Reply to Jeff Cup Girls Final Verdict

Forum Index » Soccer

Reply to "Jeff Cup Girls Final Verdict"

Subject:

Emoticons





More smilies

Text Color: Font: Close Marks

[quote=Anonymous][quote=Anonymous]Good Morning Everyone,

I am the OP. I created this because of the high level of debate between leagues. You know the saying, numbers don't lie.

Look, this is one of the very few events that has high numbers of both ECNL and GA teams. Based on statistics, this time around ECNL shows it has a much higher level of competition. I'm not talking about recruitment or anything else. Just outright competition. The stats give a VERY clear answer this year.

For those comments about Arlington and SOCA. I could take SOCA out, yes however Arlington IS currently a GA team so they get grouped as GA.

Recruitment is whole other conversation that has a lot to do with how a player takes action in the process, resources and many other things. [/quote]

I would be careful about asserting that there is any statistical relevance to your analysis. Here are some fundamental statistical flaws.
1) Comparing 2 subsets of leagues, one from an east cost set and one from a mid Atlantic region cannot be considered equal. You must have equal geographically constrained subsets to be meaningful.
2) Comparing two unequal sized subsets is statistically incorrect. On average, the ecnl to ga team ratio is around 4 to 1. In pragmatic terms, you have set up the analysis so that only 1 in 4 ecnl teams has to win in order to claim that ecnl is ‘better’ than ga. What you can state from your analysis is that when an ecnl team plays 4 games, they are more likely to win at least 1 game than a ga team is likely to win 1 of 1 game.
3) Your lack of confidence/sensitivity calculations is problematic. Given this is really an analysis of how teams who chose to go to the tournament compete for 3 games in 48 hours, you should calculate the above mentioned intervals/percentages. For example, the addition or subtraction of a single goal in anyone of the top brackets completely reorders the outcome. Variability for your analysis is quite extreme and puts a confidence interval well below 50%.
4) Numbers don’t lie, but many people don’t actually understand statistics and probability. This lack of understanding often leads to false and inaccurate conclusions.

The only real takeaway from this tournament is which teams are competitive in the top brackets.  If we assume any game with more than a 2 goal differential is not competitive, than we can analyze each bracket for competitiveness. What this shows is that each age group, the bottom one or two teams were more likely than not placed too high. This is the only thing that is VERY clear from the results.

[/quote]

Options

	Disable HTML in this message
	Disable BB Code in this message
	Disable smilies in this message

Review message

Search Recent Topics Hottest Topics