Arc of Champions - 2024

[fasbjd]

Hey all, it's that time again for this year's version of the Arc of Champions.  Unfortunately, not including our Hoosiers (to visualize where we would stand, we would be just below where Yale is in Sector 4 - similar aOE with slightly better aDE).  As a reminder, this is based on data from Bart Torvik's T-rank site and uses his pre-tournament adjusted efficiencies.  This data has been available since '08 and when plotted and compared to team finishes in the tournament, provides additional insight in predicted finishes for teams.  For instance, this year, teams like Illinois and Kentucky that seem poised for a long run, are actually quite vulnerable due to inadequate defenses...and Auburn (seeded 4) is in a prime position to win it all. [Click on image to enlarge]

 

"Brief" summary of Sectors:

Along with Auburn, there are 5 other teams that make up the Sector 1 grouping:  Houston, UConn, Purdue, Arizona, and Iowa State.  Teams in this sector have an 18.1% chance of winning it all and 37.5% chance of making the Final 4.  Over 86% of eventual champions come from this grouping (despite it only representing ~7% of the field historically).

The rest of the championships came from the Sector 2 grouping (~13% of champions and 1/3 of finalists).  Overall, this is a larger group, historically making up ~24% of the field (though this year, it only accounts for ~13% of the field) - so your chance of winning it all if you're from this group is <1%.  Nearly 41% of the Elite 8 come from this grouping with each team from this group having just over a 20% chance of making the elite 8.  This year's 9 Sector 2 teams are Tennessee, UNC, Marquette, St Mary's, Kansas, Michigan St, Creighton, Duke...and Nebraska!

Sector 3 is also a historically large group (~24% of the field, 22% of this year's field).  This grouping accounts for the last 7% of finalists and account for ~21% of the Elite 8.  That said, these team's most likely finish is the round of 64 (~46% chance) and this is the first grouping where your chances of finishing in each subsequent round go down (~33% chance of Round 32 finish, ~13% chance Sweet 16, ~4% chance Elite 8, 2.5% chance of Final 4, and <1% chance of being a finalist). 

Sector 4 has accounted for the final 5% of Final Four participants but only 10% of the Sweet Sixteen, despite accounting for 22% of the field historically (this year has over 29% of the field).  Notable teams that fall into this grouping this year include UK, Illinois, Alabama, Texas A&M, and South Carolina.  These team's chances of making the Sweet 16 hover ~7% and the Final 4 1.4%.

The barrier between Sector 4 and Sector 5 has been named in honor of one Sector 5's greatest participants:  Saint Peter's Gate - and it only seems fitting that Saint Peters again sits in the outskirts of this sector, waiting to pounce!

 

 

[Drroogh]

Thanks @fasbjd your work is always appreciated!!

[fasbjd]

Here's how the field looks going into the Sweet 16.   Sector 1 has 31.25% of the remaining field (historically makes up 24.6% of the S16), 31.25% also for Sector 2 (historically contains 42.9% of S16 teams), only 12.5% are from Sector 3 (historically 20.8%), and a robust 25% from Sector 4 (historically only contains 10% of S16 teams).  All the remaining Sector 4 teams have "championship-level" offenses (all better than 3 champions since '08) but have suboptimal defenses.  In fact, all the remaining teams except for San Diego State could be considered "championship-level" offenses.  Defensively though, there is clearly a divide.  In general, it's safe to say that "championship-level" defense could be defined as an aDE of 94 or less (that would leave only 1 outlier champion - the 2018 Villanova team that had the best offense of any tourney team since '08).  Using that definition, the remaining team with "championship-level" defense would include UConn, Arizona, Houston, UNC, Tennessee, Iowa State, and San Diego State (listed in order of best to worst offense).  Purdue's offense is amongst the all-time elite (5th best amongst tourney teams since '08) and their defense falls just between "championship-level" and 2018's Villanova's team, so it would be unreasonable to discount them as well.  I would suspect that given the quality of offenses left in the field, the defensive liabilities of the non-"championship-level" defenses will be exposed and lead to their elimination.

