Illini Basketball 2019-2020

Status
Not open for further replies.
#676      
The point the article is making is that EVEN IF Illinois were to get those extra 3.98 turnovers per game to be the best at forcing turnovers in the NCAA, it just isn't worth that much and it would likely come at the cost of introducing more defensive weaknesses elsewhere. The team that scored the most points per possession in the 2018-19 season was Gonzaga with 1.197 points per possession (https://www.teamrankings.com/ncaa-basketball/stat/offensive-efficiency). 3.98 extra possessions * 1.197 is just 4.76 points per game. And that would be the absolute ceiling since it doesn't factor in that forcing those 3.98 turnovers and running your offense eats up game clock and therefore you aren't really increasing your total possessions per game by 3.98.
You are factoring in the possession gain on offense but it doesn't appear that you are factoring in the opponent's loss of possession. For each turnover the opponent loses their average points/non-turnover possession (because points/possession already factors in the team's turnover rate and would artificially lower the points they miss out on when they do in fact turn it over) and we gain our average points/possession, no?
 
#677      
All of this is to say that you seem not to actually have understood the main points the article is making about BU's defense. The main point is that BU has his guys play on the line and away from their man. This means that Illinois players are directly between the ball and the guy they are defending instead of a few steps off the line towards the basket. It also means that there is distance between them and their defender. Both of these things make it easier for the defender to cut to the basket and beat the defender in the lane. The upside is that it makes it very difficult to pass the ball from 1 side of the court to the other which the vast majority of college offenses rely on so it disrupts teams that have not specifically prepared for Illinois' style of defense. The article author is clearly saying that the system, despite being good against lower talent teams and teams that haven't specifically prepared for it, is weak against teams that plan to play against Illinois since the backdoor cut is an easy way to get an easy 2 points. The author is also saying that the interior defenders are out of position by staying between the ball in a deny position which results in help coming late (sometimes too aggressively so that a foul is given) or not at all.

If you're going to type out a detailed response to an article, it seems like you should at least take the time to read that article and respond to some of the actual criticisms being made. Instead you just said, if we increase our turnovers and get better players we can score more points and get more rebounds (no duh).
I guess I'll start here. Is this where I'm supposed to make the joke about you being the actual author of the article?

Anyway, I read the article just fine, thanks. And I understand the tradeoffs between various styles of defenses. It's clear that Underwood's D does trade backdoor cuts for turnovers, and the whole premise of my argument in favor of his system isn't that it's clearly better at preventing the other team from scoring points, but rather that for the vast majority of teams -- say, every team that doesn't have a Mark Adams on its staff -- you're going to get similar results no matter how you choose to set up your defense. There are scraps to be had at the margins, and those are worth fighting for, but the system isn't going to be the difference between being really good and really bad.

Contrast that with the following two paragraphs from the article:

And that, frankly, is the problem. Underwood's defenses are getting exactly what they want. There are little more opponent turnovers to be squeezed from this system. In other words, you cannot simply state that the players are not running the system properly. They are. It is functioning as designed. The design just sucks.

The deal is a bad one. In exchange for those extra 2-3 turnovers (and that's all it is, in a typical 65ish possession game), Underwood's teams give up easy 2s (54.3% in Big Ten play last year), foul a ton (43.9 FT Rate, last in conference), and generally stink on the boards, for good measure (13th in defensive rebounding percentage in Big Ten play the last two seasons). All for a couple of steals.

I read the first paragraph as saying that there is no room for improvement in terms of forcing turnovers, which to my point about the distance between #1, #25, and #152 on the list being roughly equal, is false. Illinois did not maximize forced turnovers last year, and did not even come close to forcing them at the same rate as they did in 2018.

The second paragraph gets to the point I made above and in my original post about how significant the tradeoffs between turnovers and 'regular' defense are. The author claims that the turnovers provide less of a benefit than focusing on staying behind the ball would. I disagree, and offered a very light amount of analysis to back up the claim. It's my story, and I'm sticking with it.

I just disagree with the notion that we had the players to succeed defensively last season and it was merely Underwood's stubborn refusal to implement the right system that stood in our way.

