Awwal Mohammed
International University of Malaya-Wales © 2020-2022
This report covers basic approaches to using easily-understood machine learning techniques to predict "overall player ratings"1 based on data obtained from the FIFA video games, with initial inference from real life football data.
KNN Regression, Multiple Linear Regression, Voting Regressor, and Gradient Boosting Regressor algorithmic models are used in this research. Evaluation methods such as Lasso, Ridge, Cross-Validation, and Multiple KNN value comparison are explored.
FIFA ratings refer to the aggregate, weighted metric which serves as a player’s overall beneficiary. It lies within the range of 0-99 with the provision that the higher the number, the better. For example, Cristiano Ronaldo may be rated 93 while Lionel Messi may be rated 94.
This report aims to tackle the following:
- Predicting FIFA overall ratings of outfield players2 from FIFA 20, and crosschecking it against existing results.
- Identifying the feature(s) that influence(s) the ratings the most.
This is a supervised learning exercise. The report culminates with a summary of the performance of each algorithm in predicting the overall ratings.
While this research focuses on FIFA 20, there's an additional dataset introduced to cover grounds for investigative knowledge bias. One dataset has information pertaining to real life European football players statistics for inference (you can completely ignore this, and subsequent related activities if you're not curious about it), while the other contains data to be used for training and learning.
-
European Soccer Dataset (Optional Inferential Dataset)
- This dataset has 25,000 match records, 10,000 players, 11 European countries with their leagues team line
- with squad formation, detailed match events ( goal types, possession, corner, cross, fouls, cards etc…) for +10,000 matches.
-
FIFA 15-20 Complete Player Dataset (Core Dataset)
- The FIFA 20 subset is used as the core dataset. The datasets include the players' data from FIFA 15 to FIFA 20, all the expected features necessary for the machine learning excercise.
This conversion is due to programming convenience and familiarity.
⚠ Prerequisites for this section:
-
Obtain the European Soccer Dataset from Kaggle here.
-
You may save it in the path
{current_path}/Datasets/european-soccer/. -
The resulting dataset should be an SQLite file called
database.sqlite.
PreProcessSQLite.py performs the neccesary conversion.
It is imperative to separate outfield players from goalkeepers due to the latter's unique features, for this reason, outfield players will remain the target for this exercise.
Prerequisites for this section:
-
Obtain the FIFA 15-20 Complete Player Dataset from Kaggle here.
-
You may save it in the path
{current_path}/Datasets/fifa-20-complete-player-dataset/. -
The resulting dataset should be a CSV file called
players_20.csv.
To achieve this grouping, we convert the CSV to an SQLite file to quickyl run some queries over it, then convert the grouped files back to CSV format.
Once this is done, you can optionally run the resulting datasets briefly through Microsoft Excel's Intelligent Suggestions module for quick insights into the story the data might be trying to tell.
This is a visual representation of the procedure as described,
which PlayerGroup.py subsequently implements:
Not quite there yet. Some unwanted features still need to be removed. To achieve this, you'll have to combine the resulting grouped CSV files as one, then convert into SQLite format (Rebasedata and Konbert can do this for you) to easily run some SQL queries over them for further trimming.
This is a visual representation of the procedure as described,
which TrimFeatures.py implements:
This is a visual representation of the KNN Regression model,
which KNNReg.py implements,
with the subsequent output:
Automatically created module for IPython interactive environment
sofifa_id 0
age 0
overall 0
value_eur 0
wage_eur 0
..
team_position_RS 0
team_position_RW 0
team_position_RWB 0
team_position_ST 0
team_position_SUB 0
Length: 1036, dtype: int64
RMSE value for k= 1 is: 1.5700318468107581
RMSE value for k= 2 is: 1.4832396974191326
RMSE value for k= 3 is: 1.4001983986405482
RMSE value for k= 4 is: 1.3722700171613458
RMSE value for k= 5 is: 1.3862900129482285
RMSE value for k= 6 is: 1.3676825492619096
RMSE value for k= 7 is: 1.3646783279496708
RMSE value for k= 8 is: 1.3583940426106116
RMSE value for k= 9 is: 1.3515880006336216
RMSE value for k= 10 is: 1.3473863588444108
RMSE value for k= 11 is: 1.3475260466057761
RMSE value for k= 12 is: 1.357016232605769
RMSE value for k= 13 is: 1.3537041009391264
RMSE value for k= 14 is: 1.3498299212591676
RMSE value for k= 15 is: 1.3435773144854743
RMSE value for k= 16 is: 1.3486828064819392
RMSE value for k= 17 is: 1.356784821518547
RMSE value for k= 18 is: 1.360447684122609
RMSE value for k= 19 is: 1.3614244210728814
RMSE value for k= 20 is: 1.3570740584065408
predictions:
[84.66666667 78.22222222 79.55555556 80.88888889 78.55555556 81.22222222
78.44444444 79.22222222 83.66666667 81.55555556 79.11111111 78.55555556
83.77777778 84. 78.88888889 78.88888889 78.33333333 78.55555556
78.22222222 79.66666667 78.55555556 82.55555556 78.55555556 85.77777778
81.88888889 79.55555556 79.66666667 78.77777778 80. 78.33333333
79.11111111 80.66666667 78.33333333 79.66666667 78.55555556 78.55555556
79.88888889 83.88888889 78.33333333 80.11111111 78.44444444 86.22222222
84.88888889 78.77777778 80.55555556 85.22222222 78.11111111 83.
