Predictions of Scrabble game outcomes

knn.txt

=== Run information ===

Scheme:weka.classifiers.lazy.IBk -K 350 -W 0 -F -A "weka.core.neighboursearch.LinearNNSearch -A \"weka.core.ManhattanDistance -R first-last\""
Relation:     scrabble-game-states
Instances:    7273
Attributes:   8
              number_of_plays
              first_player_current_score
              second_player_current_score
              current_score_difference
              first_player_average_score
              second_player_average_score
              average_score_difference
              won_by_first_player
Test mode:split 66.0% train, remainder test

=== Classifier model (full training set) ===

IB1 instance-based classifier
using 350 similarity-weighted nearest neighbour(s) for classification


Time taken to build model: 0 seconds

=== Predictions on test split ===

inst#,    actual, predicted, error, probability distribution (number_of_plays,first_player_current_score,second_player_current_score,current_score_difference,first_player_average_score,second_player_average_score,average_score_difference)
     1     1:true     1:true         *0.546  0.454 (22,156,162,-6,14.181818,14.727273,-0.545455)
     2     1:true     1:true         *0.517  0.483 (24,166,179,-13,13.833333,14.916667,-1.083333)
     3     1:true     1:true         *0.519  0.481 (26,190,199,-9,14.615385,15.307692,-0.692308)
     4     1:true    2:false      +   0.41  *0.59  (28,196,225,-29,14,16.071429,-2.071429)
     5     1:true    2:false      +   0.438 *0.562 (30,211,233,-22,14.066667,15.533333,-1.466667)
     6     1:true     1:true         *0.853  0.147 (32,287,258,29,17.9375,16.125,1.8125)
     7     1:true     1:true         *0.956  0.044 (34,322,276,46,18.941176,16.235294,2.705882)
     8     1:true     1:true         *0.95   0.05  (36,337,288,49,18.722222,16,2.722222)
     9     1:true     1:true         *0.945  0.055 (38,346,296,50,18.210526,15.578947,2.631579)
    
     ...


=== Evaluation on test split ===
=== Summary ===

Correctly Classified Instances        1852               74.8888 %
Incorrectly Classified Instances       621               25.1112 %
Kappa statistic                          0.4961
Mean absolute error                      0.342 
Root mean squared error                  0.4043
Relative absolute error                 68.4884 %
Root relative squared error             80.7797 %
Total Number of Instances             2473     

=== Detailed Accuracy By Class ===

               TP Rate   FP Rate   Precision   Recall  F-Measure   ROC Area  Class
                 0.832     0.337      0.719     0.832     0.771      0.85     true
                 0.663     0.168      0.792     0.663     0.722      0.85     false
Weighted Avg.    0.749     0.254      0.755     0.749     0.747      0.85 

=== Confusion Matrix ===

    a    b   <-- classified as
 1047  212 |    a = true
  409  805 |    b = false