In advance of the AP and Coach's rankings coming out later today, I thought I would get things started with a look at the latest other rankings out there.
Sagarin
Duke is currently #3 in the most widely known computer ranking behind Kansas and Syracuse. Texas, Villanova, and undefeated Kentucky make up the next few teams.
Worth noting that Duke is #1 in the Sagarin predictor, which he says is a better way of projecting future results. Gotta like that!
Also, oft-discussed Kentucky has remained undefeated against a tremendously weak schedule. They have not played a single team ranked in Sagarin's top 25 and their strenth of schedule is the lowest of any team in the top 20. The woeful SEC (aside from Tenn and maybe Miss or Vandy) won't present many challenges either.
The ACC is the #1 conference (as it almost always is), but Duke is the only really highly ranked team. The ACC rankings are:
Duke - 3
Wake - 16
Ga Tech - 21
Clemson - 25
FSU - 28
Maryland - 33
Va Tech - 37
UNC - 48
Miami - 56
NCSU - 74
Virginia - 75
BC - 86
It is interesting to see how these rankings greatly differ from Ken Pomoeroy's.
KenPom
Duke is #2 in the Pomeroy rankings this morning, trailing only Kansas. Syracuse, BYU, and Texas round out the top 5. Duke has the #1 offense in the land and the #14 defense. I am not sure if the computer watched the NC State game
Kentucky, preparing to assume the #1 ranking in the human polls, again gets no love from the computers. They are #9. Interestingly, Pomeroy has some category called "luck" in which Kentucky is the #21 team in the land. Luck is not something you control (Maine, Bethune Cookman, and Western Carolina are the luckiest teams in the land according to Ken). I am not sure how Ken measures it, but this would seem to imply that Kentucky is having a very lucky season. Duke's luck rates out as #262, a very low number.
Ken ranks the ACC thusly--
Duke - 2
Maryland - 11
Clemson - 18
Ga Tech - 23
FSU - 24
Wake - 27
Va Tech - 37
Miami - 49
Virginia - 54
UNC - 55
NCSU - 65
BC - 69
You all know how much I love Pomeroy's predictions. I am slammed at the moment, but will have more on them in a bit...
--Jason "to be continued" Evans
Why are you wasting time here when you could be wasting it by listening to the latest episode of the DBR Podcast?
From kenpom.com:
"The new ones are Cons (Consistency) and Luck. The easiest one to understand is Luck, which is the deviation in winning percentage between a team’s actual record and their expected record using the correlated gaussian method. The luck factor has nothing to do with the rating calculation, but a team that is very lucky (positive numbers) will tend to be rated lower by my system than their record would suggest."
Coach's ranking is now out, and Duke checks in at #7.
A little surprised to see the treatment of Kansas St, but it suggests that recency is a big factor to voters, even within weeks (i.e., the Saturday loss to OK St had a bigger impact than the Monday win over Texas).
Just be you. You is enough. - K, 4/5/10, 0:13.8 to play, 60-59 Duke.
You're all jealous hypocrites. - Titus on Laettner
You see those guys? Animals. They're animals. - SIU Coach Chris Lowery, on Duke
Harvard got more votes than the heels in the AP poll. Tommy and Co. got 1 vote.
Q "Why do you like Duke, you didn't even go there." A "Because my art school didn't have a basketball team."
Just be you. You is enough. - K, 4/5/10, 0:13.8 to play, 60-59 Duke.
You're all jealous hypocrites. - Titus on Laettner
You see those guys? Animals. They're animals. - SIU Coach Chris Lowery, on Duke
In addition to Pomeroy and Sagarin, the other rating model I've grown to love, is LRMC (Logistic Regression/Markov Chain model), created by some research at Georgia Tech.
Read about it here:
http://www2.isye.gatech.edu/people/f...bout_lrmc.html
See the "Pure LRMC Rankings":
http://www2.isye.gatech.edu/people/f...rmc.sort0.html
The Pomeroy, Sagarin, and LRMC are 3 good models, so much superior to the mathematically flawed RPI, and the emotionally flawed AP and Coaches polls.
These models saw the weakness of Carolina early on, and never had them in the top 25, ever, this season. And it sees the weakness of Kentucky. No way is Kentucky the top team.
And if NCSU wins Tuesday, is it actually worth storming the court?
One thing that you can see from the computer rankings is that the ACC is so very NOT BAD. It doesn't have the strength at the top of some conferences, but there are no real breathers. Every single team in the ACC is in the top 100 in Sagaran, Pomeroy and LRMC.
Bracketology has unc-ch as a 9 seed right now. How many more losses before unc is headed to the NIT? If they finish 7-9 and get bounced in the ACC semis, what possible argument do they have to be in the dance?
I think I am maxing out on schadenfreude right now!!