81.55555556 80.55555556 80.11111111 79.88888889 78. 83.55555556
83.44444444 79.77777778 78.66666667 83.33333333 82.66666667 81.33333333
78. 78. 77.55555556 79.44444444 77.55555556 78.44444444
81.88888889 79.88888889 77.88888889 77.77777778 78.44444444 78.77777778
83.44444444 79.11111111 78.11111111 82.66666667 79.77777778 80.
81.44444444 83.66666667 81.66666667 78.55555556 78. 78.88888889
78.22222222 77.44444444 79. 83.44444444 82.44444444 78.55555556
81.77777778 85.55555556 78.33333333 78.88888889 83. 77.77777778
80. 90. 78.55555556 78.22222222 79.44444444 79.44444444
89.66666667 78.22222222 78.11111111 79. 78.66666667 78.88888889
83.11111111 79.88888889 78.55555556 77.77777778 77.77777778 79.22222222
81.66666667 82.66666667 83.44444444 78.33333333 79.11111111 78.33333333
79.55555556 88.33333333 78.33333333 79.66666667 79.22222222 79.77777778
82.88888889 77.77777778 79.11111111 78.11111111 78.88888889 79.55555556
83.22222222 84.33333333 79.11111111 78.88888889 78.33333333 78.88888889
78.55555556 78.11111111 77.77777778 82.44444444 79. 78.
79.11111111 78.88888889 79.11111111 84. 78.88888889 78.22222222
86.44444444 77.66666667 82.55555556 78.88888889 80.22222222 90.
87.66666667 77.77777778 81.77777778 78. 78.33333333 78.22222222
78.66666667 78.44444444 83.44444444 78.55555556 86.88888889 86.55555556
78.22222222 78.11111111 77.77777778 79.88888889 79.66666667 78.77777778
85.22222222 78.77777778 78. 79.11111111 77.88888889 79.11111111
77.77777778 81. 78.33333333 82.66666667 81.77777778 78.
80. 78.33333333 78.33333333 82.22222222 79.22222222 79.
78. 78.11111111 78.55555556 83.33333333 78.44444444 78.55555556
80.77777778 78.77777778]
93 85
613 79
494 80
376 80
999 77
..
89 85
256 82
341 81
509 79
464 80
Name: overall, Length: 200, dtype: int64
score: 0.7863803120519212
This is implemented in KNNRegTest.py.
Automatically created module for IPython interactive environment
sofifa_id 0
age 0
overall 0
value_eur 0
wage_eur 0
..
team_position_RS 0
team_position_RW 0
team_position_RWB 0
team_position_ST 0
team_position_SUB 0
Length: 1036, dtype: int64
RMSE value for k= 1 is: 1.5700318468107581
RMSE value for k= 2 is: 1.4718186029535025
RMSE value for k= 3 is: 1.4306175822583291
RMSE value for k= 4 is: 1.4004463574160917
RMSE value for k= 5 is: 1.398427688513067
RMSE value for k= 6 is: 1.3884943884174201
RMSE value for k= 7 is: 1.3890240297064842
RMSE value for k= 8 is: 1.3700764121026243
RMSE value for k= 9 is: 1.3700950674413659
RMSE value for k= 10 is: 1.3699999999999994
RMSE value for k= 11 is: 1.3755840232274716
RMSE value for k= 12 is: 1.3712828543618074
RMSE value for k= 13 is: 1.3695872483230962
RMSE value for k= 14 is: 1.3795240971776737
RMSE value for k= 15 is: 1.3788320339250097
RMSE value for k= 16 is: 1.376263623910768
RMSE value for k= 17 is: 1.3793271013777533
RMSE value for k= 18 is: 1.3878051327256837
RMSE value for k= 19 is: 1.384630714691279
RMSE value for k= 20 is: 1.3872499774734186
predictions:
[80.33333333 83. 79.33333333 81.66666667 78.33333333 79.
80.33333333 79.66666667 78. 78. 79.66666667 78.
77.33333333 86.66666667 83. 81.33333333 78. 83.
82.33333333 78. 78. 79.66666667 82.33333333 81.
77. 82. 78. 87.33333333 78.66666667 79.33333333
81.66666667 80.66666667 83. 78.33333333 87.66666667 81.33333333
79.33333333 79. 83.66666667 83.66666667 84.66666667 80.
78.33333333 77.66666667 77.66666667 80. 83. 79.
80. 81.66666667 83.33333333 79.66666667 81.33333333 79.66666667
77.66666667 79.33333333 78. 88.33333333 85. 78.66666667
77.66666667 82. 79.33333333 77.33333333 77. 78.
85.66666667 78.33333333 78. 83.66666667 83. 78.
83. 84.33333333 79. 81.33333333 78.66666667 79.
83. 81. 78.33333333 85.66666667 77.66666667 79.
