Illini Basketball 2023-2024

Status
Not open for further replies.
#137      

DeonThomas

South Carolina
By putting Hawkins into this group, BU loses all credibility.

Instead he needs to qualify his statement with something like this --- "For a 6'11 big, Coleman has a pretty decent looking 3-point shot, but he typically doesn't get enough arc on the ball. Overall, he's OK from beyond the arc. He's just not NBA good. And I've got 3-4 other guys on the roster we'll be looking to take the three before we go to CH. He's probably our 5th best option."
 
#139      
Watched my first in-person Illini game of any sport yesterday (ILL vs PUR) and quickly realized why I belong here instead. 🥹🥲
 
#140      
By putting Hawkins into this group, BU loses all credibility.

Instead he needs to qualify his statement with something like this --- "For a 6'11 big, Coleman has a pretty decent looking 3-point shot, but he typically doesn't get enough arc on the ball. Overall, he's OK from beyond the arc. He's just not NBA good. And I've got 3-4 other guys on the roster we'll be looking to take the three before we go to CH. He's probably our 5th best option."

Yeah, thought that comment was the most eye-raising one of the conference, and was a reminder that these were just practice reports. Ultimately, press conferences don't matter, so no big deal. It was good to hear how the group was doing, even if it's not very different overall from what was said last year.

Of the returning guys, it wouldn't take a ton of improvement to make the team number a heck of a lot better, and of the new guys, Domask is 36%, Guerrier was 35%, and a healthy Goode is probably 40%, so you can understand the optimism. Hawkins otoh, has never shot 30%, and his stroke just doesn't look that good imho. If I were coach, I'd tell him he's only green-lit for 2 a game, so make them count. For reference, I looked at last seasons game logs, and he had 23 games with 3 or more 3pt attempts and averaged about two 3pt attempts per FTA. Yikes
 
#141      
I did a regression analysis to understand which variables correlate most strongly to a team's wins. I'm not finished with it yet but figured I'd share some interesting nuggets it's produced so far.

*It's a small sample size. It's data from 2021-2023 and it's B1G teams only at this point. So 28 observations total.

Independent variables:
Returning minutes
Returning scoring
Total career games played on the roster
Number of RSCI top 100 players
Number of RSCI top 30 players
Number of transfers

Only one of these variables has a strong correlation to the number of team wins - the number of transfers. And it's a negative correlation. So as the number of transfers increase, the number of wins decrease.

Minutes and scoring returning are moderately correlated to more wins.

Number of top 100 and top 30 recruits is very weakly correlated to more wins. Side note, last year's team is tied with 2022 MSU for having the most talented roster of the last 2 years in terms of the number of RSCI top 100 players.

Number of career games played (ie experience) is not correlated to a team's wins.

Admittedly, it's not a very good model yet. These variables account for a little over 70% of the variation in a team's record. And on average, the model only comes within 4 games of predicting a team's win total correctly. Which is kind of a big margin considering the dynamics of a basketball season.

What the model doesn't quantify, is how one very good college player can overcome all of these variables. Purdue shouldn't have been as good as they were last year but they had Edey. Wisconsin shouldn't have been as good as they were in 2022 but they had Davis. Iowa shouldn't have been as good as they were in 2022 but they had Murray. It also doesn't account for the quality of the transfers on the roster (though it does account for their experience and RSCI ranking).

The model did correctly predict the number of wins for last year's Illinois team (20) and predicted 22 wins for the 2022 team (they ended up with 23).

For this upcoming year, it's predicting another 20 win season. The high number of returning minutes and the high number of transfers on the roster effectively cancel each other out. But, again, it really just takes one very good player to outperform this model's prediction, and I think we have that in TSJ.

Overall, I wouldn't put too much stock into it yet but I do intend to continue updating it so, hopefully over time, it'll be more valuable.
 
#142      
My guess is little slower. Harmon lists at 6'4", 180 lbs, and he never ran track at Curie HS. My guess is Harmon just barely beat Frazier, who barely beat Nnanna, so barely a sub-5, maybe 4:55 or so.
There is no reason to guess. Underwood said his pace 12.8 mph.
 
