Group Stage Scoring Models

NOTE: This is for the Final Release (Version 2.1) of the spreadsheet. The Models Page for the earlier version is here. A late change in the FIFA Regulations [pdf] back to the 2002 model sent us scrambling to update our three-model spreadsheet. With time running short, we did only what time allowed - a hack job!

The previous method required head-to-head tiebreaks for teams tied in Points alone. In the revised regulations, these tiebreaks will apply after (a) Overall Group Points (3 for a win, 1 for a tie); (b) Goal Difference [GD] (goals scored less goals allowed), and (c) Goals scored [G]. Teams still tied are evaluated as a new group "of the teams concerned." That is, in a three-way tie for second place, the three teams are evaluated as an exculsive Mini-Group.

Our modified spreadsheet will still serve three models:
here's a flipflop
In the example above, Model 2 calls for no head-to head tiebreak; Ecuador (G = 3) pips Costa Rica (G = 1). In contrast, Model 3 calls for a tiebreak as both teams have 4 Points. Costa Rica's 1-0 victory over Ecuador in the fourth group-match gives it the edge; Overall Group G never comes into play because the Mini-Group resolved the tie.

So...what happened?

What we're calling the "2005 Model" favoured head-to-head results over Overall Group GD and G in order to motivate teams to win all matches - rather than settle for qualifying for the knockout stage. It was a promising but rather complicated-to-implement, and probably unnecessary, system. A few whined but could be safely ignored. Spreadsheet writers buckled down to deliver a workable product. We were all set with brilliant features unheard of in any tournament tracker. Then out of nowhere, a butterfly flapped its wings. Someone ran a computer simulation that predicted an increased likelihood of drawing lots with that model.

We blinked. And decided to add a user-selectable third model with a second Mini-Group because it had the potential to solve further head-to-head ties before the Overall Group GD and G did their work. Mind you, the chance of the second Mini-Group being required was slim, but we found a case that it solved - brilliantly. (Who wants to see an outcome derived from drawing lots?)

Then FIFA blinked. Perhaps they took a look at our spreadsheet ... maybe someone even tried to figure out the merits of the Cascading Mini-Group Model. "Ridiculously complicated," he would've surmised. But it had the potential to lead to another butterfly-flap. Imagine a storm conjured by a hundred irate, patriotic spreadsheet-users:
"We wuz robbed — according to Model 3 of this here...er, potato spreadsheet [sic], we would've qualified and possibly gone on to win la Copa!"
That's no big deal, of course, unless their local press join in, triggering the tornado caused by a butterfly flapping its wings in a remote location that chaos theorists love.

Blink.

The story is fiction, of course. But one consequence of applying Overall Group GD and G before head-to-head tiebreakers is that our Cascading Mini-Group Model was neutered. Eunuched. Sliced, diced and tossed in the used-soccer-ball bin. It's unlikely that a second Mini-Group could resolve remaining ties. At least we think so though we can't be sure. Only the butterfly could figure that out. We're hoping he won't. Like many of us fans, he'll be watching the show. Yes, many are strapped-in prepared to ride an emotional roller coaster. One or two might even find that their ticker ain't up to it, and end up watching the final from that great green soccer field in the sky. The month will pass quickly and not without its share of controversy. And soon it'll all be over.

Enjoy it while you can.

Our obsolete Models Page is still online here if you're up for it. Faced with the prospect of updating it, I preferred to write something more entertaining. Life is short, and we've waited four years for le grand spectacle. Only the following paragraph was ported over from the old page:

Drawing Lots

If and when lots are required to settle a tie, the spreadsheet will request a precise solution. For example, in a 1-2-2-2 [3-way tie] scenario, all it needs to know is who qualified in second place and it will demote the other 'seconds' to a third place tie. There's no reason to draw lots for third and fourth places as they are deemed simply to have not qualified. In the example above, anything but a '2' entered in a tied row is redundant. Which also means ... there's no way to make an untied third place finish into a first ... so if the model is wrong, ç'est fini. There is no remedy.

P.S.

If you're a sports hack and you've read this far, how could you track the 2006 World Cup with anything but our spreadsheet?