82. 83.66666667 78.66666667 79.66666667 79.33333333 79.
79.33333333 79.66666667 79. 83. 80. 77.
77.66666667 77.66666667 79. 78.33333333 91.33333333 81.33333333
85.66666667 78.33333333 83. 82.66666667 77.66666667 77.33333333
79. 77.33333333 77.66666667 84.33333333 78.33333333 78.33333333
78. 78.33333333 77.66666667 80. 85.33333333 85.33333333
79. 78. 78. 79.66666667 77.33333333 83.66666667
78. 78.66666667 82. 77.66666667 78.33333333 80.33333333
80.33333333 83.66666667 83. 79. 83. 90.
77.66666667 80.33333333 80.33333333 78.33333333 78.33333333 80.33333333
82. 81.66666667 78.66666667 83. 88.33333333 80.33333333
78. 78.33333333 77.66666667 83. 79.66666667 79.66666667
81.33333333 82.33333333 79.66666667 79.33333333 79. 78.33333333
78. 78.66666667 79.33333333 79.66666667 83.66666667 78.66666667
78.66666667 80.33333333 78.33333333 78. 83. 78.
78. 79. 81.33333333 83.33333333 78. 78.33333333
77. 86. 80.33333333 78.66666667 79.33333333 79.
77.66666667 78.66666667 91.33333333 80. 78.33333333 78.
89. 78.66666667 85.66666667 79. 79.66666667 78.66666667
83.33333333 83.66666667]
271 82
57 86
250 82
268 82
498 80
..
503 80
348 81
700 78
90 85
187 83
Name: overall, Length: 200, dtype: int64
score: 0.7843321144936547
This is a visual representation of the MLR model,
which MLReg.py subsequently implements:
Automatically created module for IPython interactive environment
sofifa_id 0
age 0
overall 0
value_eur 0
wage_eur 0
..
team_position_RS 0
team_position_RW 0
team_position_RWB 0
team_position_ST 0
team_position_SUB 0
Length: 1036, dtype: int64
predicted values:
[86.8747943 79.98287891 79.26993736 79.18332689 79.67756544 80.60501831
80.52276787 79.29888782 80.28082223 78.12278117 79.6012055 80.21880171
79.34236509 80.01603953 81.55328082 80.1751625 78.29607511 79.04708373
78.99171736 78.43517057 82.28184572 78.90349655 80.00990784 78.13109112
78.40954873 77.88751127 79.03719414 78.30889441 80.79354622 80.75094826
79.95673579 78.42026576 79.43704207 80.19447842 81.57691416 78.37510851
80.03805467 81.60567414 78.34112407 80.06318956 79.0295774 77.52133698
79.10315286 78.45721869 76.70407925 79.20363191 79.0899063 81.02318519
78.73865861 78.88245809 78.62692699 78.04434026 83.67718072 80.00887928
78.39601612 78.20699869 84.03526274 89.10538422 79.61438461 79.76556191
77.50886018 80.86960944 79.65229288 78.7759495 80.78806178 80.18113942
80.37753506 84.05407105 87.22970394 78.35125693 80.29624459 78.06141034
78.25784608 82.93970308 79.27280813 78.44266091 87.38093126 77.67930806
79.55207892 77.72399259 80.53090342 80.55093246 81.16288182 82.16377855
96.91288402 78.86621444 78.48639451 83.53253542 79.33765381 80.54401786
81.87619389 79.28763294 80.33273613 79.11719214 91.58374862 77.76259762
83.16961109 85.57175401 77.08213524 79.33741596 95.0694012 80.47219572
79.77821668 77.54980461 80.17510672 80.25218474 88.58374993 79.5192633
79.7607276 79.89006017 79.68248429 78.36850084 77.26642242 80.00992079
78.84233939 79.8116052 81.50129592 81.13023537 82.81474215 79.00844263
79.71370357 80.39037457 79.68971909 83.05315921 80.89104015 78.58593234
79.18882001 89.96442373 80.08217066 78.80075642 78.45946475 77.13835189
77.90239693 81.93584849 83.38431464 78.72713548 78.02919278 79.48799374
82.49463363 78.45865454 77.70564695 79.11573473 78.9234846 77.23354311
81.84524116 83.95840705 79.48622504 83.31858105 81.34200108 81.29536291
79.22843531 78.87804549 81.45809371 78.90036286 80.12059674 78.86754744
78.00698225 81.97069033 80.92848317 80.61446849 83.59708188 79.38641171
82.43287793 79.28820253 81.95752266 84.2907566 82.46222136 80.19084413
80.19315988 80.23974539 78.06477574 79.06676931 78.60883052 82.18218672
82.7180019 78.5318587 92.42876462 78.4452741 78.55158694 84.9763761
81.20648969 79.75411203 81.79719949 79.63294568 80.27402199 81.98106764
82.10545967 78.91899684 78.09562845 81.52371728 82.49728332 80.70055312
77.87706404 79.85673253 79.32684502 81.76385193 81.70596012 83.072761
78.05336621 77.36345795]
This is implemented in MLRegEval.py.