#144      

DeonThomas

South Carolina
I did a regression analysis to understand which variables correlate most strongly to a team's wins. I'm not finished with it yet but figured I'd share some interesting nuggets it's produced so far.

*It's a small sample size. It's data from 2021-2023 and it's B1G teams only at this point. So 28 observations total.

Independent variables:
Returning minutes
Returning scoring
Total career games played on the roster
Number of RSCI top 100 players
Number of RSCI top 30 players
Number of transfers

Only one of these variables has a strong correlation to the number of team wins - the number of transfers. And it's a negative correlation. So as the number of transfers increase, the number of wins decrease.

Minutes and scoring returning are moderately correlated to more wins.

Number of top 100 and top 30 recruits is very weakly correlated to more wins. Side note, last year's team is tied with 2022 MSU for having the most talented roster of the last 2 years in terms of the number of RSCI top 100 players.

Number of career games played (ie experience) is not correlated to a team's wins.

Admittedly, it's not a very good model yet. These variables account for a little over 70% of the variation in a team's record. And on average, the model only comes within 4 games of predicting a team's win total correctly. Which is kind of a big margin considering the dynamics of a basketball season.

What the model doesn't quantify, is how one very good college player can overcome all of these variables. Purdue shouldn't have been as good as they were last year but they had Edey. Wisconsin shouldn't have been as good as they were in 2022 but they had Davis. Iowa shouldn't have been as good as they were in 2022 but they had Murray. It also doesn't account for the quality of the transfers on the roster (though it does account for their experience and RSCI ranking).

The model did correctly predict the number of wins for last year's Illinois team (20) and predicted 22 wins for the 2022 team (they ended up with 23).

For this upcoming year, it's predicting another 20 win season. The high number of returning minutes and the high number of transfers on the roster effectively cancel each other out. But, again, it really just takes one very good player to outperform this model's prediction, and I think we have that in TSJ.

Overall, I wouldn't put too much stock into it yet but I do intend to continue updating it so, hopefully over time, it'll be more valuable.
A really cool endeavor!

Keep us apprised. Sounds promising.
 
#145      
Yeah, thought that comment was the most eye-raising one of the conference, and was a reminder that these were just practice reports. Ultimately, press conferences don't matter, so no big deal. It was good to hear how the group was doing, even if it's not very different overall from what was said last year.

Of the returning guys, it wouldn't take a ton of improvement to make the team number a heck of a lot better, and of the new guys, Domask is 36%, Guerrier was 35%, and a healthy Goode is probably 40%, so you can understand the optimism. Hawkins otoh, has never shot 30%, and his stroke just doesn't look that good imho. If I were coach, I'd tell him he's only green-lit for 2 a game, so make them count. For reference, I looked at last seasons game logs, and he had 23 games with 3 or more 3pt attempts and averaged about two 3pt attempts per FTA. Yikes
Hawkins 3-point % might increase if he shot more of them inside 30'. :)
 
#146      
I did a regression analysis to understand which variables correlate most strongly to a team's wins. I'm not finished with it yet but figured I'd share some interesting nuggets it's produced so far.

*It's a small sample size. It's data from 2021-2023 and it's B1G teams only at this point. So 28 observations total.

Independent variables:
Returning minutes
Returning scoring
Total career games played on the roster
Number of RSCI top 100 players
Number of RSCI top 30 players
Number of transfers

Only one of these variables has a strong correlation to the number of team wins - the number of transfers. And it's a negative correlation. So as the number of transfers increase, the number of wins decrease.

Minutes and scoring returning are moderately correlated to more wins.

Number of top 100 and top 30 recruits is very weakly correlated to more wins. Side note, last year's team is tied with 2022 MSU for having the most talented roster of the last 2 years in terms of the number of RSCI top 100 players.

Number of career games played (ie experience) is not correlated to a team's wins.

Admittedly, it's not a very good model yet. These variables account for a little over 70% of the variation in a team's record. And on average, the model only comes within 4 games of predicting a team's win total correctly. Which is kind of a big margin considering the dynamics of a basketball season.

What the model doesn't quantify, is how one very good college player can overcome all of these variables. Purdue shouldn't have been as good as they were last year but they had Edey. Wisconsin shouldn't have been as good as they were in 2022 but they had Davis. Iowa shouldn't have been as good as they were in 2022 but they had Murray. It also doesn't account for the quality of the transfers on the roster (though it does account for their experience and RSCI ranking).

The model did correctly predict the number of wins for last year's Illinois team (20) and predicted 22 wins for the 2022 team (they ended up with 23).

For this upcoming year, it's predicting another 20 win season. The high number of returning minutes and the high number of transfers on the roster effectively cancel each other out. But, again, it really just takes one very good player to outperform this model's prediction, and I think we have that in TSJ.

Overall, I wouldn't put too much stock into it yet but I do intend to continue updating it so, hopefully over time, it'll be more valuable.
On transfers, it might be interesting to break it down further to transfers in, transfers out from prior season, and/or net transfers. Though number of transfers on the team is negatively correlated, the reason for getting transfers (i.e. graduation/going pro vs. transfers out) probably plays a role. With the Covid year effect, it is also probably a little screwy anyway.
 
#147      
I love that as our “big” lineup against other big front courts. Shooting though..
Yeah, that is only weight-wise bigger by 5 lbs on listing weights on the roster pages, which Underwood made sound way off on Media Day, when he said Guerrier is now at 235 when he lists at 220. Goode lists at 6'7", where Domask lists at 6'6", so Goode should be part of our "big" lineup., when done height-wise.

You also had that right that Goode is the better shooter (48.4% on FG, 42.1% on 3s, last season) compared to Domask (44.7% on FG, 34.8% on 3s), so that is a very good point for Goode starting over Domask.

The only question is whether we need a more "creative" scorer, more like Matt Mayer. Yes, Domask has done that before in forcing up more shots at SIU against pretty good defenders. If we don't want or need that (I hope we don't), then Goode should start and get more time than Domask (as you correctly pointed out). But if we need that creativity, then Domask might start just because he's more willing/capable to get shots off, say if Shannon was out somehow (sorry to even speak of that).

You convinced me to change my mind. Derek Piper also had a good example of Goode's game in Bloomington on 2/18/2023 which is convincing. Goode was 3 for 5 and 1 for 3 for a total of 7 points in that road game to Indiana. Illinois lost by 3 pts, but Piper was right; that was an excellent game for Goode. I looked through Domask's full season last year, and I couldn't find a single game where Domask shot percentages like that.
 
#148      
Yeah, that is only weight-wise bigger by 5 lbs on listing weights on the roster pages, which Underwood made sound way off on Media Day, when he said Guerrier is now at 235 when he lists at 220. Goode lists at 6'7", where Domask lists at 6'6", so Goode should be part of our "big" lineup., when done height-wise.

You also had that right that Goode is the better shooter (48.4% on FG, 42.1% on 3s, last season) compared to Domask (44.7% on FG, 34.8% on 3s), so that is a very good point for Goode starting over Domask.

The only question is whether we need a more "creative" scorer, more like Matt Mayer. Yes, Domask has done that before in forcing up more shots at SIU against pretty good defenders. If we don't want or need that (I hope we don't), then Goode should start and get more time than Domask (as you correctly pointed out). But if we need that creativity, then Domask might start just because he's more willing/capable to get shots off, say if Shannon was out somehow (sorry to even speak of that).

You convinced me to change my mind. Derek Piper also had a good example of Goode's game in Bloomington on 2/18/2023 which is convincing. Goode was 3 for 5 and 1 for 3 for a total of 7 points in that road game to Indiana. Illinois lost by 3 pts, but Piper was right; that was an excellent game for Goode. I looked through Domask's full season last year, and I couldn't find a single game where Domask shot percentages like that.

There is no reason Domask and Goode should be compared to each other. They play completely different roles and have completely different styles of play. Goode is a 3 and D forward and Domask is wing with guard skills. Furthermore comparing Goode's stats to Domask's stats 1 for 1 is completely pointless. They played completely different roles on different teams
 
Status
Not open for further replies.