Ivy Madness After 8 Games

[recspecs730]

Here’s my Ivy Madness odds after 8 games for every team.

Change after tonight:
Up: Brown (+7%), Dartmouth (+2%), Harvard (+2%)
Down: Princeton (-4%), Cornell (-3%), Columbia (-3%), Penn (-1%)

Unpacking the 3-way Tie for 2nd Place

Cornell (2-0)
Princeton (1-1)
Dartmouth (0-2)

Brown has wins over Cornell (1-0) and Princeton (1-1) which makes things interesting if it can get into a tie with either team.

Playoff Leverage for Tomorrow

Difference in playoff chances is win vs. if lose, agnostic of other results

Dartmouth: 20% (88% if win, 68% if lose)
Brown: 15% (25% if win, 10% if lose)
Cornell: 12% (99%+ if win, 87% if lose)
Harvard: 9% (14% if win, 5% if lose)
Princeton 4% (99%+ if win, 95% if lose)
Penn: 3% (3% if win, ~0% if lose)
Columbia <1% (<1% if win, 0% if lose)

Finally, differences in odds based on games other than the one you are playing in
Dartmouth: 8% on Brown-Penn (71% if Brown wins, 79% if Penn wins),
Brown: 6% on Harvard-Columbia (22% if Harvard wins, 28% if Columbia wins)
Cornell: 5% on Brown-Penn (94% if Brown wins, 99% if Penn wins)
Harvard: 5% on Cornell-Dartmouth (13% if Cornell wins, 8% if Dartmouth wins)

Finally, the distribution of wins for the 4th seed:

If you get to 8 wins, you probably avoid a tiebreak (at least vs 5th/6th place). 7-7 is the record of 4th place team more often, but results in a tiebreak about a third of the time. All this to say, is buckle up, things could get pretty interesting over the next few weeks.

[recspecs730]

Update tonight after round 9 – just 5 games left per team

Yale officially clinches (just a spot, not yet the 1 seed, though numerically that’s getting rounded to 100%)

Princeton drops 3.5% while Cornell drops 12% and Dartmouth gains most of those loses, at 15%. Princeton drops less given that it’s chances of losing to Yale were already factored in as a decent bit more likely than Cornell losing to Dartmouth.

I’m a little surprised Brown holding steady at 25% range given its strong weekend, lower than what my gut tells me given recent league form. I think my most likely explanation is Dartmouth leaping Cornell in the standings is really bad for Brown. In a 2 way tie between Brown and Princeton, H2H is tied (1-1), neither team has beat Yale (Brown still has one more shot), and #2 in league is Dartmouth (who Princeton already beat, and Brown already lost to). Harvard’s win today also hurts Brown since Harvard owns H2H currently over Brown. All this to say, the the most likely path for Brown coming into today was Dartmouth fading, and that just hasn’t happened. The gains from Princeton/Cornell losing and roughly offset by being in a worse tiebreaker situation w/ Dartmouth + Harvard wins? And finally, rightly or not analytical models/vegas still think Cornell/Princeton (and now Dartmouth) are all better than Brown. Brown still scares me the most as the 4 seed I wouldn’t want to face in the semis with them hosting. Anyway, just my musings – will be a fun race to unpack this chaos over the next few weeks.

[Northbounder]

Love it - thanks for all you do with these. I have no idea what to make of the league after today. I’m with you that Brown, by the eye test, seems like they “could” be higher than their number. Big wild card if they make it.

And boy does the season feel like a missed opportunity for Columbia after that noncon start. Obviously no proof that things go differently…but I wonder what direction they’d have gone had they held on against Yale at home (and gotten in the win column early).

[james]

as a yale fan i cheer against brown from here. they are so dangerous in a tourney format and for so many reasons.

that tells a story.

[iogyhufi]

It really hurt them to lose Rubio de la Rosa. Columbia has had awful injury luck in the Engels era.

[iogyhufi]

Looks like Saturday is the big day this week. I would think Princeton-Dartmouth and Brown-Cornell are going to have huge ILT leverages. (Luke, I dunno if I’ve said this already, but I love the matchup-leverage feature on your R Shiny site.)

Harvard-Princeton is pretty big too. This seems like a make-or-break weekend for Princeton.

[ToothlessTiger]

Huge weekend for Tigers for sure. Fortunately the games are at Jadwin. Saturday is Alumni Day for Princeton. Basketball fans will witness the only women and men doubleheader of the season. The women play Columbia at 5:30 with first place on the line. Looking for big crowds at both games.

[recspecs730]

Thanks for the kind words. Here’s the leverage at least for Friday.

Harvard 29% leverage (33% playoffs w/ win, 4% w/ loss)
Princeton 22% (97% win, 75% loss)
Brown 16% (29% win, 13% loss)
Cornell 10% (92% win, 82% loss)
Dartmouth 8% (94% win, 86% loss)

Other large swings (> 5%) based on games a team doesn’t play in:

Cornell 9% on Brown/Columba (89% if COL wins, 80% if Brown wins)
Dartmouth 9% on Harvard/Princeton (91% if Princeton wins, 82% if Harvard wins)

Harvard/Princeton seems to be the big one on Friday. Yale/Cornell to a lesser extent for Cornell. This is in part because Cornell’s existing playoff odds already factor in that a loss to Yale is more likely than Princeton losing to Harvard (and thus the latter is more detrimental).

While it’s not hard to compute leverage for Saturday, I think it’ll be more interesting after Friday given the additional context we will know by that point, so I’ll put some stuff out there w/ playoff update either Friday night or Saturday AM.