Critic Score Residual

  • From the above analysis, it is seen that all the models gives almost similar result with Critic Score. Yet the Best Training Model was found to be the "Ridge Model" as the MSE is lower than other models and the R2 score is higher by few points. With the MSE : 0.77 and R2: 0.14, we can say that the Error Value is quite high and the Ridge Score is too low therefore we can conclude that the Critic Scores are not a very reliable predictor of the future success of the video game sales. Although they do have a positive corelation with the sales yet it is a weak one.

Game Score Residual

  • From the above analysis, it is seen that all the models gives almost similar result with Game Score. Yet the Best Training Model was found to be the Ridge Model as the MSE is lower than other models and the R2 score is higher by few points. With the MSE : 0.86 and R2: 0.046, we can say that the Error Value is quite high and the Ridge Score is too low therefore we can conclude that the Game Scores are not a very reliable predictor of the future success of the video game sales. Although they do have a positive corelation with the sales yet it is a weak one.

Play Score Residual

  • From the above analysis, it is seen that all the models gives almost similar result with Play Score. Yet the Best Training Model was found to be the Ridge Model as the MSE is lower than other models and the R2 score is higher by few points. With the MSE : 0.78 and R2: 0.133, we can say that the Error Value is quite high and the Ridge Score is too low therefore we can conclude that the Play Scores are not a very reliable predictor of the future success of the video game sales. Although they do have a positive corelation with the sales yet it is a weak one.