The van Oss - Chaudhury - Good theory ( vOCGT) was checked for a small artificial set of the work of adhesion input data calculated for 9 solids and 7 liquids. Taking from the literature the data for Lifshitz - van der Waals ( LW) component and acid and base ( A and B) parameters for 7 liquids and the values of the component and the parameters for 9 solids ( close to those in the literature), the work of adhesion was calculated and its value was assumed to be free of error. Next, new values of the work of adhesion were obtained by adding a random error of normal distribution belonging to 11 distributions of a mean value equal to the errorless work of adhesion value and standard deviations from 0.1 to 60% of the mean value. The LW components and A and B parameters for these solids were back-calculated for each solid and the error level by solving 20 3-equation systems. These 9 solids were grouped in 3 sets of 3 solids in each, and for each of the solid sets the overdetermined system of equations ( of matrix 7 x 3) for these 7 liquids was solved. The root mean square errors ( RMSEs) of the LW component and A and B parameters were linear functions of RMSE of the vector ( matrix) of the work of adhesion in both solution methods of a set of equations. It was found that a solution of the 3-equation set of the vOCGT was always exact for all liquid triplets. Erroneous LW components and acid and base parameters are obtained because quite a different set of equations ( caused by an existing error in the data) is solved than in the case of error-free data. There is a linear transformation from the input error in the work of adhesion vector ( matrix) space into the output error in the solution vector ( matrix) space, and the inverse ( or pseudoinverse) of the matrix A is the transformation matrix. In the case of a 3-equation set there is a linear relationship between the total RMSE of the solution and the condition number of the matrix A. The higher the input error in the work of adhesion data the higher is the influence of the condition number on the error in the solution. The RMSE value of the solution of an overdetermined system of equations was about 10-times lower than the mean value of RMSE calculated for the same liquids used as separate triplets.