I personally don't think Maddux is a problem either, though he is overpaid at 9 million. Stat analysis is all about evaluating performance. Has Maddux been a decent performer for the Cubs? Surely. And is his ability to consistently stay in a game valuable? Surely. But no one will convince me that wins are an appropriate way to judge a pitcher's performance. There's too many variables there (mainly run support) that have nothing to do with the pitcher. If I may, I'll share an analogy from my own profession. If I assign my eighth graders a group project. I place four in a group and give them the assignment. Two students do all the work, but do it exceptionally. The other two do minimum work, but their co-workers pick up the slack. I may assign an "A" to that group for the project, but it is a poor assessment of each student's ability. If I want to know what each student is capable for, I will need a rubric that measures the individual contribution to the work of the group. The final goal of any team is a win. Each player on the team contributes to that. Now, I will go as far to say the individual pitcher contributes more every fifth day than an individual offensive player, but judging the pitcher on the final outcome is as naive as assuming that since my group of students produced an excellent project that each did "A" level work in the process.