Historical sweet 16 tournament appearances

[JMorgan99]

I posted this on my blog, and thought some of you might find it interesting.




This chart shows the number of times each team has made the sweet 16 since the tournament expanded the field to 64 teams in 1985.

All games from 1985 through this year are included. The number of appearances is included in parentheses behind each team’s name (due to space limitations, teams with 3 or less wins do not include this label).

Each team is shaded based on its winning percentage in the sweet 16 (winning percentage is through the 2009 tournament). The darker the blue, the higher the winning percentage. Teams are sorted based on their conference membership today, not based on their membership at the time of sweet 16 appearances.

Top ten teams (teams in bold will play on Thu/Fri)

1. Duke (19 appearances, 11 wins)
2. North Carolina (17 appearances, 12 wins)
3. Kansas (16 appearances, 8 wins)
4. Kentucky (15 appearances, 9 wins)
5. Arizona (12 appearances, 7 wins)
6. UConn (12 appearances, 7 wins)
7. Michigan State (11 appearances, 6 wins)
8. UCLA (11 appearances, 6 wins)
9. Syracuse (11 appearances, 4 wins)
10. Louisville (10 appearances, 5 wins)

Needless to say, it’s a strange year for the tournament. 6 of the top 10 programs won’t appear in the sweet sixteen.

[El_Diablo]

If Duke has 11 wins in 18 previous appearances, isn't the winning percentage over 50%? It appears to have the 25-50% shade. Same with UNC.

Otherwise...interesting product. Some conferences have been pretty balanced over time, but some have just been dominated by a couple teams.

[hc5duke]

If Duke has 11 wins in 18 previous appearances, isn't the winning percentage over 50%? It appears to have the 25-50% shade. Same with UNC.

Duke = 19 appearances, 11 wins, (19 minus 3) losses = 41%
UNC = 17 appearances, 12 wins, (17 minus 3) losses = 46%

minus 3 meaning each team has 3 championships since 1985, so Duke has lost (been eliminated) 16 times in its 19 Sweet Sixteen appearances, and 14 for UNC

[Eternal Outlaw]

Duke = 19 appearances, 11 wins, (19 minus 3) losses = 41%
UNC = 17 appearances, 12 wins, (17 minus 3) losses = 46%

minus 3 meaning each team has 3 championships since 1985, so Duke has lost (been eliminated) 16 times in its 19 Sweet Sixteen appearances, and 14 for UNC

Huh? Duke is 11-7 in this round in that time frame for 61%. What does Championships even have to do with the sweet 16 round?

Why are you dividing 11 by 27? What does the 27 represent? If you are trying to say that from the sweet 16 and on that Duke has a 46% win percentage that is wrong also because the 11 is wins in the sweet 16 round only and does not include the elite 8 wins, semi-final wins, and Championship wins. I have Duke at 31-15 in that time span if you take their win-loss records from the rounds of the sweet 16, elite 8, semi-finals, and championship game which is 67%.

[hc5duke]

Huh? Duke is 11-7 in this round in that time frame for 61%. What does Championships even have to do with the sweet 16 round?

Why are you dividing 11 by 27? What does the 27 represent? If you are trying to say that from the sweet 16 and on that Duke has a 46% win percentage that is wrong also because the 11 is wins in the sweet 16 round only and does not include the elite 8 wins, semi-final wins, and Championship wins. I have Duke at 31-15 in that time span if you take their win-loss records from the rounds of the sweet 16, elite 8, semi-finals, and championship game which is 67%.

I was under the impression that the winning percentages are based on games after round of 32, but the way it's written I guess it should only count the games in the round of sweet sixteen. In which case whoever created the chart did indeed color the bars wrong

[CDu]

I was under the impression that the winning percentages are based on games after round of 32, but the way it's written I guess it should only count the games in the round of sweet sixteen. In which case whoever created the chart did indeed color the bars wrong

If you are going to count all losses past the round of 32, then you'd have to also count all wins after the round of 32. So we'd have a lot more than 11 wins in that scenario. Our record beyond the round of 32 since 1985 is 31-15, which is better than 60%. Of course, I don't think that was the intended calculation.

Basically, either way it has to be a mistake. We're either 31-15 (assuming all games after the round of 32, which I don't think was the intended calculation) or 11-7 (assuming only Sweet-16 games, which I do think was the intended calculation).

By the same token, it appears that UNC, Arizona, MSU, UConn, and UCLA are shaded incorrectly as well.

EDIT: oops, I didn't fully read Eternal_Outlaw's post, which says the 31-15 part. Sorry for the double post (though at least the last sentence is still relevant).

[JMorgan99]

Hey guys, thanks for the catch and sorry for the confusion. The shading is definitely incorrect. I mistakenly uploaded an older version of this chart. It should be updated now so it is correct now. Is it possible to edit the initial post so that I can change the image link?

Here's the updated version:


The way that winning percentage is calculated, a team like Duke would have 11/18 wins in the sweet 16.

Hope that helps!

JM

[Chitowndevil]

Needless to say, it’s a strange year for the tournament. 6 of the top 10 programs won’t appear in the sweet sixteen.

Nice post. This bugged me, though. Let's do a back of the envelope calculation. In 25 tournaments there are 400 total slots in the Sweet 16. The top 10 list of teams you gave have accounted for 134 appearances, or 33.5%.

If everything is iid (yeah right), those top 10 teams would account for 16*.335 = 5.36 of the 16 slots each year, with a standard deviation of (16*.335*(1-.335))^(1/2) = 1.89. In other words, 4 out of the 10 teams being represented is not that unusual.

[JMorgan99]

Nice post. This bugged me, though. Let's do a back of the envelope calculation. In 25 tournaments there are 400 total slots in the Sweet 16. The top 10 list of teams you gave have accounted for 134 appearances, or 33.5%.

If everything is iid (yeah right), those top 10 teams would account for 16*.335 = 5.36 of the 16 slots each year, with a standard deviation of (16*.335*(1-.335))^(1/2) = 1.89. In other words, 4 out of the 10 teams being represented is not that unusual.

Awesome, thanks for that. In retrospect that should have been more obvious to me but I definitely missed it and probably would have shied away from the statistical calculation. I'll add your comment to the post on my blog.

[JMorgan99]

Although, in fairness, 4 of these 10 teams didn't even make the tournament. That is unusual, however that wasn't the point I was making in my post.

[Welcome2DaSlopes]

Hey guys, thanks for the catch and sorry for the confusion. The shading is definitely incorrect. I mistakenly uploaded an older version of this chart. It should be updated now so it is correct now. Is it possible to edit the initial post so that I can change the image link?

Here's the updated version:


The way that winning percentage is calculated, a team like Duke would have 11/18 wins in the sweet 16.

Hope that helps!

JM

Hey it's cool man, thanks for the chart, pretty good information there,