 

[ledies22]

1)Anyone else

2)Uconn

3)Houston

4)Purdue

[5fouls]

47 minutes ago, fasbjd said:

Here's how the field looks going into the Sweet 16.   Sector 1 has 31.25% of the remaining field (historically makes up 24.6% of the S16), 31.25% also for Sector 2 (historically contains 42.9% of S16 teams), only 12.5% are from Sector 3 (historically 20.8%), and a robust 25% from Sector 4 (historically only contains 10% of S16 teams).  All the remaining Sector 4 teams have "championship-level" offenses (all better than 3 champions since '08) but have suboptimal defenses.  In fact, all the remaining teams except for San Diego State could be considered "championship-level" offenses.  Defensively though, there is clearly a divide.  In general, it's safe to say that "championship-level" defense could be defined as an aDE of 94 or less (that would leave only 1 outlier champion - the 2018 Villanova team that had the best offense of any tourney team since '08).  Using that definition, the remaining team with "championship-level" defense would include UConn, Arizona, Houston, UNC, Tennessee, Iowa State, and San Diego State (listed in order of best to worst offense).  Purdue's offense is amongst the all-time elite (5th best amongst tourney teams since '08) and their defense falls just between "championship-level" and 2018's Villanova's team, so it would be unreasonable to discount them as well.  I would suspect that given the quality of offenses left in the field, the defensive liabilities of the non-"championship-level" defenses will be exposed and lead to their elimination.

 

I picked every game of the tournament based on this data.  It resulted in me missing out on some of the early round upsets, but I am set up very well from here on out.  People that picked teams like UK to got to the Final Four would have benefitted from looking at this information.    

[Drroogh]

58 minutes ago, fasbjd said:

Here's how the field looks going into the Sweet 16.   Sector 1 has 31.25% of the remaining field (historically makes up 24.6% of the S16), 31.25% also for Sector 2 (historically contains 42.9% of S16 teams), only 12.5% are from Sector 3 (historically 20.8%), and a robust 25% from Sector 4 (historically only contains 10% of S16 teams).  All the remaining Sector 4 teams have "championship-level" offenses (all better than 3 champions since '08) but have suboptimal defenses.  In fact, all the remaining teams except for San Diego State could be considered "championship-level" offenses.  Defensively though, there is clearly a divide.  In general, it's safe to say that "championship-level" defense could be defined as an aDE of 94 or less (that would leave only 1 outlier champion - the 2018 Villanova team that had the best offense of any tourney team since '08).  Using that definition, the remaining team with "championship-level" defense would include UConn, Arizona, Houston, UNC, Tennessee, Iowa State, and San Diego State (listed in order of best to worst offense).  Purdue's offense is amongst the all-time elite (5th best amongst tourney teams since '08) and their defense falls just between "championship-level" and 2018's Villanova's team, so it would be unreasonable to discount them as well.  I would suspect that given the quality of offenses left in the field, the defensive liabilities of the non-"championship-level" defenses will be exposed and lead to their elimination.

 

As always @fasbjd thanks for the input! But out of curiosity, where does IU fall in this?

[fasbjd]

37 minutes ago, Drroogh said:

As always @fasbjd thanks for the input! But out of curiosity, where does IU fall in this?

IU was around where Yale was (Sector 4).  They had ~equal aOE but IU had a slightly better aDE - so just below them on the chart

[fasbjd]

1 hour ago, 5fouls said:

I picked every game of the tournament based on this data.  It resulted in me missing out on some of the early round upsets, but I am set up very well from here on out.  People that picked teams like UK to got to the Final Four would have benefitted from looking at this information.    

I am wondering who you picked to win it all.  UConn, Houston, and (..er) Auburn all looked well positioned.  My guts says UConn to repeat but there's an interesting tidbit that I've been looking at.  I call it the "Cradle of Champions" and its the box w/in Sector 1.  9 of the 15 champions shown came from that box.  Both Auburn and Houston this year were in it.  Of course, only Houston remains.  In total, there have been 37 teams that fit in that box.  Excluding this year's Houston, 1/4 of those teams have won the title, 1/3 play in the finals, 44% make the final four, and 2/3 the elite eight.  Factoring that in, it would be hard to bet against Houston or (...er) Auburn.

 

Edited March 28 by fasbjd

[5fouls]

5 minutes ago, fasbjd said:

I am wondering who you picked to win it all.  UConn, Houston, and (..er) Auburn all looked well positioned.  My guts says UConn to repeat but there's an interesting tidbit that I've been looking at.  I call it the "Cradle of Champions" and its the box w/in Sector 1.  9 of the 15 champions shown came from that box.  Both Auburn and Houston this year were in it.  Of course, only Houston remains.  In total, there have been 37 teams that fit in that box.  Excluding this year's Houston, 1/4 of those teams have won the title, 1/3 play in the finals, 44% make the final four, and 2/3 the elite eight.  Factoring that in, it would be hard to bet against Houston or (...er) Auburn.

 

I picked UConn.  The strategy I used when two team were in the same tier was to pick the team closest to the tier line in front of them.  When I got to the Final 4 and everyone was Tier 1, without another tier 'in front', I sided with Offense over Defense.  So, my Final Four matchups were UConn over Arizona and Purdue over Houston, with UConn winning over the Boilers in a close one.