[henryo]

Davis: Defense wins championship
"Clubs with lowest numbers in goals allowed are all playoff bound"

That defense wins championships.

More to the point in this case, teams couldn't even get into position to win a title without a great defense.

Eight clubs found their way "home," landing in the 14th version of Major League Soccer's playoffs. And you might notice that these very same clubs, without exception, rank 1-8 in league stats when it comes to goals allowed.

There's an undeniable correlation. As I said, once again the defense rests on this old sports bromide. The bottom seven clubs in goals allowed are home now, conducting exit interviews, planning for the offseason and sorting out where it all went wrong.

Notable Exceptions: (teams that led the league in defence but still missed the playoffs)
2008: SJE, 38 (4th fewest goals allowed)
2007: COL, 34 (joint-3rd fewest)
2006: LAG, 37 (2nd fewest!! )
2002: DCU, 40 (4th fewest)

On the other hand, quite a number of teams that sneaked-in with below-par defence:
2008: NY, 48 (3rd most goals allowed, eve made MLS Cup Final!! )
2007: NY/KCW, 45 (joint-3rd, KCW made Conf Final)
2006: COL, 49 (joint-1st, made Conf Final)
2005: CHI/NY, 50/49 (3rd/4th, CHI made Conf Final)
2004: NY, 49 (1st)
2003: NER/COL, 47/45 (2nd/3rd, NER made Conf Final)
2002: NER/COL, 49/48 (2nd/3rd, NER made MLS Cup Final, COL made Conf Final)
2001: KCW, 53 (2nd)
2000: COL, 59 (2nd)
1998: COL/MIA/NY, 69/68/63 (1st/2nd/4th)
1997: TPB/COL, 60/59 (1st/joint-2nd, COL made MLS Cup Final too)
1996: KCW/CLB, 63 (1st/2nd, KCW made Conf Final)

Hence the "correlation" might not really hold, given the numerous exceptions over the years.

Let's discuss...

[Balerion]

Originally Posted by henryo

Hence the "correlation" might not really hold, given the numerous exceptions over the years.

Let's discuss...

Paging voros...

The "defense wins championships" mantra is a huge oversimplification. For philosophical reasons, it doesn't make sense. To win you have to score more goal than the opposition, so scoring and defending are both valuable.


The more important point is that soccer is not American football. It's a flowing game and there isn't a team clearly designated as "offense" or "defense" at any given time. This kind of statement ignores the team concept of soccer. Goals Allowed is a statistic that is heavily influenced by the quality of a team's midfield and the ability of the team's attack to keep their opponents on the defensive. It's a good quick-and-dirty tool to see teamwide strengths and weaknesses, but this statistic reflects a lot more than just the defense.

[Pegasus]

While there isn't a completely different offense and defense it surely explains why coaches don't always play the flashy midfielder. He might make some exceptional plays but also likely never wins the ball. Plus / minus ratings might help but with ten other players it could easily be very misleading.

[arkjayback]

Originally Posted by henryo

Hence the "correlation" might not really hold, given the numerous exceptions over the years.

Dumb statement. You even said it, they are exceptions. You found, what, 24 exceptions in 14 years? That's less than 2 a season.

When a stat holds true 75% of the time, theres definitely a correlation.

[henryo]

Originally Posted by arkjayback

Dumb statement. You even said it, they are exceptions. You found, what, 24 exceptions in 14 years? That's less than 2 a season.

When a stat holds true 75% of the time, theres definitely a correlation.

No worries Sir, please allow me to present the case in more details below.

Number/Percentage of teams that defied the hypothesis "Playoff Teams must have Better Defence than Non-playoff Teams and vice versa" for each season:

==========================

2009: 0 (i.e. as mentioned by Steve, all playoff-bound teams have better defence than non-playoff teams, :-)

==========================

2008: 9 = 60% (= 9/15 teams)

- Playoff team with worst defence: NYR at 48 goals.
- # of teams with equal/better defence but didn't make playoff = 4 (SJE 38, DAL 41, TOR 43 & COL 45)

- Non-playoff team with best defence: SJE at 38 goals.
- # of teams with equal/worse defence but still made playoff = 5 (KCW, RSL, CHV, NER & NYR themselves)

- Hence total = 4+5 = 9

==========================

2007: 8 = 57% (= 8/14 teams)

- Playoff team with worst defence: NYR/KCW at 45 goals.
- # of teams with equal/better defence but didn't make playoff = 3

- Non-playoff team with best defence: COL at 34 goals.
- # of teams with equal/worse defence but still made playoff = 5 (incl. NYR/KCW)

- Hence total = 3+5 = 8

==========================

2006: 3+7 = 10 = 83% (= 10/12 teams)
2005: 2+2 = 4 = 33% (= 4/12 teams)
2004: 2+2 = 4 = 40%
2003: 1+2 = 3 = 30%
2002: 2+5 = 7 = 70%
2001: 3+2 = 5 = 46%
2000: 3+5 = 8 = 66%
1999: 3+2 = 5 = 46%
1998: 4+5 = 9 = 75%
1997: 2+4 = 6 = 60%
1996: 2+3 = 5 = 50%

==========================

As such, in 14 seasons of MLS, only once (i.e. 2009) did we see the hypothesis stood.

For the rest, the correlation are simply not strong enough, none of these season have seen a 'defying rate' of less than 25% (i.e. the 75% compliance mentioned).

Even by Wikipedia Standards (50% compliance for strong correlation), only 5 out of the 14 seasons supported the hypo.

Surprised? To be honest, I'm equally surprised too, upon working out the full set of numbers. :-)