Ivy League Weekly Roundup: Tiebreaker Madness

Last Week in the Ivy League: Princeton clinched at least a share of its 27th championship. Columbia kept its postseason hopes alive by beating Penn, throwing the final week into tiebreaker chaos. The Lady Bulldogs got hot, shaking up the other playoff race.

Where We Stand: Had Penn beaten Columbia, the four Ivy League Tournament participants would be set, and the final weekend would have been about seeding and Princeton’s quest for 14-0. Instead, every team is technically within two games of a playoff spot, so get familiar with the tiebreakers.

Everything starts with Penn and Columbia, who are currently tied for fourth place at 5-7. Penn hosts Dartmouth and Harvard; Columbia visits Brown and Yale. Whichever team wins more games this weekend makes the tournament, and whichever loses more games is out. (The winner will be the #4-seed, unless it goes 2-0 and Yale goes 0-2, in which case it gets the #3-seed.)

But if Penn and Columbia tie, things get complicated. Jonathan Tannenwald has the full details, but a summary:

Penn and Columbia both go 2-0: If Yale loses to Cornell (as well as Columbia), it will lose a 3-way tie with Penn and Columbia at 7-7. This is the only scenario in which the Bulldogs miss the tournament. Otherwise, this is probably the weirdest case: Penn and Columbia’s records would differ against only Cornell (favoring Penn) and Brown (favoring Columbia). Based on the second tiebreaker, the fourth seed would come down to which of those two teams finishes higher — which, in this scenario, would be determined by their head-to-head meeting on Saturday night. (A fun implication is that neither Penn nor Columbia fully controls its own destiny now.)

Penn and Columbia both go 1-1: If Penn’s loss comes to Harvard, Columbia gets the 4-seed. If Penn beats Harvard and Columbia’s loss comes to Yale, Penn gets the 4-seed. If Penn and Columbia beat only Harvard and Yale, respectively, the winner will depend on what Brown, Cornell and Dartmouth do in their other games.

Penn and Columbia both go 0-2: Columbia gets the bid, unless Dartmouth also beats Princeton to force a three-way tie at 5-9. In that scenario, Dartmouth gets the bid. (Cornell or Brown could also tie at 5-9, but they would lose to Columbia or Dartmouth on tiebreakers.)

On the women’s side, there is also drama at the bottom of the bracket. Penn (10-1) is likely to end up #1, and Princeton (8-3) and Harvard (8-4) have also clinched tournament berths. Cornell (6-6) jumped into fourth place with an upset at Princeton (with a former First Daughter in attendance), Brown (5-7) dropped two more close games, and Yale (5-7) swept the weekend to stay alive.

The juicy part: Yale and Brown visit Cornell this weekend. If the Big Red wins both games, it makes the tournament (and gets the #3-seed if Harvard gets swept). But if it loses to Brown or Yale, and the winner also beats Columbia, tiebreaker scenarios are in play:

Cornell-Brown: Brown wins the tiebreaker on head-to-head.

Cornell-Yale: Head-to-head is split, moving to the “highest-seed” tiebreaker. Yale almost certainly owns this based on its win over Penn, unless Princeton catches Penn for the #1-seed in the final week.

Brown-Yale (both teams sweep): As above, Yale almost certainly owns the tiebreaker with its Penn win, unless Princeton catches Penn for #1.

Cornell-Brown-Yale (Brown and Yale beat Cornell, lose to Columbia): This is quite unlikely, but Brown would win this tiebreaker at 6-8, based on its 3-1 record within the trio.

Got all that? I know how I feel:

Get rid of the “record against highest seed” tiebreaker. I love logic puzzles as much as anyone, but I’m sick of figuring out all the #4-seed scenarios in play. The higher-seed procedure is complicated, distracting from the actual games that happen. It has weird potential consequences, such as the race for a playoff bid coming down to a game that does not involve either team. Above all, it is arbitrary. By definition, any teams reaching this tiebreaker have had the same record against the same schedule. Why should it matter which games within that schedule they won or lost?

There’s a better alternative sitting one row down: The ratings indices. Not only are they much simpler, but unlike the higher-seed tiebreaker, they come with real advantages. They incorporate non-conference games, rewarding play over four months rather than two. And they are better measures of performance, resulting in a better tournament. Going forward, the Ivy League should use ratings indices as the second tiebreaker, after head-to-head.

…Better yet, ditch the tiebreakers altogether. For any tie that determines who is in or out of the tournament, why can’t the Ivy League hold a play-in game? (It would keep the tiebreakers for cases that affect seeding only.) Executing postseason games isn’t a trivial task, but everyone will already at the playoff venue on Friday. And because only the 4-seed is at stake, the standard for media and attendance should be relatively low, lowering the logistical pressure. It would be fairer — not to mention more fun — to settle ties on the court.

