Pregame: Illinois vs Houston, Thursday, March 26th, 9:05pm CT, TBS

Status
Not open for further replies.
#577      
I’ve seen people say the comp game is Sparty. I actually think it’s VCU.

Quick athletic guards. Lots of ball pressure and trying force turnovers. Bigs that are smaller than us but have a big advantage in athleticism. That’s Houston and it’s also VCU.

If we can play with the same defensive bite and clog gaps like we did vs VCU, I think Houston will similarly have to settle for long contested 2s and will probably struggle to consistently score.

To me the biggest key is our passing and .5 second offense. If the ball doesn’t stick we should be able to see over the top of their pressure and create advantages on the back end. This may be an E8 type matchup but in some sense we are fortunate it’s a S16 matchup because it means we have a longer prep. My guess is we are practicing a couple of sets right now so we won’t have to think or read too much because we know where the ball is going. Think the second Nebraska game. We ran 1-2 sets over and over again and were able to consistently beat their defense and get shots at the rim.

All of that said, we could come in with the perfect gameplan and still just be overwhelmed by their toughness and athleticism and lose by 20. But we present matchup problems for them too and if we play our game we have a chance. All you can ask for at this stage.
I like this take. Many of these podcast talking about this being an elite8 game. I like the fact that we have time to prepare for this one. If we play hard nosed, we have a great chance to take this one.
 
#578      
Fun fact for you young folks, Ron Gunther was ready to hire Sampson but when he showed up at the airport in Oklahoma there was a media contingent waiting for him and he turned around and left without interviewing him. Gunther always acted in stealth mode and was pissed off. Can’t remember if that was when we ended up hiring Self or Weber though.
Believe that was just prior to hiring Self. Remember it well from the early Illinois forum days!
 
#579      
If we can play with the same defensive bite and clog gaps like we did vs VCU, I think Houston will similarly have to settle for long contested 2s and will probably struggle to consistently score.
When I watched their earlier games, they looked to be a decent mid-range shooting team. I think that might be a difference from VCU in that they can consistently hit that long 2 and VCU struggled after a point. Our ball screen defense definitely needs to be on point for this game.
 
#580      
There's a lot of talk on here about how Illinois has fared against teams like MSU and Michigan, but not much on Houston's own track record. And here's the deal: Houston has struggled just as much as us with that top tier of opponents. They are 5-6 in Q1A (we are 5-7). The toughest opponent they've beaten is Arkansas (#17 KP, #15 NET).

Looking for a similar efficiency profile to ours in the Big 12, the most similar I found was Texas Tech. High ranked offense, lower ranked defense (worse than us in both however). Houston went 1-1 against Tech, and both games were close, (yes, even the one in Houston).

In fact when you look at their Q1 games, a lot of close games, not many blowouts.

Looking at Houston's resume, I don't understand the Michigan 2.0 talk. Where have they shown they can do that to a top-10 team? They simply haven't. We aren't Baylor and we aren't Texas A&M. They may ultimately beat us, but I'd be surprised if we weren't in it to the end.
The Office Thank You GIF
 
#584      
There's a lot of talk on here about how Illinois has fared against teams like MSU and Michigan, but not much on Houston's own track record. And here's the deal: Houston has struggled just as much as us with that top tier of opponents. They are 5-6 in Q1A (we are 5-7). The toughest opponent they've beaten is Arkansas (#17 KP, #15 NET).

Looking for a similar efficiency profile to ours in the Big 12, the most similar I found was Texas Tech. High ranked offense, lower ranked defense (worse than us in both however). Houston went 1-1 against Tech, and both games were close, (yes, even the one in Houston).

In fact when you look at their Q1 games, a lot of close games, not many blowouts.

Looking at Houston's resume, I don't understand the Michigan 2.0 talk. Where have they shown they can do that to a top-10 team? They simply haven't. We aren't Baylor and we aren't Texas A&M. They may ultimately beat us, but I'd be surprised if we weren't in it to the end.
That talk is just that, talk. Complete nonsense at that. Michigan won the best conference in America by 4 games. Houston finished 3-5 against the top teams in the Big 12.
Michigan 2.0 my !!!.
 
#585      
It's a good point really. They're as odd and unique defensively as we are offensively.

They are strong on the offensive glass though. This game is a clipboard-smashing JV coaches dream, it really is all going to be decided by who boxes out better.
To be fair Houston is 26th in the country in offensive rebound rate- I know that is “green” on Torvik but not really elite as you get down to the best teams in the country now. They are only 126th in defensive reb rate- so we have to exploit that
 
#586      
To be fair Houston is 26th in the country in offensive rebound rate- I know that is “green” on Torvik but not really elite as you get down to the best teams in the country now. They are only 126th in defensive reb rate- so we have to exploit that
That doesn’t fit the narrative. Houston is ‘tough’, ‘physical’, ‘Michigan 2.0’, ‘great defensive team’, ‘great rebounding team’. How did they even lose a game? Even lost one at home where they are invincible.
 
#589      
A statistical look at the matchup. Because the teams played strong schedules (Illini 13th toughest, Houston 18th toughest), these might be meaningful comparisons:
--A/TO: Hou +1.73, ILL +1.68 (closer than I thought)
--Bench pts: ILL 20, Hou 17
--FG%: ILL 46.46%, Hou 45.09%
--FG Def: Hou 39.51%, ILL 40.93%
--Eff FG%: ILL 55.3%, Hou 52.2%
--3's Att/game: ILL 31.4 (#12), Hou 25.1 (#118)
--3's made: ILL 11 (#9), Hou 8.8 (this could mean a 6-pt ILL advantage given how equal the next 2 categories are)
--3-pt%: ILL 34.99%, Hou 34.92%
--3-pt% Def: ILL 31%, Hou 31.4%
--Fouls/game: ILL 13.3 (#2), Hou 17.9 (#221) (another possible ILL advantage if the whistles are neutral)
--FTA: ILL 20.3, Hou 16.7 (#334/361)
--FT%: ILL 78.44% (#11), Hou 77.20% (#25)
--Reb margin: ILL +10 (#7), Hou +3 (#75) (maybe ILL's biggest advantage)
--Off Reb: ILL 13.65, Hou 12.83
--Pts/game: ILL 84.7 (#17), Hou 77.5 (#136)
--Pts allowed/game: Hou 62.2 (#2), ILL 69.4 (#17)
--Hou points scored vs. ILL points allowed: 72
--ILL points scored vs. Hou points allowed: 73
 
#590      
Watched the highlights of the big 12 championship. I think we can replicate alot of the actions Arizona seemed to have success with:

1. Peat really hurt Houston in a short roll. David isn’t the athlete Peat is but he’s good at using his body to create space and is a better passer.
2. Bradley had success in ball screens. Keaton isn’t as quick but he’s a lot longer and we’ve seen him score over smaller guards/wings time and time again.
3. They were able to create isolation opportunities for Karchenkov to drive and score. From what I can see Andrej is a better athlete/driver than Karchenkov.
4. They were able to get open 3s off of ball screen action/dribble penetration. Arizona only shot 13 but made 6. We should be able to generate open looks for our shooters.

Overall I’m encouraged. I think there are opportunities if we scheme correctly and come out in attack mode. We just can’t get back on our heels.

Analytically and really schematically we are much closer to Arizona who they lost to twice than the Texas Techs and Arkansas who are their best wins. And from what I saw KU had talent but was a bit of a hot mess especially towards the end of the year. Think we’ve got a shot if we’re aggressive.
 
#591      
A statistical look at the matchup. Because the teams played strong schedules (Illini 13th toughest, Houston 18th toughest), these might be meaningful comparisons:
--A/TO: Hou +1.73, ILL +1.68 (closer than I thought)
--Bench pts: ILL 20, Hou 17
--FG%: ILL 46.46%, Hou 45.09%
--FG Def: Hou 39.51%, ILL 40.93%
--Eff FG%: ILL 55.3%, Hou 52.2%
--3's Att/game: ILL 31.4 (#12), Hou 25.1 (#118)
--3's made: ILL 11 (#9), Hou 8.8 (this could mean a 6-pt ILL advantage given how equal the next 2 categories are)
--3-pt%: ILL 34.99%, Hou 34.92%
--3-pt% Def: ILL 31%, Hou 31.4%
--Fouls/game: ILL 13.3 (#2), Hou 17.9 (#221) (another possible ILL advantage if the whistles are neutral)
--FTA: ILL 20.3, Hou 16.7 (#334/361)
--FT%: ILL 78.44% (#11), Hou 77.20% (#25)
--Reb margin: ILL +10 (#7), Hou +3 (#75) (maybe ILL's biggest advantage)
--Off Reb: ILL 13.65, Hou 12.83
--Pts/game: ILL 84.7 (#17), Hou 77.5 (#136)
--Pts allowed/game: Hou 62.2 (#2), ILL 69.4 (#17)
--Hou points scored vs. ILL points allowed: 72
--ILL points scored vs. Hou points allowed: 73
So all of that means- we win by 1!
 
#592      
A statistical look at the matchup. Because the teams played strong schedules (Illini 13th toughest, Houston 18th toughest), these might be meaningful comparisons:
--A/TO: Hou +1.73, ILL +1.68 (closer than I thought)
--Bench pts: ILL 20, Hou 17
--FG%: ILL 46.46%, Hou 45.09%
--FG Def: Hou 39.51%, ILL 40.93%
--Eff FG%: ILL 55.3%, Hou 52.2%
--3's Att/game: ILL 31.4 (#12), Hou 25.1 (#118)
--3's made: ILL 11 (#9), Hou 8.8 (this could mean a 6-pt ILL advantage given how equal the next 2 categories are)
--3-pt%: ILL 34.99%, Hou 34.92%
--3-pt% Def: ILL 31%, Hou 31.4%
--Fouls/game: ILL 13.3 (#2), Hou 17.9 (#221) (another possible ILL advantage if the whistles are neutral)
--FTA: ILL 20.3, Hou 16.7 (#334/361)
--FT%: ILL 78.44% (#11), Hou 77.20% (#25)
--Reb margin: ILL +10 (#7), Hou +3 (#75) (maybe ILL's biggest advantage)
--Off Reb: ILL 13.65, Hou 12.83
--Pts/game: ILL 84.7 (#17), Hou 77.5 (#136)
--Pts allowed/game: Hou 62.2 (#2), ILL 69.4 (#17)
--Hou points scored vs. ILL points allowed: 72
--ILL points scored vs. Hou points allowed: 73
Really excellent breakdown. I was extremely surprised that Houston is only +3 in rebounding. I have seen most of their games being a Houston native and I thought that number would be much higher.

Surprising.
 
#593      
A statistical look at the matchup. Because the teams played strong schedules (Illini 13th toughest, Houston 18th toughest), these might be meaningful comparisons:
--A/TO: Hou +1.73, ILL +1.68 (closer than I thought)
--Bench pts: ILL 20, Hou 17
--FG%: ILL 46.46%, Hou 45.09%
--FG Def: Hou 39.51%, ILL 40.93%
--Eff FG%: ILL 55.3%, Hou 52.2%
--3's Att/game: ILL 31.4 (#12), Hou 25.1 (#118)
--3's made: ILL 11 (#9), Hou 8.8 (this could mean a 6-pt ILL advantage given how equal the next 2 categories are)
--3-pt%: ILL 34.99%, Hou 34.92%
--3-pt% Def: ILL 31%, Hou 31.4%
--Fouls/game: ILL 13.3 (#2), Hou 17.9 (#221) (another possible ILL advantage if the whistles are neutral)
--FTA: ILL 20.3, Hou 16.7 (#334/361)
--FT%: ILL 78.44% (#11), Hou 77.20% (#25)
--Reb margin: ILL +10 (#7), Hou +3 (#75) (maybe ILL's biggest advantage)
--Off Reb: ILL 13.65, Hou 12.83
--Pts/game: ILL 84.7 (#17), Hou 77.5 (#136)
--Pts allowed/game: Hou 62.2 (#2), ILL 69.4 (#17)
--Hou points scored vs. ILL points allowed: 72
--ILL points scored vs. Hou points allowed: 73
I do like that some of our relative strengths could be synergistic. E.g., if Illini can get Tugler (5.1 fouls per 40) or Cenac (3.7 fouls per 40) in foul trouble, the rebounding advantage potentially could really assert itself in a game defining way.
 
#595      
What he means is that in the Kenpom algo he weights (changes) the value of a statistic (3 pt field goal just as an example) depending on opponent and venue.

Keaton Wagler's 3 pointer (or Illinois as a team) @ Purdue is worth more than Keaton Wagler (or Illinois as a team) making a 3 pointer at home against Colgate.

The issue with that is that the teams that play the "heavier" schedule are going to show up higher in all the Kenpom ratings which where the statistical bias comes in.
It isn't controversial that scoring/defending against good defenses/offenses (and playing on the road) is generally tougher, so having no adjustment for opponent/location would incorrectly bias teams that played strong/weak schedules and/or lots of home/road games. There are of course biases for individual teams (with lower/higher home court advantage, or who play better/worse than typical against good/bad teams), but this will not be systematic (it will vary in magnitude and direction for each team)

But to keep it really simple - there is no statistical correlation between Kenpon rankings and tournament wins. In fact the correlation is slightly negative. Over the past 10 years when the NCAA tournament was held fewer than 50% of the 8 highest ranking Kenpom teams prior the tourney, actually made it to the elite 8.
I've been saying all along that I think betting markets are better than KenPom (even after ignoring teams shorthanded by injury/etc), just not by a lot. But the number of top teams advancing to the E8 isn't a useful metric without comparing against betting markets (and without considering the odds KenPom gave them in the first place). This year, betting futures implied only ~4.6 of the top 8 (by betting odds) would get to the E8, and the top seeds have become more dominant in the last 5-10 years (plus seeding was/is so inaccurate that top 8 teams can face each other in the S16, like us this year), so it was presumably less in the past. So KenPom ratings probably wouldn't even have expected 50% of its top 8 to get there either over the last 10yrs.

The Prediction Tracker (link) compares computer ranking systems against betting markets- both straight up and against the spread. KenPom and Torvik aren't included, but eleven others are. They typically perform very well- this year, the betting favorite won 72% of the time, and an average of all 11 computer models matched that (slightly ahead by a negligible amount). Against the spread, the computer avg was correct 50.5% of the time with an average error of 9.1pts compared to 9.0 for the line, and the bias was within 0.07pts of the line. This is pretty amazing considering the lack of data during the early part of the season and no information on injuries, etc. There's enough year-to-year variation, though, that I wouldn't expect computer rankings to regularly beat the line, and certainly not by enough to be profitable.

Perhaps the tournament is especially hard since it involves teams from different conferences that haven't faced each other yet, and (after the first round or two) teams that are likely very similar in quality. Mark Moog's site (link) doesn't compare ranking systems against betting markets, but does allow you to separate out by time range. Over the last 10 tournaments, KenPom accurately predicted 69.3% of game winnners, and to me that's close enough to the full-season computer ranking results to maintain reasonable confidence in it for the tourney.

Edit: here were the top 8 implied by betting odds (after adjusting for difficulty of draw): Arizona, Michigan, Duke, Florida, Houston, Iowa State, Illinois, Purdue. Aside from some slight differences in order, this is identical to Torvik and KenPom. So they will by definition perform equivalently to the betting odds in # of teams in the E8.

Who ranks higher in Kenpom (statistically speaking) is a money loser (if you are professional spread better) in the NCAA tourney.
I'd love to see a source for this, especially given the above links. And a model has to be meaningfully better than the line just to break even, so a money loser can still be very accurate.
 
Last edited:
#598      
I honestly think we are going to win by double digits. I don't think Houston can stop us for a whole game and I think our size will be too much for them. It will be close at the half, and then we will break away.
 
#599      
I hate living in the west coast man. Sporting times are ridiculous
It's worse here on the east coast, as I now live in Connecticut near NYC, late games last well past my bedtime. :)
 
#600      
Fun fact for you young folks, Ron Gunther was ready to hire Sampson but when he showed up at the airport in Oklahoma there was a media contingent waiting for him and he turned around and left without interviewing him. Gunther always acted in stealth mode and was pissed off. Can’t remember if that was when we ended up hiring Self or Weber though.
Self I believe.
 
Status
Not open for further replies.
Back