Anything more than a superficial look at those cancer statistics used to support one method of treatment against another should fill one with dismay.
Figures indicating the success of one treatment are contradicted by others. Symposia allowing interchange among experienced men in a field serve only to bring together opposing statistics, each supposedly collected in a scientific manner. Such differences are now so usual that they no longer produce animated discussions. They are becoming a traditional rite of cancer conferences.
After collecting some results on patients treated in my department and considering them for publication, I began to wonder how a statistician would respond if he were asked to comment candidly at the end of the article.
He might write "The arithmetic is accurate and I am satisfied that no patients with this diagnosis have been excluded from the study. The results are broken down into stages and as presented