Well they are judging a guy on past events that are deemed more reproducible than raw goal/assists totals. So its easier to replicate than something that involves more luck like a shooting percentages. Its the higher likelihood of replicability (creating good chances) than the replicability of a sudden spike in shooting percentage.
- sditulli
I disagree. If you take a large sample, say 2-3 years, and look at a players "raw" goal/assist/shot/shooting% totals (I find that term "raw" funny when NOT dealing with advanced stats), that should account for fluctuations which should represent what to expect as opposed to how many chances/shots he gets over a segment of games.
Basically, my whole take on analytics is that they're a good tool to verify things and make broad assumptions on what is likely to happen in the near future. But using them as a be-all/end-all tool and using them as a definitive source of what will happen in the future, as the blog author does, is making erroneous assumptions. Also, simply out-CORSIing a team does not mean you will beat them in a game, or in a best-of-seven series. That only happens when your team scores more goals than the other team, despite the amount of shots/chances you have.