Pregame: Illinois vs Michigan, Sunday, January 27th, 5:00pm CT, BTN

[Foggy Notion]

Originally Posted by zpfled

Our resume is really pretty good. Our trajectory is what's concerning. Nebraska helped, and tomorrow could really help. If we can get back to playing well, we'll be in fantastic shape. Our little skid is no big deal if it's just a 3-game skid. If it's the norm for B1G season, then we're screwed.

Hey, we're on a one-game winning streak. Trajectory is on the upswing!

[trevdv]

I like how people are saying that next 5 games is the "gauntlet." Really, in my opinion, the next 7 games are. After that road game at Minnesota, we have Purdue (with whom we've lost to 8 straight times) and Northwestern on the road (who just beat Minnesota too). If you ask me, those games are real challenges too.

[WesterveltVictoryCigar]

Originally Posted by trevdv

I like how people are saying that next 5 games is the "gauntlet." Really, in my opinion, the next 7 games are. After that road game at Minnesota, we have Purdue (with whom we've lost to 8 straight times) and Northwestern on the road (who just beat Minnesota too). If you ask me, those games are real challenges too.

Well, the reality is we only have 2 "comfortable" games left, home games vs PSU and Nebraska. The rest are all relatively daunting in some respect.

[danielb927]

Originally Posted by zpfled

But if we win tomorrow, those odds go up, since they're updated after each game, right?

Yep. Our chances in the other games would go up since our ratings would get a boost. Without that, here's how it breaks down. About a 3x boost to the chance of getting 3+ wins, but still a 1-in-5 longshot. 2 wins or more, though, becomes a 2 to 1 favorite.

Win tomorrow, 5-game stretch probabilities become:

1 win: 34.9%
2 wins: 43.6%
3 wins: 18.2%
4 wins: 3.1%
5 wins: 0.2%

[trevdv]

Originally Posted by danielb927

Yep. Our chances in the other games would go up since our ratings would get a boost. Without that, here's how it breaks down. About a 3x boost to the chance of getting 3+ wins, but still a 1-in-5 longshot. 2 wins or more, though, becomes a 2 to 1 favorite.

Win tomorrow, 5-game stretch probabilities become:

1 win: 34.9%
2 wins: 43.6%
3 wins: 18.2%
4 wins: 3.1%
5 wins: 0.2%

Wow...beat the #2 team in the nation, and we don't even have a 1 in 4 chance of getting three wins?

[danielb927]

Originally Posted by trevdv

Wow...beat the #2 team in the nation, and we don't even have a 1 in 4 chance of getting three wins?

Right now we're not much better than a 1 in 4 chance of getting two wins, if you believe those numbers! So that'd be a pretty nice bump.

I'd guess the rating boost we'd get would put it right about at 25%, maybe a tad better. It really is a gauntlet, our estimated chances of winning the next 5 games individually are 23% (UM), 19% (@MSU), 38% (UW), 21% (IU), 12% (@Minn). Obviously that's all based on one system's rankings, but it's a pretty good indicator. Tough to pull 2/4 when you're a strong underdog in almost every game.

Hopefully we can take care of the tough part tomorrow first

[FeelYourPaign]

Originally Posted by danielb927

Yep. Our chances in the other games would go up since our ratings would get a boost. Without that, here's how it breaks down. About a 3x boost to the chance of getting 3+ wins, but still a 1-in-5 longshot. 2 wins or more, though, becomes a 2 to 1 favorite.

Win tomorrow, 5-game stretch probabilities become:

1 win: 34.9%
2 wins: 43.6%
3 wins: 18.2%
4 wins: 3.1%
5 wins: 0.2%

These are cumulative, I gather. Win exactly 1 game with a 35% probability, meaning that there is a 35% probability of losing all of the remaining 4 games.
Win 2 games means that 44% of the time, after winning game 1 we would only 1 of the remaining 4.
And so on and so forth.

Assume that was done right. Looks pretty reasonable.

[danielb927]

Originally Posted by FeelYourPaign

These are cumulative, I gather. Win exactly 1 game with a 35% probability, meaning that there is a 35% probability of losing all of the remaining 4 games.
Win 2 games means that 44% of the time, after winning game 1 we would only 1 of the remaining 4.
And so on and so forth.

