning in qt((1 - level)/2, df): NaNs produced
## Warning in qt((1 - level)/2, df): NaNs produced
## Warning in qt((1 - level)/2, df): NaNs produced
## Warning in qt((1 - level)/2, df): NaNs produced
## Warning in qt((1 - level)/2, df): NaNs produced
plot of chunk unnamed-chunk-94
Newton Rush Yards Plot of Data Yards against opponents
## Warning in qt((1 - level)/2, df): NaNs produced
## Warning in qt((1 - level)/2, df): NaNs produced
## Warning in qt((1 - level)/2, df): NaNs produced
## Warning in qt((1 - level)/2, df): NaNs produced
## Warning in qt((1 - level)/2, df): NaNs produced
## Warning in qt((1 - level)/2, df): NaNs produced
## Warning in qt((1 - level)/2, df): NaNs produced
## Warning in qt((1 - level)/2, df): NaNs produced
plot of chunk unnamed-chunk-95
Newton TDs Plot of Data Yards against opponents
## Warning in qt((1 - level)/2, df): NaNs produced
## Warning in qt((1 - level)/2, df): NaNs produced
## Warning in qt((1 - level)/2, df): NaNs produced
## Warning in qt((1 - level)/2, df): NaNs produced
## Warning in qt((1 - level)/2, df): NaNs produced
plot of chunk unnamed-chunk-96
Machine Learning Alogorithm (Pass Yards)
Data Splitting
In this data splitting action, I am splitting the data into 60% training set and 40% testing set.
## Location Opp Def.Rank.Yds PassYds PassTD
## Away:42 ATL :11 Min. : 1.00 Min. :114.0 Min. :0.000
## Home:42 NOR :11 1st Qu.:10.00 1st Qu.:194.0 1st Qu.:1.000
## TAM : 8 Median :18.00 Median :231.5 Median :1.000
## MIN : 4 Mean :18.63 Mean :235.4 Mean :1.488
## SEA : 4 3rd Qu.:28.00 3rd Qu.:277.0 3rd Qu.:2.000
## ARI : 3 Max. :32.00 Max. :432.0 Max. :5.000
## (Other):43
## Int RushYds RushTD
## Min. :0.0000 Min. : 4.00 Min. :0.0000
## 1st Qu.:0.0000 1st Qu.: 23.75 1st Qu.:0.0000
## Median :1.0000 Median : 39.00 Median :0.0000
## Mean :0.8333 Mean : 40.67 Mean :0.5476
## 3rd Qu.:1.0000 3rd Qu.: 53.25 3rd Qu.:1.0000
## Max. :4.0000 Max. :116.00 Max. :3.0000
## ## [1] 52 8## [1] 32 8K-Fold Cross Validation
We will then use the K-Fold process to cross-validate the data by splitting the training set in to many, smaller data sets.
Here, I am creating 10 folds and setting a random number seed of 32323 for the study. Each fold has approximately the same number of samples in it.
set.seed(32323)
Passfolds <- createFolds(y=CNdf$PassYds,k=10,list=TRUE,returnTrain=TRUE)
sapply(Passfolds,length)## Fold01 Fold02 Fold03 Fold04 Fold05 Fold06 Fold07 Fold08 Fold09 Fold10
## 74 75 76 76 76 76 75 76 76 76Passfolds[[1]][1:10]## [1] 1 2 3 4 5 7 8 9 10 11Machine learning alogorithm decisioning
First I wanted to determine the optimal machine learning model to use. The first test I used the Decision Tree approach. I followed up by testing the Random Forest approach.
Machine learning using Decision Trees
The first task was to determine model fit.
Next, construct a Decision Tree graph
fancyRpartPlot(modelFitPass)
plot of chunk unnamed-chunk-100
Machine Learning Alogorithm (Rush Yards)
Data Splitting
In this data splitting action, I am splitting the data into 60% training set and 40% testing set.