Automatically created module for IPython interactive environment
sofifa_id 0
age 0
overall 0
value_eur 0
wage_eur 0
..
team_position_RS 0
team_position_RW 0
team_position_RWB 0
team_position_ST 0
team_position_SUB 0
Length: 1036, dtype: int64
predicted values:
[79.40361188 83.55486616 81.26935389 80.42627552 78.96459486 77.67666383
78.73827333 79.41593791 77.77116033 78.16103164 97.4660832 77.23137748
86.81600328 76.81955798 80.05663759 78.96210958 80.27110505 81.04851716
78.68319823 78.45633012 81.80349478 81.4980994 79.13071163 79.89028273
77.32858504 80.00879513 82.87576176 80.12204373 77.2806339 79.79710615
81.62573841 78.56275373 78.32515851 83.12310018 79.84451354 78.1694399
79.21211357 78.36624133 83.71347022 80.08086447 84.79650572 81.79482146
77.37925677 79.60018492 80.15336693 78.48446607 79.1487607 78.38785964
79.98373812 78.34131857 80.94055847 85.41116137 79.8378149 80.91723177
79.93314862 78.06396441 90.66206746 78.37770273 82.33729768 79.17513625
77.35149402 79.15112245 78.84941524 78.6350503 78.02655455 79.88049548
82.26505602 78.3163484 79.48591601 81.6581002 80.13699552 82.86257194
80.40764975 80.03022497 78.87806953 79.99323897 81.18445367 83.14113558
78.17952985 79.72807329 79.12030415 78.94076627 83.65470293 78.28088579
79.38551553 78.48161506 81.77626225 78.78729366 79.60288408 77.39360559
84.93322437 78.28501946 79.04618823 85.59028857 77.2222882 79.56884039
80.9722789 83.93047723 79.43192999 79.61063578 79.22010825 79.53893457
82.56982223 81.32809656 79.98885033 77.04719699 82.94436761 78.76803857
84.45925102 78.11985891 81.35158585 80.98575625 79.64213251 79.94322338
78.60185508 82.84114098 80.51287329 82.87912627 88.16969988 78.8077043
81.28053633 84.46183976 80.06993357 82.16452983 81.33834315 78.01221207
78.89213423 78.48444117 78.74473552 93.4422006 88.61192856 79.00350348
81.04168014 78.23641009 81.72100382 78.82834833 78.13222522 81.94940254
78.98629566 79.39420975 83.15449298 77.67136974 78.92474634 78.83530033
81.72467206 78.8251622 82.55032577 79.37621408 78.18910526 90.52235784
78.1714016 77.31157394 79.06599898 80.1302792 81.64796724 77.31921165
78.11445321 79.53637569 80.24809392 79.65016543 92.85863224 79.44325882
79.53725036 79.34232542 78.88990941 79.55587099 80.82808883 79.16443232
84.88677667 77.68277603 77.39837335 80.11404591 82.28394053 78.33019422
79.56867135 90.87860951 77.71333032 81.95247804 77.28671842 79.67124147
77.80390413 79.23741156 80.60859116 79.72157188 80.6709539 78.99153519
77.226052 78.5025757 83.14298659 80.5327565 78.88535612 77.60836817
77.83411542 79.54316201 92.82923417 80.15032254 81.70800993 83.02083051
79.52211637 83.29302473]
score: 0.8616732626399296
cross validation measured values
[79.08913886 83.82119152 80.99591729 80.2216745 78.58090688 77.72066444
77.43627201 79.26671555 77.93841185 78.30394704 96.56826366 78.27951058
86.75621817 77.90873441 80.15142402 79.17811601 80.84911502 81.20145916
78.80289615 78.48519971 81.59596556 81.08069728 78.60519142 79.73183797
77.76516456 79.48521157 81.62868704 80.10090817 77.89670843 79.45573956
81.31739567 79.28155965 78.2905625 82.62681717 80.60956486 77.13567079
79.33043564 78.04629077 83.05276451 80.03226861 84.76425774 80.55965921
77.38933779 79.25528963 80.77631364 79.23807994 79.94047221 78.46094429
80.30061031 78.68442914 81.05741503 84.02140127 80.28283772 80.19884035
79.1806556 77.84041901 89.87575603 78.89926629 81.50797431 80.06772229
78.75858223 79.1366315 79.01726001 79.3170256 78.41535055 80.11663746
81.91777393 78.89408009 79.61059849 81.33870787 80.15672202 82.25621448
80.33483966 80.68801592 78.91526387 79.99729749 81.19460664 82.42893661
78.08684368 80.29256109 79.22629444 78.87737393 82.81045078 78.90028803
79.41719548 78.87213142 81.14030483 78.73717551 79.73727115 77.67397196
85.01266236 78.44844494 78.71843055 83.9630373 77.39062557 79.72893076
80.02139323 83.31152404 79.34241493 79.70393392 79.16319424 79.74044515
81.05737023 80.8494603 79.78674586 77.76842696 81.9909123 78.98126758
83.41068787 78.44373687 80.74778979 80.92530264 79.21160364 79.68123803
78.90037982 81.97230435 80.16161501 81.92799225 85.67780237 78.79622781
81.22473051 84.19925366 80.34619349 81.49552258 81.40796505 78.29971881
79.19321256 78.39927242 78.8396382 91.41057265 87.44350615 79.44693594
80.43540692 78.57190102 81.07935135 79.34488978 79.58831809 81.18767796
79.68104556 79.3474458 82.63157986 79.15713248 79.44889476 79.66153342
81.24538527 79.76829081 81.82946497 79.90661735 78.84097759 89.37033396
79.19115155 77.72291951 79.29578174 80.45160543 80.92730325 77.90455148
78.55791769 80.30838085 80.44002786 80.07305364 92.71038807 78.57125214
80.16246932 79.53306883 78.14423982 79.11586393 81.23920678 79.50821902
83.00116375 77.62975279 77.13185349 80.21600608 80.71167264 78.69294978
80.20065501 90.65482664 77.21841349 77.23611875 77.50055592 79.06161376
78.09282908 79.17411 79.53656241 79.12986282 80.36987396 77.88277873
77.64657577 77.44234746 82.30924642 80.98135788 78.35044369 77.53676448
78.69834324 81.92276809 90.98634074 79.38152221 80.5501763 82.21377907
79.164157 83.07097474]
cross validation scores at every 10th interval:
[0.83540359 0.80428478 0.86765576 0.64867744 0.91046375 0.72172856
0.89447911 0.81609669 0.68839869 0.77260819]
This is a visual representation of the Voting Regression model,
which VotingReg.py implements:
Automatically created module for IPython interactive environment
sofifa_id 0
age 0
overall 0
value_eur 0
wage_eur 0
..
team_position_RS 0
team_position_RW 0
team_position_RWB 0
team_position_ST 0
team_position_SUB 0
Length: 1036, dtype: int64
RandomForestRegressor Predictions:
[77.1 78.4 79.6 83.4 83.2 78.1 84.5 78.2 77.6 77.1 78.1 78. 79.1 78.
80.7 79.9 77.8 80.2 83.6 80.6 78.1 78.2 79.6 79.6 82.5 88.6 77.2 79.3
78.1 81.8 79.3 88.8 86.1 79. 81.2 78.2 89.3 78. 83.8 79.6 79.9 78.2
78. 78.1 79.4 82.3 77.1 89.3 80. 79.2 77. 83. 77.2 77.2 79.2 79.8
82.2 78.9 82.5 83.7 81.6 79.3 78.1 78.1 78. 80.1 79.7 77.1 79.7 82.9
78.4 83.7 77.3 78.4 79.4 77.7 77. 77.2 80. 81.4 78.7 80.6 79.3 79.3
79. 80. 84.1 79.8 79.5 85.2 78. 79.6 84.7 81.1 86.1 82.4 78.7 78.
79.2 79.5 79.9 81.3 82.5 78.5 78.1 77.1 79.1 82.1 78.6 79.4 78.4 78.9
81.6 86.3 80.5 83.6 82. 87.4 78.5 77.8 81.6 77.1 83.8 78.1 80.6 78.1
79.4 78.1 80.1 87.7 80.2 82.5 78. 83.9 91.2 78.2 80.5 78.5 78.1 86.2
85. 81. 79.9 79.6 78. 86.5 77. 78.4 79.4 79.3 78.3 79.2 79.7 85.2
80.2 80.2 77.2 78.2 82.9 77.4 81.9 79.9 78.4 77.1 79.3 79.2 82.1 80.
81.1 78. 87.2 85.5 79.4 81.1 78. 83.7 77.9 82.9 79. 79.9 79.3 83.
83.3 83.1 81.4 84.5 79.1 77. 82.5 80.2 81.7 79.9 78.9 78.5 81.5 78.9
82.5 79.1 84.6 81.8]
Linear Regression Predictions:
[78.03836684 78.75788189 80.3805173 82.90392869 83.29585635 78.89490794
82.45084736 78.31622613 77.68986592 77.31412932 79.31727904 77.83087213
77.56585883 78.4552804 79.60328697 79.88662966 78.63595091 79.38387256
83.04672164 80.38932452 79.14339234 78.69824323 77.93508276 79.20566013
81.76832845 90.21182395 77.76864644 79.34087261 78.58027771 81.1927403
79.21209311 92.92481315 85.84749452 79.04314007 80.10917765 78.92308551
92.68461335 78.91580508 83.1113927 79.79145235 80.13800859 78.54264552
77.91572175 77.69298048 80.19708684 80.90788001 77.74558136 92.53787032
79.88548843 79.45480313 77.23278525 82.56277232 78.4437192 78.45036348
78.92710013 79.55791094 81.66993457 79.34723793 81.4120579 82.58996381
82.03327102 78.98752069 78.1440222 77.76900082 78.87211662 80.01978968
78.98321613 77.00474972 79.83448976 81.43079953 78.56732488 83.23826362
77.55471234 79.21750731 79.08908676 77.64902032 77.83114085 76.80110644
80.14959285 82.39615273 80.3790081 80.59053875 78.86678322 78.82294858
78.85489377 79.29071045 83.15715507 78.78236543 78.43891291 85.52153112
78.069948 79.50513422 84.85828545 80.8201056 84.00821609 81.43457539
