KenPom: Seeding doesn't matter

[skibum]

Who’s No. 1? Who Cares.

Read the full article if you're interested, but the essentials are this:
1. Seeding doesn't matter for the top teams (think 1-3 seeds in each region)
2. Overall strength of region does matter

Or, in other words: it's better to be a 3 seed in a weak region than a 1 seed in a tough region, aka "Gonzaga: don't get pissed off if you don't get a 1 seed, you're actually better off with a 2 or 3 in the west than you are with a 1 in a different region."

He briefly mentioned that seeding probably does play a part in Cinderella teams, but the overall point was that if you're a team that's in the discussion for a 1 seed, it doesn't really matter if you actually get it or not...

 

[jgisland]

Obviously getting a three seed in a weak division could be better than a 1 seed in a good division but history doesn't exactly add up to what KenPom is saying. You definitely don't want to be a 8 or 9 seed.

This chart shows how each seed has fared over the six rounds of each tournament since 1985:

Seed	1st	2nd	Sweet 16	Elite Eight	Final Four	Finals
1	100	87	70.7	41.3	22.8	14.1
2	95.7	63	46.7	22.8	10.9	4.3
3	83.7	47.8	22.8	13	8.7	3.3
4	80.4	43.5	15.2	9.8	2.2	1.1
5	69.6	37	5.4	4.3	2.2	0
6	68.5	38	13	3.3	2.2	1.1
7	64.1	19.6	6.5	0	0	0
8	45.7	9.8	6.5	3.3	1.1	1.1
9	54.3	3.3	1.1	0	0	0
10	35.9	17.4	6.5	0	0	0
11	31.5	12	4.3	2.2	0	0
12	31.5	15.2	1.1	0	0	0
13	19.6	4.3	0	0	0	0
14	16.3	2.2	0	0	0	0
15	4.3	0	0	0	0	0
16	0	0	0	0	0	0

 

[Medford M.]

Obviously getting a three seed in a weak division could be better than a 1 seed in a good division but history doesn't exactly add up to what KenPom is saying. You definitely don't want to be a 8 or 9 seed.

This chart shows how each seed has fared over the six rounds of each tournament since 1985:

Seed	1st	2nd	Sweet 16	Elite Eight	Final Four	Finals
1	100	87	70.7	41.3	22.8	14.1
2	95.7	63	46.7	22.8	10.9	4.3
3	83.7	47.8	22.8	13	8.7	3.3
4	80.4	43.5	15.2	9.8	2.2	1.1
5	69.6	37	5.4	4.3	2.2	0
6	68.5	38	13	3.3	2.2	1.1
7	64.1	19.6	6.5	0	0	0
8	45.7	9.8	6.5	3.3	1.1	1.1
9	54.3	3.3	1.1	0	0	0
10	35.9	17.4	6.5	0	0	0
11	31.5	12	4.3	2.2	0	0
12	31.5	15.2	1.1	0	0	0
13	19.6	4.3	0	0	0	0
14	16.3	2.2	0	0	0	0
15	4.3	0	0	0	0	0
16	0	0	0	0	0	0

History can almost be set aside at the point though because parity is at an all time high. A 16 over a 1 will happen in the next 10 years. I would be curious what that chart would look like over the last 10 seasons.

 

[tante]

The 8 seed is such a strange anomaly.

 

[skibum]

Obviously getting a three seed in a weak division could be better than a 1 seed in a good division but history doesn't exactly add up to what KenPom is saying. You definitely don't want to be a 8 or 9 seed.

My immediate thought was similar to yours when I read the headline - a 15 or 16 seed obviously sucks. Then I read the article, and realized that what he was really arguing is that if you're an elite team (i.e. you're not getting an 8 or 9 seed), then ultimately there isn't much, if any, difference between getting the 1, 2 or 3 seed, and that the overall strength of the region you're placed in is more important than what seed you get.

 

[Nor Cal Buff]

Obviously getting a three seed in a weak division could be better than a 1 seed in a good division but history doesn't exactly add up to what KenPom is saying. You definitely don't want to be a 8 or 9 seed.

This chart shows how each seed has fared over the six rounds of each tournament since 1985:

Seed	1st	2nd	Sweet 16	Elite Eight	Final Four	Finals
1	100	87	70.7	41.3	22.8	14.1
2	95.7	63	46.7	22.8	10.9	4.3
3	83.7	47.8	22.8	13	8.7	3.3
4	80.4	43.5	15.2	9.8	2.2	1.1
5	69.6	37	5.4	4.3	2.2	0
6	68.5	38	13	3.3	2.2	1.1
7	64.1	19.6	6.5	0	0	0
8	45.7	9.8	6.5	3.3	1.1	1.1
9	54.3	3.3	1.1	0	0	0
10	35.9	17.4	6.5	0	0	0
11	31.5	12	4.3	2.2	0	0
12	31.5	15.2	1.1	0	0	0
13	19.6	4.3	0	0	0	0
14	16.3	2.2	0	0	0	0
15	4.3	0	0	0	0	0
16	0	0	0	0	0	0

You're not looking at it from the right perspective. 1 seeds on average are going to be better than than 2 seeds who are going to be better on average than 3 seeds, etc. That is not in question. What KenPom is addressing is what the impact on an individual team being mis-seeded is. That means what is the impact on a team with 1 seed talent being placed as a 2 or 3 seed and the ultimate analysis is that it doesn't make much of an impact at all.

 

[jgisland]

You're not looking at it from the right perspective. 1 seeds on average are going to be better than than 2 seeds who are going to be better on average than 3 seeds, etc. That is not in question. What KenPom is addressing is what the impact on an individual team being mis-seeded is. That means what is the impact on a team with 1 seed talent being placed as a 2 or 3 seed and the ultimate analysis is that it doesn't make much of an impact at all.

I get what KenPom is saying and agree with everything you said. My point is that seeding may not have much effect on the top couple of seeds it does play a part farther down. CU is a prime example this year, they could be anything from an 8 to a 12. Looking historically I would rather be a 10 than a 12 ths year.

 

[Buffnik]

The thing that stands out to me is how good of a job the Committee does with the seeding. For all I complain, they do get it very close to correct. Some of the discrepancy on the 11- and 12- lines are completely due to the best mid-major rarely getting a better seed than that in years past.