Were the Miami Marlins “lucky” to win 2 World Series?

Kevin Brown of the Florida Marlins delivers to the plate against the Atlanta Braves 14 October in game six of the National League Championship Series (NLCS) at Turner Field in Atlanta, GA. AFP PHOTO/Steve SCHAEFER (Photo by STEVE SCHAEFER / AFP) (Photo credit should read STEVE SCHAEFER/AFP via Getty Images)
Kevin Brown of the Florida Marlins delivers to the plate against the Atlanta Braves 14 October in game six of the National League Championship Series (NLCS) at Turner Field in Atlanta, GA. AFP PHOTO/Steve SCHAEFER (Photo by STEVE SCHAEFER / AFP) (Photo credit should read STEVE SCHAEFER/AFP via Getty Images) /
facebooktwitterreddit

It is often said that playoffs are a “crapshoot” and that the best team doesn’t always win. It is often said that it’s the “luckiest” team or the “hottest” team that gets those championship rings. Research however shows that it’s simply NOT true. Luck plays a part in the playoffs, but the best team usually does win. Let’s take a look at whether the championships won by the Miami Marlins (then Florida Marlins) in 1997 and 2003 are “deserving” by a true best team in the Majors, or if they were won by the “luckiest” team.

The Miami Marlins deserved to win our two championships.

It is often incorrectly assumed that the team with the best record is the best team. The problem is that trades and injuries take place and change the team’s actual talent “real potential”. If a team acquires new players at the trading deadline and plays much better, is it fair to use the prior pre-trade record to define the team’s ability at the moment?

Here’s what I once wrote about the 1997 Miami Marlins:

"They finished the season as one of the top teams in ERA and FIP and swept the San Francisco Giants (who had a worse real and Pythagorean record) in the NLDS. They then decisively beat the Atlanta Braves in the NLCS and then beat the Cleveland Indians (who had a worse record) 4-3 in the World Series."

The Miami Marlins (easier to use the current name) failed to win our division, but made the playoffs as the NL Wild Card team with a 92-70 record (the same Pythagorean) Our NLDS opponents the San Francisco Giants had a worse record (90-72) and an even worse Pythagorean record (80-82). We then decisively beat the Atlanta Braves in the NLCS (4-2). The World Series were close and we just barely beat the Cleveland Indians 4-3, that said they had a worse record (86-75 and 85-76 Pythagorean). While records shouldn’t matter too much (as I said above), they can still be used and this shows that the best team won the 1997 World Series.

What about the 2003 Miami Marlins? Here’s what I wrote about that team…

"They were better than their division winning Atlanta Braves in the second half (42-25 to Atlanta‘s 40-29). The San Francisco Giants (who finished with the best record in the NL) were just a little bit better (43-24), but both Florida and San Francisco had an identical 18-8 record in September. It seems to me that they “figured it out” in the second half and were among the elite teams in the Majors despite their 91-71 overall record. The New York Yankees (101-61) may have been better in the World Series, if not for shifting around hitters to new positions, bad luck in close games and injuries."

The Miami Marlins were just as good if not better than the San Francisco Giants, who we beat 3-1 in the NLDS. We had a better record than the Chicago Cubs (who we beat 4-3 in the NLCS) and it’s only the New York Yankees, who you can argue could’ve actually been the better team. One of the games in that series went into extra innings and another had a Bombers starter fall to injury (David Wells). One-run games have a big luck factor, so it’s an argument that can be made.

Ultimately, the Miami Marlins deserved to win the World Series in 1997 and probably in 2003 as well, or at least weren’t that “lucky” to do it. We deserve our championships and should receive more recognition for having them and for being more successful than most teams, based on our first season of existence.