| 1 | 
                    
                        
            [q]I made no such claim as that makes no sense to claim. Where do you have a dataset featuring these randomly balanced games? [/q]
                     | 
                    1 | 
                    
                        
            [q]I made no such claim as that makes no sense to claim. Where do you have a dataset featuring these randomly balanced games? [/q]
                     | 
                
                
                    | 2 | 
                    
                        
            Then your claims are mutually exclusive.
                     | 
                    2 | 
                    
                        
            Then your claims are mutually exclusive.
                     | 
                
                
                    | 3 | 
                    
                        
            \n
                     | 
                    3 | 
                    
                        
            \n
                     | 
                
                
                    | 4 | 
                    
                        
            - You say that it's nonsensical to claim that a randomly-balanced dataset cannot be used to evaluate players.
                     | 
                    4 | 
                    
                        
            - You say that it's nonsensical to claim that a randomly-balanced dataset cannot be used to evaluate players.
                     | 
                
                
                    | 5 | 
                    
                        
            \n
                     | 
                    5 | 
                    
                        
            \n
                     | 
                
                
                    | 6 | 
                    
                        
            - You say that it's invalid to evaluate whether a subset of a dataset balanced using (elo, whr, and all games) increases or decreases predictivity of trueskill, because these games were balanced using elo and whr over all games and not using trueskill over only-teams.
                     | 
                    6 | 
                    
                        
            - You say that it's invalid to evaluate whether a subset of a dataset balanced using (elo, whr, and all games) increases or decreases predictivity of trueskill, because these games were balanced using elo and whr over all games and not using trueskill over only-teams.
                     | 
                
                
                    | 7 | 
                    
                        
            \n
                     | 
                    7 | 
                    
                        
            \n
                     | 
                
                
                    | 8 | 
                    
                        
            [tooltip=Well,
             
            a 
            person 
            who 
            actually 
            understands 
            the 
            subject 
            matter 
            cannot.
             
            Evidently 
            _you_ 
            can!]You 
            can't 
            have 
            both.
            [/tooltip]
                     | 
                    8 | 
                    
                        
            You 
            can't 
            have 
            both.
                     |