As for the technical questions on the stats, apologies for not showing my work. I used Pomeroy's tempo-free data for everything. The raw, non-adjusted numbers you've cited have inherent flaws that strip much of their value when doing analysis. Ten offensive turnovers for Wisconsin is a much worse outcome than the same number for Illinois, because Illinois plays at a pace that results in more possessions and more opportunities to turn the ball over. It's much in the same vein as hitting ten home runs in 100 at bats is far more impressive than hitting 10 homers in 200 at bats.

And on your calculation of points per possession where you've cited Gonzaga's total number, note that I was using the PPP for possessions where a shot was taken (and where there was no turnover) to get to 1.27 for my average. You have to exclude the possessions where a turnover was committed to get to a proper value there.

For the vast majority of teams, meaning those that are not outliers in being well-coached, like Texas Tech, or equally poorly coached, I just don't think the system matters all that much. We should try and improve around the margins where possible, but that's not what we're talking about when we talk about missing the dance.
 
#678      
You are factoring in the possession gain on offense but it doesn't appear that you are factoring in the opponent's loss of possession. For each turnover the opponent loses their average points/non-turnover possession (because points/possession already factors in the team's turnover rate and would artificially lower the points they miss out on when they do in fact turn it over) and we gain our average points/possession, no?

It's more complicated...

TDLR: If you treat a forced turnover as adding your teams average points per possession and subtracting your opponents points per possession, you're doing it wrong.

For example consider a 2 minute game where Team A starts with the ball and the average possession is 20 seconds long if no turnover is forced.
Team A would have 3 score-able possessions at shot clock times 2:00-1:40, 1:20-1:00, and 0:40-0:20.
Team B would have 3 score-able possessions at shot clock times 1:40-1:20, 1:00-0:40, and 0:20-0:00.

If Team B forced a turnover 5 seconds into the game:
Team A would have 3 score-able possessions at shot clock times X2:00-1:55X, 1:35-1:15, 0:55-0:35, and 0:15-0:00.
Team B would have 3 score-able possessions at shot clock times 1:55-1:35, 1:15-0:55, and 0:35-0:15.
So both teams still have equal score-able possessions but one of Team A's duration was reduced by 5 seconds (the time into the possession they were turned over). So unless Team B forces the turnover right at the end of the first 20 second duration, they don't gain an additional scoring opportunity. If you extend this over a 40 minutes game, the average duration of each scoring opportunity is not going to be exactly 20 seconds, so the first turnover may or may not actually change the number of scoring opportunities either team receives or even the duration of time for those scoring opportunities.

Another scenario, If Team B forces a turnover 5 seconds into the game, Team A forces a turnover 5 seconds into Team B's second possession, and Team B forces a turnover 5 seconds into Team A's 4th possession:
Team A would have 3 score-able possessions at shot clock times X2:00-1:55X, 1:35-1:15, 1:10-0:50, X0:30-0:25X, and 0:05-0:00.
Team B would have 3 score-able possessions at shot clock times 1:55-1:35, X1:15-1:10X, 0:50-0:30, and 0:25-0:05.
This is more interesting, if the total duration of the elapsed time before turnover is less than 20 seconds (the average possession duration) as in this scenario, and the delta in turnover is 1 or less, there are no additional scoring opportunities generated. However, you can see that if the time elapsed before turnovers had totaled 20-40 seconds Team B with their additional forced turnover would have 1 more scoring opportunity than Team A (and the duration of that opportunity would be 0-20 seconds depending on the time elapsed prior to those turnovers). If the turnovers added up to 40-60 seconds though, the teams would again have the same number of scoring opportunities.

3 turnovers per 2 minutes is a pretty high clip. That's 1.5 per minute or equivalent to 60 turnovers in a 40 minute game... A typical game has closer to 25 forced turnovers or ~6 turnovers per 10 minutes.