A three- or four-way tie makes this plan more complicated, requiring two days of play-in games starting Thursday. But the league has set up playoffs on four days’ notice before, even with higher stakes. (This would require the Princeton and Penn women to move their final-Tuesday game to an earlier date, as the men did this year.) It does mean some play-in participants would have to play four games in four days if they reached the final — but many other leagues already do so. Besides, given that most multi-team ties for fourth place would involve sub-.500 teams, it’s a fair disadvantage.

Weekly Awards:

Player of the Week: Devin Cannady, Princeton — Cannady lived up to the number on the back of his jersey, hitting three-pointers all over the Empire State (12 total in two games). After a slow start at Columbia, he sparked Princeton’s second-half run with deep shots, and he kept going for 26 points in Ithaca. The sophomore now leads the league by a wide margin with 37 treys in Ivy play.

Rookie of the Week: Bryce Aiken, Harvard — Aiken might be a freshman, but it appears he’s already taken the Harvard-Yale rivalry to heart: After dropping 27 points in New Haven, Aiken scored 22 in last weekend’s rematch, adding five assists and three steals. He’s battling Miye Oni for not only the Rookie of the Year award, but also the Swaggiest Ivy Player award, and he fortified both cases on this play:

Play of the Week: Conor Voss throws down the second-best dunk of Columbia’s season:

Power Rankings:

  1. Princeton (12-0) — Guess who has the best three-point shooting percentage in Ivy play? You won’t — it’s Amir Bell, at 56%. Bell shot an abysmal 3-25 out of conference; his turnaround is one of many factors driving Princeton’s 12-0 start.
  2. Harvard (10-2) — Read David Tannenwald’s senior-week profile of Zena Edosomwan. After taking an unusual path to high-school stardom, Edosomwan’s college career has been up and down, but his experience on and off the court should inspire future top-100 prospects to take Ivy League offers seriously.
  3. Yale (7-5) — Before conference play began, I wrote about how Yale’s hot outside shooting was unsustainable, but I didn’t expect regression to hit this hard: The Bulldogs are dead last in three-point percentage, as Ray noted this weekend, even after an 11-22 outing at Dartmouth. Alex Copeland has been the worst offender, plunging from 42% out of conference to 16% in Ivy play.
  4. Penn (5-7) — Penn got the three-point shooting it needed at Columbia on Saturday — but the Quakers were undone by inaccuracy closer to the hoop. They shot just 3-15 from 4-10 feet, an area where the league shot 37% last season. All three makes were from AJ Brodeur; Penn’s guards were 0-9 from an area where they tend to fling low-percentage shots.
  5. Columbia (5-7) — Grant Mullins is on pace for the Ivy Alumni Player of the Year award, averaging 11.5 ppg on 43% three-point shooting for Cal in Pac-12 play. The Daily Californian has more on Mullins’ season and background.
  6. Cornell (3-9) — Matt Morgan’s counting stats are well down from last year, but he’s still capable of individual brilliance — as Penn discovered in Friday’s second half. Chucking from NBA range over the Quakers’ zone, the sophomore scored 15 points (and assisted three more) in a six-minute span, erasing a 43-31 deficit en route to a game-high 26.
  7. Brown (3-9) — Another acceptable Ivy League tiebreaker is a head-coach dunk contest. Mike Martin looks like the clear #1-seed (unless any other coaches want to submit their own videos?):

  1. Dartmouth (3-9) — Dartmouth hosted Yale and an NBA All-Star Game broke out: The Big Green scored 1.18 points per possession and lost by double digits, 99-86. They’ve allowed 1.12 ppp for the full Ivy season, and only Brown is saving them from the defensive basement.

3 thoughts on “Ivy League Weekly Roundup: Tiebreaker Madness

  1. D Gates

    Your suggestion to use a slam dunk contest to settle tiebreakers is a good one. Here is an even better one: In FBS football, when there are not enough teams with at least a 6-6 regular season record to fill out all the berths in contracted bowl games, the NCAA fills in the few necessary additional teams by selecting from the 5-7 teams those that have the best academic standing.

    That is probably the only tangible way that the NCAA rewards academics in any sport, despite all the marketing materials focused on “student-athletes.”

    Let’s break tournament tiebreakers by US News rankings. All the schools are already focused on improving their rankings as it is, so let’s reward them for good performances.

  2. EK Rin

    Yale’s Copeland may not be hitting threes but still efficiently scoring at pretty high rate. One of top 3 guards in league efficiency it seems.

  3. Seawolf06

    One thing for sure.. the tournament has added much needed urgency/significance/fun/drama to the games played by teams in the middle of the standing towards the end of the regular season. Adding the tournament was a great idea.

Comments are closed.