79.36584305 78.61644679 78.99869728 79.40334852 80.11925802 80.40682583
81.55158233 79.53997527 79.50693303 77.44344485 79.31890433 82.1451289
76.75236124 79.42851605 79.01554888 79.27204876 81.85232303 87.80343161
79.99223174 83.01292018 82.15784367 89.38594477 77.16629057 77.35227191
81.52091167 77.54975783 82.33171803 78.39208757 79.92825503 79.78559788
79.05022805 78.3069711 80.21275063 90.59930479 80.06492639 81.9879527
78.97582126 83.85590902 97.43840635 79.94164473 80.22613963 80.72847007
79.04949028 85.00355795 84.90151608 80.25978307 79.40682983 79.55598935
79.10043448 86.60667029 77.79023599 79.6559588 79.68533615 79.10553413
78.94961754 78.25676928 79.18823376 85.46157113 81.35407524 79.43014979
77.25845746 78.31820192 81.72879285 77.75951146 81.17928582 79.57246342
77.67666327 77.47608848 79.10885883 81.59630826 82.23878869 79.65440451
80.70255437 78.93048119 88.54737709 85.0654466 79.5512132 79.51123185
79.7487101 82.90368887 78.78018286 82.28285973 79.06686076 79.28340752
78.61179801 82.21390065 83.21908908 82.84143436 79.89298314 82.87648768
79.9610413 76.96174788 82.77711252 79.89086537 81.14861375 80.6052213
78.6683207 77.89406325 80.66017983 78.52518561 81.70803785 78.70725557
83.83106035 80.6117924 ]
Voting Regressor Predictions:
[77.56918342 78.57894095 79.99025865 83.15196434 83.24792818 78.49745397
83.47542368 78.25811307 77.64493296 77.20706466 78.70863952 77.91543607
78.33292942 78.2276402 80.15164349 79.89331483 78.21797545 79.79193628
83.32336082 80.49466226 78.62169617 78.44912161 78.76754138 79.40283007
82.13416422 89.40591197 77.48432322 79.32043631 78.34013886 81.49637015
79.25604655 90.86240658 85.97374726 79.02157003 80.65458883 78.56154276
90.99230667 78.45790254 83.45569635 79.69572617 80.01900429 78.37132276
77.95786088 77.89649024 79.79854342 81.60394 77.42279068 90.91893516
79.94274422 79.32740156 77.11639263 82.78138616 77.8218596 77.82518174
79.06355006 79.67895547 81.93496729 79.12361897 81.95602895 83.14498191
81.81663551 79.14376034 78.1220111 77.93450041 78.43605831 80.05989484
79.34160807 77.05237486 79.76724488 82.16539977 78.48366244 83.46913181
77.42735617 78.80875366 79.24454338 77.67451016 77.41557042 77.00055322
80.07479642 81.89807637 79.53950405 80.59526938 79.08339161 79.06147429
78.92744689 79.64535523 83.62857753 79.29118272 78.96945645 85.36076556
78.034974 79.55256711 84.77914273 80.9600528 85.05410804 81.91728769
79.03292153 78.30822339 79.09934864 79.45167426 80.00962901 80.85341292
82.02579116 79.01998763 78.80346651 77.27172243 79.20945216 82.12256445
77.67618062 79.41425802 78.70777444 79.08602438 81.72616151 87.0517158
80.24611587 83.30646009 82.07892183 88.39297239 77.83314528 77.57613595
81.56045584 77.32487892 83.06585902 78.24604379 80.26412751 78.94279894
79.22511403 78.20348555 80.15637532 89.1496524 80.13246319 82.24397635
78.48791063 83.87795451 94.31920317 79.07082236 80.36306982 79.61423503
78.57474514 85.60177897 84.95075804 80.62989154 79.65341491 79.57799467
78.55021724 86.55333515 77.395118 79.0279794 79.54266808 79.20276706
78.62480877 78.72838464 79.44411688 85.33078556 80.77703762 79.8150749
77.22922873 78.25910096 82.31439642 77.57975573 81.53964291 79.73623171
78.03833163 77.28804424 79.20442942 80.39815413 82.16939435 79.82720225
80.90127718 78.46524059 87.87368855 85.2827233 79.4756066 80.30561593
78.87435505 83.30184443 78.34009143 82.59142987 79.03343038 79.59170376
78.955899 82.60695033 83.25954454 82.97071718 80.64649157 83.68824384
79.53052065 76.98087394 82.63855626 80.04543269 81.42430688 80.25261065
78.78416035 78.19703162 81.08008991 78.7125928 82.10401892 78.90362778
84.21553018 81.2058962 ]
This is a visual representation of the GBR model, which GBR.py implements:
Automatically created module for IPython interactive environment
sofifa_id 0
age 0
overall 0
value_eur 0
wage_eur 0
..
