NCAA Bracketology - The Week of February 18th

[orangeBlue]

Originally Posted by AHSIllini32

As much as I love the optimism, let's not label others as being pessimistic for not expecting to win if not both of those tough games on the road. I think we have a legit chance in each one especially OSU, but there's a fine line between optimism and insanity.

I prefer "enlightened skepticism."

Lets not forget how difficult it is to win on the road in the Big 10

Particularly, this time of year. Everyone is battling to improve their chances and/or their seed.

[blmillini]

Originally Posted by AHSIllini32

The BPI is a crock of you-know-what.

That thing just proves that numbers can be used any all kinds of ways that can be valid or invalid. It simply doesn't pass any kind of reality check.

[Dan]

Illinois a 6 seed in Jerry Palm's updated bracket-

http://www.cbssports.com/collegebasketball/bracketology

[SCillini]

Originally Posted by Dan

Illinois a 6 seed in Jerry Palm's updated bracket-

http://www.cbssports.com/collegebasketball/bracketology

What a difference three weeks makes. After the Michigan game I was bemoaning another Illinois NIT appearance. Now a projected six seed! If we get one road win and a BTT win, this should be about where we land.

[MA Illini]

I wasn't trying to be pessimistic or even predict the outcome of the games. Just trying to say if we finish 3-2 I think we will end up with a 6-7 seed and 4-1 will be needed in order to have a chance at a 5 seed. It will be difficult to get to the 5 seed without that 4th win. I think 5-0 will move us to the 4 seed if we also get a couple BTT wins.

I certainly think we can beat anyone, anytime. I think the UM, OSU and @Iowa game will tell us if we have fully turned the corner. If we are looking back at a 8-1 or 9-0 stretch in B1G play that would be huge heading into the NCAA. But we also have to remember that this it he B1G, it doesn't take much of an off night to drop a game. Just ask Indiana. We could be the best team in the country; doesn't mean we will go undefeated

Originally Posted by WizardBill

We just beat the #1 team in the country. We are playing great. We should win at least one if not both of those games. Pessimism is still supreme here, I see.

I expect to win 5 or 6 more BEFORE the BTT.

[IlliniJeffrey]

Originally Posted by FeelYourPaign

I am disappointed to see that Mich does not play during the week this week.

They've been sort of bad (for a top 10 team) lately. I've seen what JG can do with an extra few days of practice and I assume that the days off will benefit scUM, who will be just that much tougher. Oh well, once we take care of PSU, hope we have enough time to be ready for the Blue.

That happened before the Wisconsin game up there also. When guys like Bo Ryan and John Bielien have a week to prepare and we don't, that definitely puts JG at a disadvantage.

[SCillini]

Originally Posted by WizardBill

We just beat the #1 team in the country. We are playing great. We should win at least one if not both of those games. Pessimism is still supreme here, I see.

I expect to win 5 or 6 more BEFORE the BTT.

It'll be hard to win six more before the BTT since we only have five regular season games remaining.

[MA Illini]

Originally Posted by blmillini

That thing just proves that numbers can be used any all kinds of ways that can be valid or invalid. It simply doesn't pass any kind of reality check.

The BPI is interesting. Everyone different rating scale always has teams that is doesn't work well with.

If you look at the individual game BPI numbers, it is pretty clear why IL is ranked pretty low. We have had some absolutely terrible games. The NW loss we had a BPI or 15. We also have an number of wins where our BPI was in the 50-60s by beating bad teams in close games, especially at home.

However, you also see that we have 11 games with a BPI of over 94. But when you are trying to average a BPI of 80-90 it takes a lot of good games to average out a 15 (Florida is No 1 with a season BPI of 91, a BPI of 80 would put us at No 25).

It the difference between trying to get an A in a class or a B. If you want an A and you get a D on one test, it is impossible to make up for. But if you want a B, just get a few As to make up for the one D.

This is an inherent problem with any rating scale that is ranking 347 teams, but when you really only care about the top 50. If you want the dynamic range to rank all 347, then you will lose sensitivity in ranking any subset of those 347. The RPI has the same issue.

[Botb9]

I personally have a problem with any rating scale that is conjured up by ESPN. See: BPI, QBR, and the CY predictor.

[bigtiny]

6 would be great. Projected 3 seeds seem substantially weaker than projected 2 seeds this year.