If Team B forces turnovers the first 4 times that Team A holds the ball, and Team A forces a turnover only the 2nd time Team B has it (all turnovers taking place 5 seconds into the possession) and we extend the game clock to 10 minutes:
Team A would have 13 score-able possessions at shot clock times X10:00-9:55X, X9:35-9:30X, X9:25-9:20X, X9:00-8:55X, 8:35-8:15, 7:55-7:35, 7:15-6:55, 6:35-6:15, 5:55-5:35, 5:15-4:55, 4:35-4:15, 3:55-3:35, 3:15-2:55, 2:35-2:15, 1:55-1:35, 1:15-0:55, 0:35-0:15.
Team B would have 16 score-able possessions at shot clock times 9:55-9:35, X9:30-9:25X, 9:20-9:00, 8:55-8:35, 8:15-7:55, 7:35-7:15, 6:55-6:35, 6:15-5:55, 5:35-5:15, 4:55-4:35, 4:15-3:55, 3:35-3:15, 2:55-2:35, 2:15-1:55, 1:35-1:15, 0:55-0:35, 0:15-0:00.
Obviously turning your opponent over 4 times more often is great but in this scenario it only resulted in 3 additional score-able possessions over your opponent (which is roughly equal to the margin of forced turnovers you get, but not exactly since if the shot clock were 15 seconds shorter in this example, it would have only been an extra 2 possessions as explained by the earlier examples).

TDLR: If you treat a forced turnover as adding your teams average points per possession and subtracting your opponents points per possession, you're doing it wrong.
 
#681      
TDLR: If you treat a forced turnover as adding your teams average points per possession and subtracting your opponents points per possession, you're doing it wrong.
I don't want to turn this into an argument about how many workouts per week you get if you work out every other day, but I'm not sure what this all refers to. My original post spoke only about the impact that forcing a turnover has on that specific possession and nothing else, and if that's what you're looking at, the impact should be the expected number of points the other team would get if they wound up with a shot (i.e. the 1.27 I cited). There may be other impacts, like maybe a couple of easy buckets by the defending team or whatever, but I don't have the data to measure them and won't pretend that I have the data analysis skills to factor in multiple variables, so I just left them out.

As for adding possessions to the game, you are correct that shorter possessions mean more of them, but I don't think that's to either team's advantage. That's all part of my argument that tempo-free stats are the only ones you should use when looking at stuff like this.
 
#682      
Thanks for sharing that article. It had some pretty interesting stuff in it, but my only minor quibble with the author's defensive analysis is that it's factually incorrect. His two main premises -- that Giorgi provided enough rim protection for the defense to be reasonably effective, and that the team forced turnovers at an elite rate -- are contradicted by the data that he cites.

With respect to turnovers, Illinois was #25 in the nation in turnover rate last year. That's good, but not elite. The distance between #1 and #25 is basically the same as the distance between #25 and the NCAA average. Contrast that with 2018, where Illinois was #4 and a short distance from the very top in that category. There is absolutely room to improve and improve significantly, and as a general rule Underwood's teams have been better at forcing turnovers than last year's team was.

Illinois was also very bad at blocking shots last year, which is directly related to being bad at stopping opponents from making twos. I love Giorgi to death, but he is not a plus defender in the post. He was the best qualifying shot blocker on the team last year, but ranked #17 in the conference in block percentage. Twelve teams had a player higher on that list than Giorgi, and twelve teams had better team block percentages and twelve teams had better two-point defenses.

To hang numbers on his analysis, the distance between Illinois and an average team on turnover percentage is 3.2%. Over a standard 68-possession game, that's 2.17 turnovers added. The average team scores 1.27 points per possession where a shot is taken (i.e. where they don't turn the ball over), so that adds up to 2.76 points per game.

On the other hand, Illinois was roughly as bad at stopping twos last year as they were good at forcing turnovers. The difference between Illinois and an average team on twos was 3.5%. I don't have exact numbers handy on the average number of twos taken per game, but Illinois shot a shade under 38 twos a game last year. We were basically average (or slightly better) in terms of turning the ball over and offensive rebounding, shot a shade fewer threes than average, but were soundly below average at getting to the line. Regardless, I think we shot a more or less average number of twos last year; adjusted for tempo, we'll call it 37.

Dropping our two-point defense to league average would result in 1.3 more made field goals out of that 37 per game, or 2.6 points. That's not a difference of any significance from what is gained by turning the other team over at a higher rate.

I just don't see how this team could have been anything other than average in terms of protecting the rim regardless of their defensive system. They had three post players on the roster. One was Giorgi, who is an undersized 5 who generally plays below the rim. One was Adonis, who was less mobile than Robert Parish when he spent that year on the Bulls' bench at the end of his career. And then there was Samba, who had the tools but looked like a drunken Boston terrier trying to hump a greased watermelon every time he got further than three feet from the basket and was not ready for B1G competition.