team_position_RS 0
team_position_RW 0
team_position_RWB 0
team_position_ST 0
team_position_SUB 0
Length: 1036, dtype: int64
MSE: 0.3163
size of prediction: 200
prediction:
[77.50066071 80.28166753 83.31445071 83.94106728 77.56628816 82.04748441
80.90129235 82.15505717 80.05711914 82.20841276 78.53161703 79.26524484
78.54827103 79.85468865 77.61092536 77.28587469 79.94905361 81.86297481
78.77746635 79.86319322 78.43557098 79.3266269 86.13456282 78.31285015
77.43495862 85.26201168 85.5060164 77.5130349 81.19944164 79.46969341
80.6592216 83.39882666 77.84822156 82.10207437 85.87807291 82.25905858
77.86798187 78.27475122 79.05309264 87.87071948 77.46228843 83.96412312
77.24764857 83.94333067 82.1354546 87.65383486 79.04751774 79.86771727
80.16661881 78.8387045 82.24575361 78.56874647 89.48778287 78.8906047
81.11172271 79.33231374 82.28744502 77.79663792 77.2501399 78.60015319
77.44469501 82.96250222 79.85404315 77.22090607 78.2743279 78.38763939
78.01734783 79.67180228 86.46634985 78.0259167 77.7755417 78.72465838
79.88594339 78.31981828 83.05113843 78.19249458 77.14882744 81.3548976
77.33279517 78.19106181 79.17693421 79.22502099 79.68512926 78.54160983
84.51476788 77.4935367 78.30159344 85.03773847 82.02317544 79.53437061
79.49703968 78.83153536 78.81653283 77.78965391 77.25507205 78.76247257
79.1869876 81.82490134 86.01759127 79.49883842 78.35573612 78.20093623
77.09192911 83.23322834 78.74506866 82.30301661 84.01836932 77.09045648
77.00857863 79.91414467 78.61494809 80.75245044 83.54869157 78.48421261
80.05918243 78.93763055 80.07493731 89.58083503 84.50994691 79.09247717
80.69234627 79.67676377 79.12016482 83.00120896 79.05543383 79.70385982
77.13686445 81.83966241 78.80526508 82.71278731 78.40441402 77.2585265
78.25354423 79.47757313 77.2592911 79.85877906 78.50837313 79.24923496
77.30182277 77.84623743 81.60059987 78.53042619 77.64705151 78.17631989
87.25798996 77.18834078 84.16967432 78.7989075 86.53967546 81.39753561
79.91475212 83.42456783 83.56030036 77.78314278 82.32831566 80.1248439
84.88697932 78.55719757 80.15247011 78.427732 79.47073472 82.98286937
78.09574948 80.38084754 77.37849072 78.48133298 82.23017174 80.62664838
85.51626085 88.53508341 82.20420421 78.446413 79.0367478 78.9351531
78.58657158 79.65101048 85.90143557 77.77731956 80.91362764 78.14178555
82.45367028 82.15640929 79.27690007 79.71285012 82.1594059 77.98438734
78.47671966 77.81387676 81.16954797 77.83113335 82.31143793 82.87862077
88.24980231 78.5633741 78.72232146 83.93781777 77.55942918 79.12881541
84.5657212 81.97213144]
test score: 0.9642
Prerequisites for this section:
- You most likely will encounter an error installing XGBoost via
pip. - You'll find instructions on the vendor's website.
GBR_XGBoost.py implements this evaluation.
Automatically created module for IPython interactive environment
sofifa_id 0
age 0
overall 0
value_eur 0
wage_eur 0
..
team_position_RS 0
team_position_RW 0
team_position_RWB 0
team_position_ST 0
team_position_SUB 0
Length: 1036, dtype: int64
MSE: 0.5716
size of prediction: 200
prediction:
[80.03222639 84.09668454 77.4990349 83.90897344 82.75383275 79.08834487
84.58305946 89.5579993 77.86457629 78.67911865 80.87253465 78.36024874
82.18699615 79.62991533 78.29861031 84.85317724 78.42132316 78.72223459
82.310451 78.28817555 79.11350053 78.9786203 82.12323901 84.47856302
78.30486517 79.39080017 78.52160768 82.16451208 77.64708466 84.03558402
79.17416666 80.12226739 81.79389443 85.51204361 78.07534453 84.28879268
81.95918051 79.13588716 80.99331227 78.73419168 77.13144436 79.46984439
79.00573627 82.13973596 79.66391275 77.34965516 88.01157852 77.74126079
77.45649445 89.27455075 80.26830984 77.81757566 78.98811374 77.42444373
80.08777601 80.18912312 79.25385843 82.27037325 80.77382527 79.23805112
77.12710556 77.43700652 78.31070568 84.85043363 79.19690423 86.82779088
77.48798525 83.14243547 78.59483253 79.92419543 78.5991908 80.6732223
78.10356097 78.50146965 80.07938471 82.55160325 82.87679393 78.11979762
77.2203668 81.95982516 84.81777841 78.39617945 84.01292525 77.80269557
81.20365214 79.08039902 78.84005939 79.43974147 79.40510677 78.34284183
86.85758254 79.97198645 82.0095076 77.44057997 78.29609098 79.02671975
78.29591132 81.43450791 84.73088187 77.15577228 78.03960032 82.58139374
