FRB Update: The Anglo-Welsh Cup

A few weeks ago, FRB wrote about the problems with determining the order of finish in a group based on games played against another group. It was mentioned that this system was being used in the Anglo-Welsh Cup in rugby union. Well, the group stage of that cup ended today, and sure enough there was a fiasco.

The 16 teams in the Cup were divided into four groups of four. Each team played four games against the teams in a different group, and only the group winners advanced to the semifinals. Groups 2 and 3 worked out O.K. this year, but groups 1 and 4, which played each other, did not.
The trouble was that Group 4 turned out to be a whole lot better than Group 1. After 16 games, Group 4 was 12-3-1. The final standings of this superior group were:

Harlequins	4-0
Bath	3-1
Ospreys	3-1
Exeter	2-1-1

And the weaker group:

Newcastle	2-2
Wasps	1-3
Cardiff	0-3-1
Leicester	0-4

Yes, three teams with winning records from Group 4 are out of the Cup, while Newcastle, with a worse record than any of them, will advance.

At least the World Club Challenge, which FRB wrote about in that earlier post, had a reason for using this bizarre system. That event was planned as a showcase of European and Oceania rugby, and by putting the teams from each region in different groups it was assured they would play only teams from the other region. But the Anglo-Welsh Cup groups are essentially arbitrary.
Until two years ago, teams played only three games against the teams in their own group. But there were complaints that three games meant that half the teams only got one home game, so the current system was concocted to assure two home games for all.  

There is no perfect sports system. Each competition has its own needs. But there are systems that just don't work at all, and this is one of them.