Assume that was done right. Looks pretty reasonable.

Yep, that's what I was going for with those - total wins over the 5 game stretch.

[Illini70math]

I don't think some of you are interpreting these numbers correctly if I understand what you are saying.

These numbers are all BEFORE any of the five games are played. They are the most likely distribution of outcomes based on the current likelihood of winning each of the games (according to the model). Everything changes if we win the first game, and not just because the ratings will shift slightly. You would then have to go to a new model based on the remaining four games because the first game is no longer part of the probability computation.

It's a little like the old joke about taking a bomb on board a plane with you because the odds of two bombs on a plane are incredibly low. Once an event has taken place, you should no longer include it in figuring the remaining probabilities.

[trevdv]

My Gonzaga and Ohio State pre-game videos were pretty popular, so I thought I'd make one for Michigan too. Enjoy!

http://www.youtube.com/watch?v=WEmgxis7Jak

EDIT: Dan can you make that YouTube link viewable on the forums? I can't seem to get it to work...

[jrichisamazing]

I like the fact that so many ranked teams have gone down on the road today and this past week. Lets keep that trend going tomorrow night!

[danielb927]

Originally Posted by Illini70math

I don't think some of you are interpreting these numbers correctly if I understand what you are saying.

These numbers are all BEFORE any of the five games are played. They are the most likely distribution of outcomes based on the current likelihood of winning each of the games (according to the model). Everything changes if we win the first game, and not just because the ratings will shift slightly. You would then have to go to a new model based on the remaining four games because the first game is no longer part of the probability computation.

It's a little like the old joke about taking a bomb on board a plane with you because the odds of two bombs on a plane are incredibly low. Once an event has taken place, you should no longer include it in figuring the remaining probabilities.

Sure, you'd have to update the probabilities for each game after the UM game because the ratings will change. But those changes won't cause more than a 3-5% swing in the chance we win any individual games, so pretending we beat UM and then figuring the chance of 0-4 more wins in this stretch isn't a bad approximation. Both of the sets of numbers are calculated from the probabilities of each individual, so I did "go to a new model" in pretending the UM game was a win by just using the probabilities for the 4 individual games left.

[zpfled]

Originally Posted by danielb927

Yep. Our chances in the other games would go up since our ratings would get a boost. Without that, here's how it breaks down. About a 3x boost to the chance of getting 3+ wins, but still a 1-in-5 longshot. 2 wins or more, though, becomes a 2 to 1 favorite.

Win tomorrow, 5-game stretch probabilities become:

1 win: 34.9%
2 wins: 43.6%
3 wins: 18.2%
4 wins: 3.1%
5 wins: 0.2%

Shouldn't that be 100%?

[danielb927]

Originally Posted by zpfled

Shouldn't that be 100%?

No, that's the chance that we'd lose the next 4 after beating UM. It's exactly 1 win, not at least 1.

By comparison, if we lose tomorrow, all of those probabilities just shift down, e.g. we'd have an 18.2% chance of getting two wins in the 5-game stretch.

[Illini70math]

Danielb927 wrote:

Win tomorrow, 5-game stretch probabilities become:

1 win: 34.9%
2 wins: 43.6%
3 wins: 18.2%
4 wins: 3.1%
5 wins: 0.2%
----------------------------------------------------

zpfled asked: Shouldn't that be 100%?
----------------------------------------------------

OK, Danielb, I think we were just thrown off by not reading closely enough. You left off the title and one of the cases (a now trivial case because the game has been played), and we didn't notice:

Title: Chances of winning exactly X out of 5 games given that we have already won against UM:

0 wins: 0% (this is the 100% issue zpfled and I asked about)
1 win: 34.9%
2 wins: 43.6%
3 wins: 18.2%
4 wins: 3.1%
5 wins: 0.2%

Hopefully our play on Sunday will require you to redo the work with the updated Pomeroy ratings so we have exact info!

[Dan]

No. 2's visit 'an unbelievable opportunity'

"Guys like challenges. I love challenges. We've got some competitive guys in our locker room, that is what it is. They'll have some butterflies. If they don't, something's wrong," Groce said. "They should feel that edge and feel that way when they get a chance to compete. (Today) is another opportunity to do that."

Marcus Jackson

Groce has experience against Michigan

"He's a really excellent coach," Beilein said. "They do a very good job of putting pressure on the rim offensively and defending people.

"You don't get to that point and have the success he's had by accident.

"We saw that last year. After they beat us fair and square, they won again (beating South Florida in the NCAA Tournament) and then one more big foul shot and they would have beaten North Carolina and been in the Elite Eight."

Tupper

Illinois to test Michigan's perimeter defense

While Michigan encountered some trouble guarding the three point shot in the first half against Purdue, it's safe to say that was mostly due to tremendous individual shooting by the Boilermakers than any significant defensive failure on the part of the Wolverines.

Beilein referenced the second half of the Purdue game as the standard for guarding shots from beyond the arc against Illinois.

"I hope we do as good a job on those two as we did on Byrd in the second half," Beilein said.

http://www.umhoops.com/2013/01/26/no...meter-defense/

[danielb927]

Originally Posted by Illini70math

Danielb927 wrote:

Win tomorrow, 5-game stretch probabilities become:

1 win: 34.9%
2 wins: 43.6%
3 wins: 18.2%
4 wins: 3.1%
5 wins: 0.2%
----------------------------------------------------

zpfled asked: Shouldn't that be 100%?
----------------------------------------------------

OK, Danielb, I think we were just thrown off by not reading closely enough. You left off the title and one of the cases (a now trivial case because the game has been played), and we didn't notice:

Title: Chances of winning exactly X out of 5 games given that we have already won against UM:

0 wins: 0% (this is the 100% issue zpfled and I asked about)
1 win: 34.9%
2 wins: 43.6%
3 wins: 18.2%
4 wins: 3.1%
5 wins: 0.2%

Hopefully our play on Sunday will require you to redo the work with the updated Pomeroy ratings so we have exact info!

I will happily avoid homework and redo things Monday if that's the case!

Actually, even a close loss (<9 points) would give us a boost in Pomeroy.

[mdonsbach]

Originally Posted by trevdv

My Gonzaga and Ohio State pre-game videos were pretty popular, so I thought I'd make one for Michigan too. Enjoy!

http://www.youtube.com/watch?v=WEmgxis7Jak

EDIT: Dan can you make that YouTube link viewable on the forums? I can't seem to get it to work...

Feel free to make those for every game. Excellent work!

[jmm44]

Before the Nebraska game, I had a dream that we beat them by 20. A few nights ago, I dreamt that we beat Michigan by one point in OT. I don't know if this means anything, but I do know that I apparently spend way too much time thinking about Illini basketball.

[trevdv]

Originally Posted by jmm44

Before the Nebraska game, I had a dream that we beat them by 20. A few nights ago, I dreamt that we beat Michigan by one point in OT. I don't know if this means anything, but I do know that I apparently spend way too much time thinking about Illini basketball.

Please, keep these dreams coming!! If we win in OT barely, I will bow down to you. :laugh:

[OrangeAndBlue217]

Originally Posted by trevdv

Please, keep these dreams coming!! If we win in OT barely, I will bow down to you. :laugh:

LOL right! If those dreams continue to come true I will refer to you as chosen one.

[illiniray]

This team has not played well at home, it seems all year. Most if not all the big wins were on the road. Don't know if they are just tight, or what; but sure hope they show up, but I don't fell good about this game.

[lstewart53x3]

Originally Posted by illiniray

This team has not played well at home, it seems all year. Most if not all the big wins were on the road. Don't know if they are just tight, or what; but sure hope they show up, but I don't fell good about this game.

The majority of the bad teams we have played have been at home, so that's probably why we haven't played well in AH. We played just fine against OSU at home.

[trevdv]

In the car on the way to Champaign and it fully hits me: with Michigan primed to move to #1 tomorrow, this is our biggest home game since???

https://twitter.com/ALionEye

Wake Forest, maybe? Wow, it sure makes this game seem bigger than I originally thought if the equivalent to this one is the '04 WF game...

[DaytonIllini]

This game is winnable. If we come out with our A game, which I grant you we have not seen for awhile, I could imagine a win. Hopefully their freshmen play a little like freshmen and we steal a win here. I certainly would not say we should be favored or anything but I think we match up against Michigan better than we do against MSU or Minnesota. Glenn Robinson scares me though.