Saves above expected, the true measure of a closer's success

What is the role of a closer in a baseball team?
Well, as the term suggests, he should close a game..after all, we don't call these guys "comeback firestarters" or "strange bearded guys with a tendence to make games just a little bit too exciting", do we?
A closer walks out of the dogout and with methodic precision takes care of the other guys, better not more than three, adding an "x" on the winning column of his team.
With this in mind, I checked all the pitchers in the Majors who, in the last 20 years, entered at least 10 games in a season with a save opportunity in sight, and saw what they did with that opportunity.
Only a game ended with a save was good, not those situations when they pitch so poorly that they blow a 3 run lead but so lucky that their team scores at least a run in the bottom half of the inning to award them the most undeserved of the wins (while the starting pitcher is fuming in the locker room..).
As a first point, it turned out that almost every year, 4 times out of 5 (actually, 20.644 out of 25.399 times, or 81.28%) the guys in this group got their save.
Applying this percentage, as it slightly changes through the year, to the save opportunities they had in the season, we can get what the expected number of saves for every pitcher would be.
For example, in 2004, the average percentage of saves was 82.722%; Trevor Hoffmann had 45 save opportunities in that season, so applying that percentage, we would expect him to save 37.225 games.
Hoffmann actually saved 41 games, thus getting 3.775 saves above the expected number that the average closer would get to his team in that season.
Mariano Rivera got the ball with a save in hand 57 times, that year, so with the expected percentage of 82.772%, he should have saved 47.151 games. He actually came up with 53 saves, or 5.849 saves above the expected.
5 saves above the expected is a terrific feat, usually good enough to rank a closer in the top 3 of a single year, but it's not good enough to make the short list of the best seasons a closer had in the last 20 years.
Eric Gagne leads the pack with his immaculate 2003, whe he was a perfect 55 out of 55, gaining a 9.532 SAE that is simply astonishing. Trevor Hoffman is close second with his 1998 campaign, while Dennis Eckersley showed his brilliance with 2 seasons in the top 7.
Everybody's favorite closer, Mariano Rivera, is not in the short list, but his hall of fame career shows through when you consider the resiliency with which, in 14 full seasons as a closer, he saved 554 games out of 617, or 89.789% of his save opportunities, which is 8 points above the Leagues' average.
With an average of 3.419 saves above the expected in the years from 1997 to 2010, he really is in a league of his own.