He (unofficially) earned $49 last year, blowing the doors off the field. Only three other A.L. starters earned $30: Roy Halladay ($36), Francisco Liriano ($33), and Mike Mussina ($30). Expanding the earnings criteria for a very good starter, 12 starters earned $20 or more. This actually trumps 2005, when only 10 starters earned $20 or more.
Santana and Halladay were the only A.L. pitchers to earn $20 or more two years running. They are also the only A.L. pitchers to earn $30+.
Roto veterans know that there's a difference between overspending for the best hitter and overspending for the best pitcher. For example, you wouldn't think about spending $55 on David Ortiz. His 54 HR and 137 RBI would go a long way toward winning you the HR/RBI categories. But there are enough guys out there who come close enough to matching his production. If an owner goes $55 on Papi, it increases your odds of getting Vlad for $37 and getting a bargain. But someone goes $55 on Johan, you're either going to get Halladay or you're going to run the risk of getting screwed.
Let's look at 2005's $20+ class and see what happened to them in 2006:
It's not a pretty picture. Only Colon and Buehrle tanked, but these aren't guys you would have wanted to rely in to win in 2006.
The drawback to pushing Johan to $55 is that he's not the Pedro Martinez of 1998-2000 or the Greg Maddux of 1994 and 1995. In those years, they each earned in the high $60s/low $70s and paying $55 for one of them was guaranteed profit. So far, Johan's best Patton $ year to date saw him earn $53 in 2004.
You're paying for stability, in other words, but you do have to be careful. I can't tell you what 2007's Top 10 will look like, but I can tell you that Santana is the guy most likely to be on it. However, that just means that someone is all the more likely to get this year's Chien-Ming Wang and pull a $15 profit out of their hat. So you have to build your staff correctly around a Johan Santana to win.