clock menu more-arrow no yes

Filed under:

Do Expansion Teams get better by just existing?

New, 5 comments

Part of what needs to be addressed this offseason is how much roster churn is necessary, and how much would actually hurt the Sounders chances in 2010.  There is an expectation amongst many fans that a team should get better just by playing together, and if that is the case, then the Sounders should have very little churn.

Now, it might also be the case that past expansion teams had high churn (Toronto is one that did such), but there is possible statistical evidence that expansion teams tend to learn their league, learn their teammates in the second season.

I was inspired to research this because David Falk asked Adrian

What personnel moves or training adjustments are Sounders FC considering to address the lack of goals in the second half of 2009?

Hanauer: "We want to score more goals next season.  Part of that may be accomplished by players who have a season of MLS under their belts, some may come from better training and preparation, and some may be the result of personnel changes. Getting better in the off-season is a priority.  Period.  We are not going to speak about specific positions or players." 

Two of Adrian's three parts are related to no longer being an expansion side. Is his suggestion valid?

Frankly - Yes. At least from a quick and dirty check.  This quick and dirty check works under the assumption that expansion sides on average have a similar churn to the other bad teams in the league. Yes, the Fire and Sounders were not bad teams, but most expansion teams are.

On average expansion sides' offense improves by 18.7% in their second year.  Defense improves by 12.7% in that same time period.  To quickly check if this was just a case of regression I compared the worst two offensive teams in those same expansion years and their offense only improved by 6.9% and their defense got 2.7% worse.

Yes there are sample size issues.  Yes, there are roster churn issues. Certainly this is a rough guide.

But, roughly, Adrian is right. Another year playing together, another year training together should help the Sounders more than the rest of the league.  In fact extrapolating their numbers the Sounders would have a Goal Differential of +20 with a standard MLS level churn, and a standard MLS expansion side improvement curve.  This has a very low accuracy.  All it does is demonstrate that panic moves are not best, but sticking to the multi year plan is likely the best idea.

Numbers normalized for amount of matches played

Expansion 2nd % Difference exp ag 2nd ag % Diff
1.9375 1.5 0.774194 1.40625 1.125 0.8
1.4375 1.3125 0.913043 2.125 1.84375 0.867647
0.9375 1.40625 1.5 2.03125 1.53125 0.753846
0.96875 1.40625 1.451613 2.09375 1.3125 0.626866
0.833333 1.133333 1.36 1.633333 1.433333 0.877551
1.066667 1.2 1.125 1.266667 1.666667 1.315789
1.266667 ? 0.966667 ?
1.187308 0.873617
1.34375 1.6875 1.255814 1.84375 1.09375 0.59322
1.40625 1.03125 0.733333 1.5625 1.65625 1.06
1.0625 0.9375 0.882353 1.40625 1.3125 0.933333
1.25 1.125 0.9 1.15625 1.53125 1.324324
1.033333 1.333333 1.290323 1.5 1.3 0.866667
1.266667 1.833333 1.447368 1.6 2.066667 1.291667
1.133333 1.3 1.147059 1.433333 1.533333 1.069767
1.233333 1.1 0.891892 1.3 1.4 1.076923
1.068518 1.026988

The top section displays expansion sides, bottom displays the poor performers from the same year.