Maybe a switch to a more conservative system would have made some difference, but I really don't think that last year's team had the length, athleticism, and experience to be effective on the defensive end. We've just looked out of place in the B1G. I have hopes that the players we're adding will change that, but it remains to be seen.

Not trying to pick on you specifically for this, but the narrative around Underwood supposedly switching defenses and miraculously turning the fortunes of OSU around really isn't that accurate. I'm sure that Underwood made some adjustments, but the biggest thing that happened during OSU's good stretch of results is that the quality of opponents weren't nearly as good as the ones they played during their losing streaks.

OSU started the conference season 0-6, then went 9-1, and finished the season 0-4. Those 11 losses included 9 games against teams that were in the top 20 of the Pomeroy rankings, with one bad loss to #70 Texas and one loss to #30 K-State. Those teams included some very good offenses, including Kansas (twice), Baylor (twice), Iowa St. (three times), and Michigan.

Contrasting that, the 9 wins included only one top-20 win (vs. WVU), with the rest coming against the bottom half of the conference. The season basically played out how you might expect it to for a team squarely in the middle of the Big 12, but by a quirk of scheduling they finished and ended the season with most of their tough games.

I don't have a problem with folks calling Underwood's pre-Illinois track record into question generally because the more I look at it, the thinner it gets. But to me, there just isn't that much compelling evidence that the issue is with the defensive system he's installed. The fortunes of his defenses have tracked fairly closely to how many shots they have blocked, which seems to indicate that personnel, or a lack of length and athleticism, is a major part of the issue.

That's still Underwood's problem to solve, of course. But there's a difference between folks saying that they don't like watching Illinois give up back door layups from an aesthetic standpoint -- maybe your parents were gunned down in cold blood in front of the Monarch Theatre by a back door cut, who knows -- and folks saying that Underwood is failing specifically because of the style he plays.

Either way, I think that the players on the court were capable of playing defense at the level we saw last year, and the only choice is really what flavor of mediocre you prefer. I prefer uptempo, attacking defense that gets burned more than it should. If you prefer watching the other team pass the ball around for 28 seconds before scoring anyway, you do you.

That's a huge advantage in the packline defense. You can cover up lack of athleticism by defending a smaller area and always having help D just a step away. And you can cover up a lack of height by not letting them in the paint nearly as much. In Groce's final year we were 35th in adjD. That was with Maverick Morgan, Leron Black, and Michael Finke. All 3 of those guys played below the rim and were slow as molasses.

I also think the packline is much easier to adapt to each opponent's strengths and weaknesses. Playing Indiana drove me crazy last year. They had one of the worst 3pt% in the country, yet we're pressuring them like crazy because that's what we do. Sometimes this defense over-complicates things. Just force them into jumpshots.

And let's not forget the extreme amount of FT's given up. IMO the new rules make this system almost impossible to play without consistently giving up a lot of FTA's.
 
#683      
for the vast majority of teams -- say, every team that doesn't have a Mark Adams on its staff -- you're going to get similar results no matter how you choose to set up your defense. There are scraps to be had at the margins, and those are worth fighting for, but the system isn't going to be the difference between being really good and really bad.

Certain defenses schemes work better against certain offensive schemes and that's just a fact. If you disagree with that then I don't really know where to start.

When the Lakers and Bulls were dominating with the Triangle offense, teams figured out that to beat that offense you had to focus on limiting the ability of the interior plays to produce. Quick doubles, overplaying the ball, and trapping opponents can result in forcing the post player into a turnover which forces the team to take lower percentage outside shots. Hard ball pressure also clogs the lanes and creates poor spacing which shuts down the triangle offense.

At high levels of college ball, certain offenses are prevalent and teams would be well suited to implement defenses that are strong against those prevalent strategies. A coach would be wise to look at the types of offense played by the schools in their conference and the teams that they might face in the NCAA tournament and run a defense that works well against that offense.

Saying you can run just any defense is clearly false as the "stand at the half court line and link arms defense" would probably not work as well as say the pack-line. Sure there may be a handful of options that would yield similar results, but that doesn't mean that Underwood's system is one of those options. And ultimately that's what the author of the article is trying to say, the style of defense that Underwood is using has never successfully been used at a high level in College basketball. That doesn't mean it can't be, but the claim being made is that there is no evidence that is will work.

Let me go on the record as saying I am actually a fan of Underwood's defense system. I think it's fun to watch and I think it can work with the right players (high basketball IQ, high stamina or a deep bench, and comparable talent to your opponent).

And that, frankly, is the problem. Underwood's defenses are getting exactly what they want. There are little more opponent turnovers to be squeezed from this system. In other words, you cannot simply state that the players are not running the system properly. They are. It is functioning as designed. The design just sucks. The deal is a bad one. In exchange for those extra 2-3 turnovers (and that's all it is, in a typical 65ish possession game), Underwood's teams give up easy 2s (54.3% in Big Ten play last year), foul a ton (43.9 FT Rate, last in conference), and generally stink on the boards, for good measure (13th in defensive rebounding percentage in Big Ten play the last two seasons). All for a couple of steals.
I read the first paragraph as saying that there is no room for improvement in terms of forcing turnovers, which to my point about the distance between #1, #25, and #152 on the list being roughly equal, is false. Illinois did not maximize forced turnovers last year, and did not even come close to forcing them at the same rate as they did in 2018.

The author very clearly did not say that there is "no room for improvement in terms of forcing turnovers" as they clearly identified that there are approximately 2-3 additional turnovers per game that Illinois could force. Also, Illinois had 16.6 forced turnovers per game in 2017-18 (https://www.teamrankings.com/ncaa-basketball/stat/opponent-turnovers-per-game?date=2018-04-03) vs. 15.6 in 2018-19. To say that "Illinois... did not even come close to forcing them at the same rate as they did in 2018" just doesn't seem to fit reality. That's only 1 extra turnover per game. And as the author noted, the ceiling is probably about 2-3 more than we got. I'll say it again, the fact that the gap between #1 and #25 is the same as #25 and #[doesn't really matter] is a meaningless statement. All that matters is how much those additional turnovers are worth, and what it's going to cost you to get them.

The author's point is NOT that it's impossible for Illinois to force more turnovers. The point being made is that Underwood's method of forcing more turnovers, comes at the sacrifice of playing solid interior defense and is a bad decision because the extra turnovers are not worth giving up easy inside shots and the slew of other defensive weaknesses like being out of position to block shots or rebound or helpers being too far away to impact the play.

The author claims that the turnovers provide less of a benefit than focusing on staying behind the ball would. I disagree, and offered a very light amount of analysis to back up the claim. It's my story, and I'm sticking with it.

You say you understand this is the claim being made but your post didn't actually provide any support for why Underwood's defense is capable of being as good as other proven defenses in college basketball. It's fine if you just want to vaguely wave your hands and say, "meh we would have had a sucky defense no matter what." But it's just wrong to say that a coaches choice of defensive scheme doesn't matter.
 
#684      
I don't want to turn this into an argument about how many workouts per week you get if you work out every other day, but I'm not sure what this all refers to. My original post spoke only about the impact that forcing a turnover has on that specific possession and nothing else, and if that's what you're looking at, the impact should be the expected number of points the other team would get if they wound up with a shot (i.e. the 1.27 I cited). There may be other impacts, like maybe a couple of easy buckets by the defending team or whatever, but I don't have the data to measure them and won't pretend that I have the data analysis skills to factor in multiple variables, so I just left them out.

As for adding possessions to the game, you are correct that shorter possessions mean more of them, but I don't think that's to either team's advantage. That's all part of my argument that tempo-free stats are the only ones you should use when looking at stuff like this.

I think you may be reading things out of order and confusing yourself. My response was to AttentionDeficit (which is why I quoted his post). They were saying that I should factor in the points gained for the extra possession earned when forcing a turnover AND also subtract out the points the other team would have gotten if the turnover wasn't forced. My post was saying that is not a correct way of viewing the value of a forced turnover.
 
#685      
That's a huge advantage in the packline defense. You can cover up lack of athleticism by defending a smaller area and always having help D just a step away. And you can cover up a lack of height by not letting them in the paint nearly as much. In Groce's final year we were 35th in adjD. That was with Maverick Morgan, Leron Black, and Michael Finke. All 3 of those guys played below the rim and were slow as molasses.

I also think the packline is much easier to adapt to each opponent's strengths and weaknesses. Playing Indiana drove me crazy last year. They had one of the worst 3pt% in the country, yet we're pressuring them like crazy because that's what we do. Sometimes this defense over-complicates things. Just force them into jumpshots.

And let's not forget the extreme amount of FT's given up. IMO the new rules make this system almost impossible to play without consistently giving up a lot of FTA's.

Thanks for more clearly stating what I am clearly failing to say. Different defensive schemes have different strengths and weaknesses and it's just idiotic to claim that the defense you choose to run doesn't have an impact. (which is what dbeverly26 is clearly saying).
 
#687      

skyIdub

Winged Warrior
rQcD4Ex.gif
 
#688      
Thanks for sharing that article. It had some pretty interesting stuff in it, but my only minor quibble with the author's defensive analysis is that it's factually incorrect. His two main premises -- that Giorgi provided enough rim protection for the defense to be reasonably effective, and that the team forced turnovers at an elite rate -- are contradicted by the data that he cites.

With respect to turnovers, Illinois was #25 in the nation in turnover rate last year. That's good, but not elite. The distance between #1 and #25 is basically the same as the distance between #25 and the NCAA average. Contrast that with 2018, where Illinois was #4 and a short distance from the very top in that category. There is absolutely room to improve and improve significantly, and as a general rule Underwood's teams have been better at forcing turnovers than last year's team was.

Illinois was also very bad at blocking shots last year, which is directly related to being bad at stopping opponents from making twos. I love Giorgi to death, but he is not a plus defender in the post. He was the best qualifying shot blocker on the team last year, but ranked #17 in the conference in block percentage. Twelve teams had a player higher on that list than Giorgi, and twelve teams had better team block percentages and twelve teams had better two-point defenses.

To hang numbers on his analysis, the distance between Illinois and an average team on turnover percentage is 3.2%. Over a standard 68-possession game, that's 2.17 turnovers added. The average team scores 1.27 points per possession where a shot is taken (i.e. where they don't turn the ball over), so that adds up to 2.76 points per game.

On the other hand, Illinois was roughly as bad at stopping twos last year as they were good at forcing turnovers. The difference between Illinois and an average team on twos was 3.5%. I don't have exact numbers handy on the average number of twos taken per game, but Illinois shot a shade under 38 twos a game last year. We were basically average (or slightly better) in terms of turning the ball over and offensive rebounding, shot a shade fewer threes than average, but were soundly below average at getting to the line. Regardless, I think we shot a more or less average number of twos last year; adjusted for tempo, we'll call it 37.

Dropping our two-point defense to league average would result in 1.3 more made field goals out of that 37 per game, or 2.6 points. That's not a difference of any significance from what is gained by turning the other team over at a higher rate.

I just don't see how this team could have been anything other than average in terms of protecting the rim regardless of their defensive system. They had three post players on the roster. One was Giorgi, who is an undersized 5 who generally plays below the rim. One was Adonis, who was less mobile than Robert Parish when he spent that year on the Bulls' bench at the end of his career. And then there was Samba, who had the tools but looked like a drunken Boston terrier trying to hump a greased watermelon every time he got further than three feet from the basket and was not ready for B1G competition.

Maybe a switch to a more conservative system would have made some difference, but I really don't think that last year's team had the length, athleticism, and experience to be effective on the defensive end. We've just looked out of place in the B1G. I have hopes that the players we're adding will change that, but it remains to be seen.

Not trying to pick on you specifically for this, but the narrative around Underwood supposedly switching defenses and miraculously turning the fortunes of OSU around really isn't that accurate. I'm sure that Underwood made some adjustments, but the biggest thing that happened during OSU's good stretch of results is that the quality of opponents weren't nearly as good as the ones they played during their losing streaks.

OSU started the conference season 0-6, then went 9-1, and finished the season 0-4. Those 11 losses included 9 games against teams that were in the top 20 of the Pomeroy rankings, with one bad loss to #70 Texas and one loss to #30 K-State. Those teams included some very good offenses, including Kansas (twice), Baylor (twice), Iowa St. (three times), and Michigan.

Contrasting that, the 9 wins included only one top-20 win (vs. WVU), with the rest coming against the bottom half of the conference. The season basically played out how you might expect it to for a team squarely in the middle of the Big 12, but by a quirk of scheduling they finished and ended the season with most of their tough games.

