Wednesday, May 4, 2011

Win, Lose - But No Random Draw

TigerBlog Jr. had a lacrosse tournament this past weekend, something that is not all that uncommon for him.

Almost all of the tournaments he plays eventually crown a champion, and they get highly competitive, especially in the summer. The one from this weekend was a bit lower key, though it was playing off to winner in each of its four divisions.

These tournaments have highly varying formats, numbers of teams entered and ways of determining the winner. Most end up with a playoff setup, and they have quite different ways of figuring out tiebreakers.

Most of the summer tournaments have divisions, and teams play a round-robin to set up the playoff round. Only one tournament that he's played in has been a straight elimination tournament (with confusing consolation rounds after each team loses its first game).

TigerBlog is always fascinated by how the tiebreakers are designed, especially given his professional background.

In all cases, they start out with head-to-head, though not every team plays every other team in every tournament. Or, in some cases, there are three-way ties that need to be broken.

As an aside, TB is amazed at how many times he's heard a parent say something like "how come they're ahead of us if we beat them," without considering that their team also lost to a third team, one that the first team defeated.

Other tiebreaking considerations include goal-differential, with a maximum number of goals that can counted from a given game, this to discourage running up the score. Other tournaments only use goals allowed.

The tournament this past weekend had two six-team divisions, and each team played three preliminary round games. After that, a series of tiebreakers would be applied to come up with a ranking of 1-6 in each division, and then there'd be cross-over games to determine first, third, fifth, etc.

The nice man running the operation was trying his best to keep up with all the various outcomes of games, but it got a bit complicated at times, since not every team played every other team. At one point, he said that it would be impossible to figure out who would be playing where and when in the cross-over games at that point - until TB came up with it for him.

He thanked TB, who said that it was nothing, that it was what he did for a living.

In the first two years of the Ivy League lacrosse tournaments, TigerBlog has tried to figure out all the permutations of results to determine if anyone had already clinched a spot in the field of four or had been eliminated.

What he's learned is that it's nearly impossible to figure out until there are only two weeks left in the season, and even then it's not easy.

Then there's the question of whether or not the tournaments are worth it.

TigerBlog is strongly in the "no basketball tournament" camp, and he has been for decades now.

For lacrosse, though, he doesn't mind the tournament, precisely for the same reason he doesn't want a basketball tournament. In the case of basketball, it would destroy the credibility of the regular season; in the case of lacrosse, it enhances the regular season.

Heading into the final weekend of the 2010 season, all seven men's teams had a chance at the Ivy tournament and by extension, the NCAA tournament. In the end, the Ivy tournament didn't add any teams to the NCAA tournament, but it also didn't keep any out. Princeton and Cornell would have been in without an Ivy tournament, and they were in anyway.

This year, six of the seven men's teams had a shot at the Ivy field entering the final weekend. It remains to be seen the effect that tournament has on the NCAA field, but TB can't see any of the three Ivy bubble teams - Penn, Yale, Harvard - not making the field with an Ivy tournament loss that otherwise would have made it.

Of course, for Princeton, the season ended a week short of the Ivy tournament. Without an Ivy tournament, Princeton would be playing its final regular season game this weekend.

Without an Ivy tournament, though, the last few games of the season would have been meaningless for Princeton, who would have been mathematically eliminated from the NCAA tournament without the possibility of a .500 record.

If TB had to change one thing about the Ivy tournament - and about any determination of a tiebreaker in the Ivy League - it would be to eliminate the possibility of a random draw.

Had Yale come back to beat Harvard and Dartmouth beaten Brown instead of losing in overtime, then the fourth spot would have been decided by a random draw.

There's too much riding on these outcomes to have them determined randomly. TigerBlog knows of a few times where the league found itself having to determine important placings or even NCAA bids by random draw (2003 men's and women's lacrosse automatic bid; 2006 women's basketball three-way playoff pairings), and in every case, TB couldn't help but think there had to be a better way.

For starters, eliminating random draws protects the Ivy League from ever being accused of rigging a random bid (something TB never has even remotely suspected, but there are probably those who thought that was the case).

So what would be the best way, if there's, say, a three-way tie where the teams are all 1-1 against each other and have the same results against every other team?

How about RPI? Goal (or point)-differential head-to-head? Goal-differential against the top team remaining in the standings?

Yes, any goal-differential situation opens up the door to running up the score, but honestly, how often could that possibly come up? And maybe go with goals allowed, rather than goal differential.

Hey, it works on the youth level.

Summing up, TB would like the Ivy League to never have a basketball tournament, keep the lacrosse tournaments and eliminate random draws.

Is that asking too much?

No comments: