Maximal information analysis: I-various Wayne State plots and the most common likelihood principle

Statistical analysis using all moments of the likelihood L(y|α) (y being the data and α being the fit parameters) is presented. The relevant plots for various data fitting situations are presented. The goodness of fit (GOF) parameter (currently the χ2) is redefined as the isoprobability level in a multidimensional space. Many useful properties of statistical analysis are summarized in a new statistical principle which states that the most common likelihood, and not the tallest, is the best possible likelihood, when comparing experiments or hypotheses.