I haven’t been able to watch as many ACC games this year as in the past but the Sagarin, Pomeroy and LRMC ratings confirm what my eyes have seen. The Devils are clearly the best team in the league and UNC is struggling. In my view, the ratings also confirm the top six teams in the league although no one has mentioned that Maryland is highly rated in both Pomeroy and LRMC ratings (second to Duke among ACC teams). I don’t expect it to last but the Terps have recently played better and more consistent basketball than I can recall seeing them play for a few years. They are a good passing team, share the ball, put forth good efforts on defense, the boards and going after loose balls, and, for now, have figured out how to put the ball in the basket. Good team play will only take you so far but they have beaten Florida State and came within an open 12-foot shot of beating Wake. Going against the superior talent of Duke and Georgia Tech will probably be too much.
gw67
Exactly. The ACC might have fewer teams at the top, but going through a conference season with no breathers (even BC is well in top 100 in both sagarin and kenpom) is really difficult. Contrast that with the Big East and Big 10, which have multiple teams over 100 (Rutgers, Depaul, Providence, Iowa, Indiana, Penn St). Clearly the Big East and Big 10 are tough conferences, but it may help to play a cupcake every now and then, if only to rest and heal up.
I think it has finally sunk into my thick skull that the NCAA field is unusually devoid of frightening teams; who is going to get hot in March? Why not Duke?
I find it interesting that the computer has the ACC with 9 teams as tournament worthy - top 65 - and KenPom has TEN (!) worthy teams. No way we get that many in, especially as the ACC season advances and we get to beating each other up more. Interesting situation for UNC, with another loss or two they could be outside looking in -- though bias/tradition unfortunately plays a big role come Selection Sunday... Miami, VA, NCSU and BC all looking pretty dubious right now. But, heck, I think the ACC would be thrilled with 7-8 teams getting in. Pretty funny that those years that produce no one (or two) great teams, are the one's that "help" the conference in terms of getting the greater number of teams in (albeit with middling seed numbers). Hmmmm.
You are mistaken about what consititutes "tournament-worthy" -- don't feel bad, every March dozens of coaches and commentators make the same mistake.
Let's get it straight -- the 65-team NCAA is NOT designed to include the top 65 teams. There's nothing like that in the rules and guidelines for the selection committee.
The NCAA field is comprised of two parts:
(1) 31 automatic conference qualifiers
(2) 34 at large teams
Now, that first category has nothing to do with a team's national rank. All you have to do is win your conference's automatic bid. In 30 of the 31 conferences, that means winning your league's postseason tourney. The Ivy awards its bid to the regular season champion.
Many of those teams WILL be among the nation's top 65 teams ... but quite a few will not be. For instance, whichever team comes out of the Ivy League (probably Cornell or Harvard) will not be a top 65 team. Neither will the champions of the Patriot League or the OVC or the SWAC.
The second category is supposed to be for the 34 strongest teams that don't have an automatic bid. I say supposedly, because I believe these choices have been corrupted by pressure from smaller conferences to grab bids that they don't deserve. With 6-8 members of the 10-team selection committee usually from the smaller leagues, they usually manage to siphon off 2-4 at-large bids for lesser teams from weaker conferences. But, officially, NCAA guidelines demand that these 34 bids go to the best teams.
When you put the best 34 at large and the top teams that also win conference titles together, you usually get 41-45 of the top teams in the country (depending as always on how many conference tournament upsets there are and how many at large bids the smaller conferences can steal in the committee).
So to look at the ACC rankings and say that nine or 10 ACC teams "are in the top 65" is meaningless. In the real world, the current Sagarin and Pomeroy rankings both would put seven teams in the field -- 48 UNC (in Sagarin) and 49 Miami (in Pomeroy) would seem to be outside the bubble.
Of course, the most important of the computer rankings is the RPI. You guys might raise objections to it ... but the committee has paid a lot more attention to it over the years. It's not as good a ranking than Sagarin or Pomeroy, but it's a better predictor of NCAA selection. As of the moment, the ACC's RPI rankings are:
3 Duke
21 Wake
29 Georgia Tech
39 FSU
43 Clemson
54 Maryland
70 VPI
73 UNC
84 NC State
90 Virginia
95 Boston College
99 Miami
By that mesasure, the ACC only has four or five sure bets (Clemson is right at the usual cutoff point).
Of course, the computer rankings are only a small part of it. Plus, we have six weeks to go before selection Sunday and a lot can chance by then.
Last edited by Olympic Fan; 01-25-2010 at 04:45 PM.
Thanks for the links to LRMC, but careful how you regard a team like Kentucky. They are almost certainly underrated by the computer rankings because they're a young team with high upside. Therefore, while they may have struggled early in the season (while still remaining unbeaten), they have improved and will continue to improve throughout the season. For example, in Pomeroy, I know Kentucky spent a good deal of the early part of the season ranked between 40 and 70 and as of today they're ranked 9th and will probably continue to climb. What would be really useful would be if Pomeroy, Sagarin, and LRMC showed separate rankings based on the last 10 games or perhaps weighed recent games more heavily in their ratings. Kentucky would probably rank better in those. (As an aside, Duke is probably a bit overrated in the computer rankings due to the experience of our players).