NCAA Bracket Strategy: Can You Stomach Picking Virginia to Advance Deep into the Tournament?

This year’s NCAA Tournament is wide open, or at least that’s been the saying all year. Yet, when it comes to picks in the brackets, many familiar faces are present. We like brand names. Heck, when we did our podcast and said what we think would happen, my picks look like everyone else’s. The most common Final Four on ESPN: #1 Kansas, #2 Michigan State, #1 North Carolina, #2 Oklahoma (and Oklahoma is the choice because there is not confidence in the rest of that region). That exact foursome is in over 10% of the brackets.

But let’s talk strategy. If you pick the exact same Final Four as everyone else, then your chances of winning depends on things like correctly getting the winner of St. Joseph’s versus Cincinnati and who emerges from the Utah/Seton Hall/Gonzaga group to get to the Sweet 16. So if you are going to go not only heavy chalk, but also popular chalk, you have to differentiate. It’s just a matter of where.

The smaller the pool, the more conservative you want to be. You don’t want an 8-seed winning. Chances are, if an 8-seed wins, you can still win if you get others right up until that point. But you do want to look for value among teams that have a reasonable chance.

Here are the top nine teams in terms of either public confidence or Ken Pomeroy champ odds, and you probably want one of these as your champ unless you are in a huge pool.

You’ll notice that the brackets are giving the public top seeds a lot of attention. Over 60% of the brackets are picking one of Michigan State, North Carolina, and Kansas to win the title. There is some good reason for this, but it also means there is little value. It’s not as extreme as last year when 60% were picking just Kentucky and I recommended picking someone else, because it’s a three-way split.

At the other end, what do Villanova and Virginia have in common? They have disappointed the last two tourneys as top seeds. Interestingly, this same perception phenomenon does not attach to Kansas, who has failed to advance to the Sweet 16 each of the last two years as a top seed, and lost in that round as a 1-seed three years ago. Of course, Kansas won a title in 2008, and lost in the title game in 2012 (they’ve underperformed their seed in every other year besides those in the last decade).

And so, my prime suspicion–rightly or wrongly–is bracket participants just don’t trust these teams.

I went back over the last 20 years of tournaments, and found all occasions where a team was (a) seeded as a 6 or better in each of the previous two years; (b) failed to win as many games as you would expect based on seed in either; and (c) was again seeded as a 6 or better in a third straight year.

There were 43 such occasions. The expected wins, in that third year, for the group based on their seed was 94.2.

The actual wins? 101. So these underperformers did better than expected in year 3.

I don’t see evidence that teams that lose earlier than you expect continue to do so. Some do, but I can point to many counter-examples the other way.

Some of those were traditional powers. Kansas, like I said, was dreadful from 2009 to 2011 in the tourney. I had them in the Final Four as a 2-seed, and they were a value because most people had North Carolina in excessive numbers. Duke in 2010 was a huge value, I won a pool that year because they were the #1 seed with the lowest support, despite great efficiency stats. Go look at their tourney results for the previous 5 years to see why.

Okay, you say but we trust those teams that are ALWAYS high seeds more than Virginia, who has emerged and has no history of doing it. I’d point you to Florida in 2006. It’s easy to play revisionist history knowing that they won back-to-back titles, but they were awful in the tournament from 2001 to 2005. They were 7th in Ken Pom but undervalued as a 3-seed, and rolled to their first national title.

Point being, maybe Virginia and Villanova have this non-clutchy gene, but I’m not sure we can deduce that from the last two years.

Here are the odds differences for the Final Four (for everyone with a 7% chance or better in either):

And here are the Sweet 16:

I would more disregard the bottom of this list–it’s all the top 4 seeds basically–and focus on the top. A pool contest will overvalue the top, because you have to keep advancing teams. You don’t want to pick every upset. I like to work backwards with my bracket. Pick the teams in the Final Four, down to the Sweet 16, and see which upsets I think are good value plays, against teams I didn’t have advancing deep anyway.

Indiana is a good value play IF you aren’t picking Kentucky to win it all (I think Kentucky going deep is also a good contrarian play though, too).

Wichita State, despite the buzz, is still a good value. Iowa is a good value if you don’t have Villanova advancing deep. Same with VCU and Oklahoma, and Wisconsin and Xavier.