Duke/Michigan/Florida/Zona > NewMexico/Louisville/Syracuse/Butler


So just jumping from a 7 seed to a 6 seed greatly improves ones chances of going to the second weekend with an upset win.

[MA Illini]

Just an updated on the bracket project matrix discussion from yesterday. IL is up to an aggregated seed average of 7.5 for the 77 brackets.

For the 46 updated yesterday, we are a 7.0. I wonder how is us beating Purdue and NW vs. getting some publicity from the mock selection committee putting us at the 5, which surprised a lot of people but caused more people to look at our resume in detail.


http://bracketproject.50webs.com/matrix.htm

[IlliNYC]

Originally Posted by peace davids

Apparently not. They must be looking at BPI.

39 Kentucky SEC 4 76.5 In - Projected At Large
55 Illinois Big Ten 8 72.0 Out

http://espn.go.com/mens-college-bask...ype/tournament

Obviously, we are way too excited with this low of a BPI

I can't believe ESPN continues to show the BPI. It's such an embarrassment that even their own commenters don't take it seriously.

[blmillini]

Originally Posted by MA Illini

The BPI is interesting. Everyone different rating scale always has teams that is doesn't work well with.

If you look at the individual game BPI numbers, it is pretty clear why IL is ranked pretty low. We have had some absolutely terrible games. The NW loss we had a BPI or 15. We also have an number of wins where our BPI was in the 50-60s by beating bad teams in close games, especially at home.

However, you also see that we have 11 games with a BPI of over 94. But when you are trying to average a BPI of 80-90 it takes a lot of good games to average out a 15 (Florida is No 1 with a season BPI of 91, a BPI of 80 would put us at No 25).

It the difference between trying to get an A in a class or a B. If you want an A and you get a D on one test, it is impossible to make up for. But if you want a B, just get a few As to make up for the one D.

This is an inherent problem with any rating scale that is ranking 347 teams, but when you really only care about the top 50. If you want the dynamic range to rank all 347, then you will lose sensitivity in ranking any subset of those 347. The RPI has the same issue.

And, if it were done correctly that would be a fine argument. It is obvious to me that the SOS component of that index is total garbage.

When you take a class and are attempting to attain any grade you expect the tests that comprise that grade to be relevant to the topic. If the test that caused your grade to slip was inconsistent with what any reasonable professor would have taught in the class, then that professor needs to be fired and needs to change... much like the BPI.

[jcgoldie]

Originally Posted by Dan

Illinois a 6 seed in Jerry Palm's updated bracket-

http://www.cbssports.com/collegebasketball/bracketology

... with potential matchups with 3 seed Syracuse and 2 seed Michigan in rds 2 and 3... I'll take it! Also love that Mizzou is now in the bubble

[scubadunk]

http://espn.go.com/mens-college-basketball/bracketology

10 seed on Lunardi's Bracketology. Playing 7 seed NC State.

[scubadunk]

ESPN keeps talking up their BPI system that is severely flawed. A 11 lost Stanford team and an 11 loss Dayton teams are rated ahead of us. Now they post this article on their blog. ESPN is severely flawed.

http://espn.go.com/blog/collegebaske...-on-the-bubble

[Ransom Stoddard]

Originally Posted by scubadunk

ESPN keeps talking up their BPI system that is severely flawed. A 11 lost Stanford team and an 11 loss Dayton teams are rated ahead of us. Now they post this article on their blog. ESPN is severely flawed.

http://espn.go.com/blog/collegebaske...-on-the-bubble

It's like they are making up new statistics (variance) in an effort to justify their bias against the B1G. Something tells me that if Illinois had more games on ESPN channels than the BTN they would be getting more love from the slapheads in Bristol.

[MA Illini]

Originally Posted by blmillini

And, if it were done correctly that would be a fine argument. It is obvious to me that the SOS component of that index is total garbage.

When you take a class and are attempting to attain any grade you expect the tests that comprise that grade to be relevant to the topic. If the test that caused your grade to slip was inconsistent with what any reasonable professor would have taught in the class, then that professor needs to be fired and needs to change... much like the BPI.

I agree with you that there seems to be a lack of SOS adjustment. It is clearly part of BPI, but perhaps not weighted appropriately.