## Location Opp Def.Rank.Yds PassYds PassTD
## Away:42 ATL :11 Min. : 1.00 Min. :114.0 Min. :0.000
## Home:42 NOR :11 1st Qu.:10.00 1st Qu.:194.0 1st Qu.:1.000
## TAM : 8 Median :18.00 Median :231.5 Median :1.000
## MIN : 4 Mean :18.63 Mean :235.4 Mean :1.488
## SEA : 4 3rd Qu.:28.00 3rd Qu.:277.0 3rd Qu.:2.000
## ARI : 3 Max. :32.00 Max. :432.0 Max. :5.000
## (Other):43
## Int RushYds RushTD
## Min. :0.0000 Min. : 4.00 Min. :0.0000
## 1st Qu.:0.0000 1st Qu.: 23.75 1st Qu.:0.0000
## Median :1.0000 Median : 39.00 Median :0.0000
## Mean :0.8333 Mean : 40.67 Mean :0.5476
## 3rd Qu.:1.0000 3rd Qu.: 53.25 3rd Qu.:1.0000
## Max. :4.0000 Max. :116.00 Max. :3.0000
## ## [1] 52 8## [1] 32 8We will then use the K-Fold process to cross-validate the data by splitting the training set in to many, smaller data sets.
Here, I am creating 10 folds and setting a random number seed of 32323 for the study. Each fold has approximately the same number of samples in it.
set.seed(32323)
Passfolds <- createFolds(y=CNdf$RushYds,k=10,list=TRUE,returnTrain=TRUE)
sapply(Passfolds,length)## Fold01 Fold02 Fold03 Fold04 Fold05 Fold06 Fold07 Fold08 Fold09 Fold10
## 74 75 76 75 76 76 76 76 76 76Passfolds[[1]][1:10]## [1] 1 2 3 5 6 8 9 10 11 12Decision tree model fit Rush Yards Defense Rank.
Pass Yards decision tree graph
fancyRpartPlot(modelFitRush)
plot of chunk unnamed-chunk-104
Machine learning using Decision Trees
The first task was to determine model fit.
Output from Decision Tree Model vs. Yards Passing
Current week (9) opponent L.A. Rams (formerly St. Louis Rams). The defense is currently ranked # 10 according to www.pro-football-reference.com.
According to the model, there is a 60% chance that Cam Newton will pass for 240 yards. This takes in to account defense rank ONLY. There is an 18.0% chance Cam will rush for 25 yards. Once again this takes in to account defense rank ONLY.
We will now use the LM model for prediction (Pass Yds)
## 1 2 3 4 5 6 7 8
## 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174
## 9 10 11 12 13 14 15 16
## 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174
## 17 18 19 20 21 22 23 24
## 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174
## 25 26 27 28 29 30 31 32
## 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174
## 33 34 35 36 37 38 39 40
## 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174
## 41 42 43 44 45 46 47 48
## 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174
## 49 50 51 52 53 54 55 56
## 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174
## 57 58 59 60 61 62 63 64
## 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174
## 65 66 67 68 69 70 71 72
## 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174
## 73 74 75 76 77 78 79 80
## 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174 195.5174
## 81 82 83 84
## 195.5174 195.5174 195.5174 195.5174Predicted passing points
## 1 2 3 4 5 6 7 8
## 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694
## 9 10 11 12 13 14 15 16
## 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694
## 17 18 19 20 21 22 23 24
## 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694
## 25 26 27 28 29 30 31 32
## 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694
## 33 34 35 36 37 38 39 40
## 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694
## 41 42 43 44 45 46 47 48
## 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694
## 49 50 51 52 53 54 55 56
## 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694
## 57 58 59 60 61 62 63 64
## 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694
## 65 66 67 68 69 70 71 72
## 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694
## 73 74 75 76 77 78 79 80
## 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694 7.820694
## 81 82 83 84
## 7.820694 7.820694 7.820694 7.820694We will now use the LM model for prediction (Rush Yds)
## 1 2 3 4 5 6 7 8
## 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227
## 9 10 11 12 13 14 15 16
## 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227
## 17 18 19 20 21 22 23 24
## 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227
## 25 26 27 28 29 30 31 32
## 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227
## 33 34 35 36 37 38 39 40
## 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227
## 41 42 43 44 45 46 47 48
## 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227
## 49 50 51 52 53 54 55 56
## 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227
## 57 58 59 60 61 62 63 64
## 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227
## 65 66 67 68 69 70 71 72
## 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227
## 73 74 75 76 77 78 79 80
## 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227 24.73227
## 81 82 83 84
## 24.73227 24.73227 24.73227 24.73227Predicted rushing points
## 1 2 3 4 5 6 7 8
## 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227
## 9 10 11 12 13 14 15 16
## 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227
## 17 18 19 20 21 22 23 24
## 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227
## 25 26 27 28 29 30 31 32
## 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227
## 33 34 35 36 37 38 39 40
## 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227
## 41 42 43 44 45 46 47 48
## 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227
## 49 50 51 52 53 54 55 56
## 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227
## 57 58 59 60 61 62 63 64
## 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227
## 65 66 67 68 69 70 71 72
## 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227
## 73 74 75 76 77 78 79 80
## 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227 2.473227
## 81 82 83 84
## 2.473227 2.473227 2.473227 2.473227We will now use the LM model for prediction (Pass TDs)
## 1 2 3 4 5 6 7
## 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535
## 8 9 10 11 12 13 14
## 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535
## 15 16 17 18 19 20 21
## 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535
## 22 23 24 25 26 27 28
## 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535
## 29 30 31 32 33 34 35
## 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535
## 36 37 38 39 40 41 42
## 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535
## 43 44 45 46 47 48 49
## 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535
## 50 51 52 53 54 55 56
## 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535
## 57 58 59 60 61 62 63
## 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535
## 64 65 66 67 68 69 70
## 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535
## 71 72 73 74 75 76 77
## 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535
## 78 79 80 81 82 83 84
## 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535 0.8045535Predicted rushing points
## 1 2 3 4 5 6 7 8
## 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214
## 9 10 11 12 13 14 15 16
## 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214
## 17 18 19 20 21 22 23 24
## 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214
## 25 26 27 28 29 30 31 32
## 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214
## 33 34 35 36 37 38 39 40
## 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214
## 41 42 43 44 45 46 47 48
## 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214
## 49 50 51 52 53 54 55 56
## 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214
## 57 58 59 60 61 62 63 64
## 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214
## 65 66 67 68 69 70 71 72
## 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214
## 73 74 75 76 77 78 79 80
## 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214 3.218214
## 81 82 83 84
## 3.218214 3.218214 3.218214 3.218214We will now use the LM model for prediction (Rush TDs)
## 1 2 3 4 5 6
## -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085
## 7 8 9 10 11 12
## -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085
## 13 14 15 16 17 18
## -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085
## 19 20 21 22 23 24
## -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085
## 25 26 27 28 29 30
## -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085
## 31 32 33 34 35 36
## -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085
## 37 38 39 40 41 42
## -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085
## 43 44 45 46 47 48
## -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085
## 49 50 51 52 53 54
## -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085
## 55 56 57 58 59 60
## -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085
## 61 62 63 64 65 66
## -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085
## 67 68 69 70 71 72
## -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085
## 73 74 75 76 77 78
## -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085
## 79 80 81 82 83 84
## -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085 -0.05003085Predicted rushing TD points
## 1 2 3 4 5 6
## -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851
## 7 8 9 10 11 12
## -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851
## 13 14 15 16 17 18
## -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851
## 19 20 21 22 23 24
## -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851
## 25 26 27 28 29 30
## -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851
## 31 32 33 34 35 36
## -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851
## 37 38 39 40 41 42
## -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851
## 43 44 45 46 47 48
## -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851
## 49 50 51 52 53 54
## -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851
## 55 56 57 58 59 60
## -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851
## 61 62 63 64 65 66
## -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851
## 67 68 69 70 71 72
## -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851
## 73 74 75 76 77 78
## -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851
## 79 80 81 82 83 84
## -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851 -0.3001851Total predicted points using linear regression models
## 1 2 3 4 5 6 7 8
## 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195
## 9 10 11 12 13 14 15 16
## 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195
## 17 18 19 20 21 22 23 24
## 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195
## 25 26 27 28 29 30 31 32
## 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195
## 33 34 35 36 37 38 39 40
## 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195
## 41 42 43 44 45 46 47 48
## 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195
## 49 50 51 52 53 54 55 56
## 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195
## 57 58 59 60 61 62 63 64
## 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195
## 65 66 67 68 69 70 71 72
## 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195
## 73 74 75 76 77 78 79 80
## 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195 13.21195
## 81 82 83 84
## 13.21195 13.21195 13.21195 13.21195Outcome: Cam Newton’s predicted point total for Week 9 against the LA Rams is 13.21 points. NFL.com is predicting that Cam’s point total will be 15.44 points
##
##
## processing file: QB_ML_Experiment_2016.Rmd## Error in parse_block(g[-1], g[1], params.src): duplicate label 'Points'