79.92627697 79.43682059 81.51273748 78.4977044 82.04990563 78.94958348
83.1724797 77.89164883 78.26969192 82.4432361 77.5549681 79.49820752
79.91824286 84.85586383 79.55528423 79.01372867 79.29515224 89.70541514
82.39701381 78.01315304 81.78587048 79.3689853 90.62610548 88.92374345
78.39089172 81.33968295 87.71122281 79.36390568 86.00138423 79.3174761
77.17504858 79.09568792 79.34677587 80.28176461 80.10116818 78.99135755
78.20928761 80.43413371 77.99559847 77.07870184 77.22699687 79.8173736
79.25695813 78.55781575 79.09321119 81.30323944 79.17703957 82.02496929
83.04925076 90.0804803 78.49530844 79.3867113 78.41856478 80.89064543
77.28325627 78.36591959 78.51530599 78.30921988 77.68304834 78.96262263
83.55328106 79.43232714 79.03043688 82.40277178 78.60538041 78.65725581
79.02929328 83.33022268 78.06864695 78.40603654 81.10151935 77.22853695
79.13065378 78.17660846 78.14502607 78.1200517 78.25065498 81.20491606
77.5237754 80.68407877 80.70213674 80.21050993 78.3948194 79.23724727
77.82331637 80.03810194 85.40056719 86.4962502 78.15580661 78.54686391
79.080409 79.20619623 87.21497868 85.04668303 78.34278693 79.44005024
81.51716489 78.01341495]
test score: 0.9438
XGB predictions:
[79.83954 84.4557 77.47257 84.12741 83.20348 79.71965 84.507614
89.625 78.0602 78.661 80.64833 78.12047 82.4132 79.83159
78.38893 85.62745 78.54127 78.999245 82.55367 78.563034 79.382576
79.468605 83.43974 84.006615 77.85358 79.73917 78.19146 82.58214
77.62072 84.11112 79.54766 80.08623 81.31419 85.5844 78.25554
84.378136 81.88007 78.74612 80.12361 78.14174 76.9646 79.90122
78.03387 82.06241 80.692795 77.23939 88.83091 77.466995 77.07657
89.121635 81.034294 77.63385 79.61537 76.92048 79.50215 80.21723
79.386856 82.24164 81.00054 78.81714 76.97433 77.35805 77.87139
84.98284 79.02129 86.419304 77.24919 82.99761 78.72895 79.58036
78.10361 81.53009 77.98868 78.335884 79.988335 83.097336 83.42798
77.75666 77.01154 81.84401 85.19378 78.474 83.81032 78.052536
81.28176 79.183525 78.722916 79.9721 79.40503 77.88457 86.89866
80.04575 83.17865 77.27263 78.17345 79.19656 78.13317 81.018196
84.698654 76.94766 77.98242 83.193825 81.39867 79.4087 81.16841
78.24879 82.1293 79.13059 83.393974 77.80798 77.89779 82.71875
77.591385 79.13242 80.12771 84.70757 79.64734 79.000305 79.13862
89.196785 82.98979 78.04829 81.712296 78.55921 90.49355 88.083244
78.30779 81.04324 88.64356 78.98328 87.57836 79.71369 77.01918
79.051926 78.898705 80.5618 80.590225 78.92508 78.23409 80.62751
77.89316 76.95264 76.738014 79.87418 79.37796 78.22808 78.67744
81.558495 79.19377 82.00043 83.21194 90.49752 79.06238 79.36658
77.84551 81.10486 76.89615 78.581696 77.48683 78.50707 77.101974
79.19367 84.44372 79.21375 79.4178 82.26863 78.20399 78.556244
79.087 83.19202 77.80677 78.01735 80.80606 77.066 78.702385
78.10479 77.89611 78.19101 78.05385 81.28604 77.4743 80.98116
80.73555 80.46555 78.54537 79.23917 77.39644 80.081 85.877945
86.48322 77.84128 78.39153 79.63295 79.049034 87.65402 85.10107
78.00571 79.45789 81.05464 77.94737 ]
Accuracy: 75.50%
ModelComparison.py implements this procedure.
Automatically created module for IPython interactive environment
KNNReg| Mean = 0.630142 STD = 0.179611 Variance = 0.032260
MLR| Mean = 0.812274 STD = 0.068709 Variance = 0.004721
VReg| Mean = 0.857354 STD = 0.060742 Variance = 0.003690
GBR| Mean = 0.854664 STD = 0.074611 Variance = 0.005567
In this report, four simple regression-based algorithms and models were used to predict the overall player ratings from FIFA 20, of 200 outfield players based on a training data set consisting of 1000 players from which we removed the 200. The initial datasets had thousands of players and features but underwent pre-processing and cleaning.
The KNN regression model came out with an 81.66% accuracy; MLR, 88.93%; voting regressor, 95%; GBR, 95.94%.
The variable importance graph from the GBR algorithm shows that monetary features like wage, value, and players’ age, contribute the most to overall player ratings.
Footnotes
-
The FIFA video games consist of several ratings for each player. The aggregate of a given set of ratings is referred to as the overall rating, which is what this research intends to investigate. ↩
-
Goalkeeprs are precluded from this report as thier characteristics are completely different from outfield players. ↩