However, most SOS adjustments suffer the same issue. SOS is a good metric, but when you are using it to adjust a W-L record it has issues. The problem is that it probably works well for the middle 200 of the 347 teams, but not for the extremes (bottom 75 and top 75). For an extreme example, which team do you think is better:

Both teams are 3-0, one teams has played a schedule with an average RPI of 100 (this is Illinois average RPI of openents, ranked the 8th toughest schedule), the other teams plays a schedule with an average RPI of 112 (this woudl ranked a team with about the 35th toughest schedule).

The team with the higher RPI SOS, right? But what if team 1 played the following three games:

against RPI #50, RPI #50, RPI #200

And Team #2 played against:

RPI #1, RPI#50, RPI #285

Which team would you think is better now? The problem is that for teams in the top 50, it doesn't really matter if they play the team ranked #200 or the team ranted #285, they should win 95%+ percent of the time. However, the win percentage estimate difference for playing #1 vs. #50 would be huge for a team that is about the 50th best team. So any top 50 or top 100 team playing schedule #1 would be have a better expected win percentage than schedule 2, but schedule 1 is rated as a tougher schedule. This is part of why the selection committee focused on good wins. Because you only really find out information about top teams when they play other top teams, not bottom dwellers.

The NCAA needs to come up with metrics aimed at tournament teams - i.e. the top 34 at large teams. Diluting these metrics by using the same metrics for all 347 teams creates problems. Other measuring techniques beside RPI try to get additional information from each game with things like margin of victory, but these all have their issues because inside of each game, the goal is to win and that is the only outcome that really matters. This is obviously a complex problem and why it is discussed so often.

[troyk]

Originally Posted by scubadunk

ESPN keeps talking up their BPI system that is severely flawed. A 11 lost Stanford team and an 11 loss Dayton teams are rated ahead of us. Now they post this article on their blog. ESPN is severely flawed.

http://espn.go.com/blog/collegebaske...-on-the-bubble

The language use there is nothing but trying to downplay our efforts. He focuses on the negatives and even skews facts to further downplay us.

"against Hawaii", not "at Hawaii". And he claims we barely beat Auburn at home, when that was a semi-home game.

Regardless, easily a piece looking to garner hits from Illini fans, which is all ESPN wants. Their BPI is utter trash, considering Dayton is ranked ahead of us in it.

[FeelYourPaign]

I hope that we either get all the way to a 6 seed or stay low at 11.
That way, your second game should be against the 3 seed.
They seem weaker than the 2's.

Unless, of course, we become the 3 seed ourselves.
Then 3's are AWESOME !

[douglascountyillinifan]

Originally Posted by Dan

Illinois a 6 seed in Jerry Palm's updated bracket-

http://www.cbssports.com/collegebasketball/bracketology

Wow, the WEST in this projection has to be one of the weakest brackets ever.

[Illwinsagain]

Originally Posted by scubadunk

http://espn.go.com/mens-college-basketball/bracketology

10 seed on Lunardi's Bracketology. Playing 7 seed NC State.

Well, we are seeded the same as Wichita St., La Salle and UNC. Most years I would be happy being seeded the same as UNC, but not this year.

[VaIllini14]

Hopefully we can get into the east region and get to the sweet 16 so I can go to dc and watch them get into the final four!

[AEX]

Originally Posted by Ransom Stoddard

It's like they are making up new statistics (variance) in an effort to justify their bias against the B1G. Something tells me that if Illinois had more games on ESPN channels than the BTN they would be getting more love from the slapheads in Bristol.

Yeah, that article is a joke. Here's the argument for why we should be on the bubble.

But let’s not forget about four losses by at least 14 points, one of which was at home to Northwestern. And three of the Illini's wins aren’t so impressive. They won by two or fewer points against Hawaii (in overtime), Gardner-Webb (at home) and Auburn (at home).

If the argument is based mainly on close wins, it's really not an argument at all.

[Illwinsagain]

Originally Posted by scubadunk

ESPN keeps talking up their BPI system that is severely flawed. A 11 lost Stanford team and an 11 loss Dayton teams are rated ahead of us. Now they post this article on their blog. ESPN is severely flawed.

http://espn.go.com/blog/collegebaske...-on-the-bubble

I think I am now dumber for having read those explanations. They credit UVA for playing better against good opponents and explain away losses to bad teams due to an injury. However, UI is criticized for losing (oops) I mean winning 2 close games against lower teams while suffering jetlag. Also playing their best games against better competition. So, close wins are now a reason for being on the bubble. :rolleyes: