Bracketology Is Not a Science
By Zach
So today saw the release of Joe Lunardi's latest "Bracketology." I always found this an exercise in lunacy. First, Lundardi boasts that most years, he only misses one team on his picks: what he means is, that of the 65 teams he predicts, 64 of them usually make the tournament. Sounds impressive, until you realize that 34 of those teams were automatic entries, so the real number is 30 of 31. Still sounds good, right? Well, not really. Most national analysts do just as well, because so much of the chricteria is made public. Still, ESPN trumpets this guy as some sort of magical sage who studies tea leaves or sticks and divines if your school will be a four-seed or a five-seed. Never mind that his predictions as far as that goes are much more frequently wrong. He's on ESPN.com, so he must know what he's talking about.
Now, he's paid to do this for a living, so I can't say for sure that he's wrong. But I find it very hard to believe a few teams are in, or seeded where he says.
Team A:
15-10, 44 RPI, 20 SOS, 5-7 vs. RPI top 50, 2 losses to teams outside RPI top 100.
Lundardi says: 10 seed.
I say: Umm, NIT. RPI is solid, SOS is good, but record isn't, and neither are those two bad losses. They've still got a chance to play themselves in, and I know the field is fairly weak at the bottom this year, but I just don't see it right now.
Team B:
22-6, 15 RPI, 27 SOS, 4-4 vs. RPI top 50, 1 loss to team outside RPI top 100.
Lunardi says: 4 seed.
I say: It's not their seeding I disagree with, it's their seeding relative to Team C.
Team C:
22-5, 30 RPI, 78 SOS, 4-1 vs. RPI top 50, 2 losses to teams outside RPI top 100, two wins over Team B.
Lunardi says: 7 seed.
I say: Joe Lunardi is a moron. Most people have them as a four or five seed, and considering they're on a six-game winning streak and are one game back of the conference lead, they're a six at absolute worst.
Team D:
21-5, 7 RPI, 43 SOS, 5-4 vs. RPI top 50, 1 loss to team outside RPI top 100.
Lunardi says: 2 seed
I say: 3 or 4. They're still benefiting from a hot start to the season, and a high ranking in the polls, plus playing in a major conference. But when you look at their resume, it's kind of thin for that high of a seed. They don't have a win over either of the top two teams in the conference, and they split with the third-best. A couple of earlier wins look far less impressive now that those teams have collapsed. To reach that high of a seed, methinks they'll need a good run in their conference tournament (which could happen, of course). But as of now, a two-seed is a reach.
Team E:
25-3, 9 RPI, 87 SOS, 2-3 vs. RPI top 50, no losses to teams outside RPI top 50.
Lunardi says: 2 seed
I say: Prove it. Their resume is built on two big early season wins, neither of which is as impressive now (wins over #17, 50 in latest RPI). Lost close games to #3, #6, and #30. Beyond that, their schedule is light: 18 games against teams outside the RPI top 100 (a product of their conference). The gaudy record and ranking may influence selection, but it's been a long time since this team faced a real test, which scares me when it comes to high seeds.
I have no problem with trying to predict which 65 teams will make the tourament. I don't even care if you try and seed them. I do have problem with treating the writings of one schmoe as gospel. Too much can happen between now and Selection Sunday for this latest Bracketology to be more than a very faint outline.
By the way, for those interested, the teams above: Indiana, UCLA, Washington, Pittsburgh, Gonzaga.
Now, he's paid to do this for a living, so I can't say for sure that he's wrong. But I find it very hard to believe a few teams are in, or seeded where he says.
Team A:
15-10, 44 RPI, 20 SOS, 5-7 vs. RPI top 50, 2 losses to teams outside RPI top 100.
Lundardi says: 10 seed.
I say: Umm, NIT. RPI is solid, SOS is good, but record isn't, and neither are those two bad losses. They've still got a chance to play themselves in, and I know the field is fairly weak at the bottom this year, but I just don't see it right now.
Team B:
22-6, 15 RPI, 27 SOS, 4-4 vs. RPI top 50, 1 loss to team outside RPI top 100.
Lunardi says: 4 seed.
I say: It's not their seeding I disagree with, it's their seeding relative to Team C.
Team C:
22-5, 30 RPI, 78 SOS, 4-1 vs. RPI top 50, 2 losses to teams outside RPI top 100, two wins over Team B.
Lunardi says: 7 seed.
I say: Joe Lunardi is a moron. Most people have them as a four or five seed, and considering they're on a six-game winning streak and are one game back of the conference lead, they're a six at absolute worst.
Team D:
21-5, 7 RPI, 43 SOS, 5-4 vs. RPI top 50, 1 loss to team outside RPI top 100.
Lunardi says: 2 seed
I say: 3 or 4. They're still benefiting from a hot start to the season, and a high ranking in the polls, plus playing in a major conference. But when you look at their resume, it's kind of thin for that high of a seed. They don't have a win over either of the top two teams in the conference, and they split with the third-best. A couple of earlier wins look far less impressive now that those teams have collapsed. To reach that high of a seed, methinks they'll need a good run in their conference tournament (which could happen, of course). But as of now, a two-seed is a reach.
Team E:
25-3, 9 RPI, 87 SOS, 2-3 vs. RPI top 50, no losses to teams outside RPI top 50.
Lunardi says: 2 seed
I say: Prove it. Their resume is built on two big early season wins, neither of which is as impressive now (wins over #17, 50 in latest RPI). Lost close games to #3, #6, and #30. Beyond that, their schedule is light: 18 games against teams outside the RPI top 100 (a product of their conference). The gaudy record and ranking may influence selection, but it's been a long time since this team faced a real test, which scares me when it comes to high seeds.
I have no problem with trying to predict which 65 teams will make the tourament. I don't even care if you try and seed them. I do have problem with treating the writings of one schmoe as gospel. Too much can happen between now and Selection Sunday for this latest Bracketology to be more than a very faint outline.
By the way, for those interested, the teams above: Indiana, UCLA, Washington, Pittsburgh, Gonzaga.
posted by Zach at 10:40 AM
1 Comments:
Lunardi's predictions are even more meaningless considering he changes his bracket weekly. He starts doing this in January, when teams typically have played only two or three conference games.
readydownhuthuthut.blogspot.com
Post a Comment
<< Home