I don't have a problem with folks calling Underwood's pre-Illinois track record into question generally because the more I look at it, the thinner it gets. But to me, there just isn't that much compelling evidence that the issue is with the defensive system he's installed. The fortunes of his defenses have tracked fairly closely to how many shots they have blocked, which seems to indicate that personnel, or a lack of length and athleticism, is a major part of the issue.

That's still Underwood's problem to solve, of course. But there's a difference between folks saying that they don't like watching Illinois give up back door layups from an aesthetic standpoint -- maybe your parents were gunned down in cold blood in front of the Monarch Theatre by a back door cut, who knows -- and folks saying that Underwood is failing specifically because of the style he plays.

Either way, I think that the players on the court were capable of playing defense at the level we saw last year, and the only choice is really what flavor of mediocre you prefer. I prefer uptempo, attacking defense that gets burned more than it should. If you prefer watching the other team pass the ball around for 28 seconds before scoring anyway, you do you.
My coworker and I have a theory that getting rid of the 5 second closely guarded rule really hurt defenses like what BU wants to run. That used to put more pressure on the ball handlers and probably caused more bad passes that could be converted to turnovers. No data to back it up, but just a thought
 
#689      
Thanks for more clearly stating what I am clearly failing to say. Different defensive schemes have different strengths and weaknesses and it's just idiotic to claim that the defense you choose to run doesn't have an impact. (which is what dbeverly26 is clearly saying).
Dial it back there about 20%, buddy.

I'm not sure how to reconcile this, but we just disagree on what the article is saying. When I read this:

There are little more opponent turnovers to be squeezed from this system. In other words, you cannot simply state that the players are not running the system properly. They are. It is functioning as designed. The design just sucks.

...I think that I'm pretty clearly reading an assertion that there's no path to forcing more turnovers, and therefore, there's no path to an improved performance from the defense that doesn't involve changing the focus from pressure/deny to something else. And when I read this:

The deal is a bad one. In exchange for those extra 2-3 turnovers (and that's all it is, in a typical 65ish possession game), Underwood's teams give up easy 2s (54.3% in Big Ten play last year), foul a ton (43.9 FT Rate, last in conference), and generally stink on the boards, for good measure (13th in defensive rebounding percentage in Big Ten play the last two seasons). All for a couple of steals.

...I think that I'm pretty clearly reading an assertion that last year's team caused about 2-3 turnovers above an average team (which it did), and traded that for all of the bad outcomes that he then lists.

The author's point, somehow, is both that Illinois can't force more turnovers, and also that the turnovers that it does cause cost more than they're worth. I'm arguing that the second point is often assumed but not necessarily true, as is borne out by the numbers that I ran comparing Underwood's team to a hypothetical team that is average in both forcing turnovers and allowing two-pointers.

As for offering evidence that Underwood's defense can work, I'm not sure what you want me to do. I offered an argument against the conventional wisdom that the downside of playing pressure D exceeds the upside. You described the workings of two different defenses in great detail, but then jumped to a conclusion that one is better than the other without offering anything meaningful in terms of evidence other than saying those more conservative defenses are 'proven'. Is that because Virginia used the pack line last year, and Virginia was good? Or did I miss something else here?

Strip away all of the junk and this is just a classic scouts vs. stats argument we're having, and those never get reconciled. (At least, not until the scouts start understanding the stats.) I just think that the x's and o's side of coaching has a minuscule impact on a team's success as compared to getting the best five guys you can out on the court. I have trouble reconciling that with the fact that Underwood is on a way shorter leash in some circles than he would be if he was channeling Dick Bennett, which toasts my biscuits because I think the Bennetts -- all of them -- should be executed in the town square for taking every ounce of entertainment out of college basketball.
 
#690      
That's a huge advantage in the packline defense. You can cover up lack of athleticism by defending a smaller area and always having help D just a step away. And you can cover up a lack of height by not letting them in the paint nearly as much. In Groce's final year we were 35th in adjD. That was with Maverick Morgan, Leron Black, and Michael Finke. All 3 of those guys played below the rim and were slow as molasses.

I also think the packline is much easier to adapt to each opponent's strengths and weaknesses. Playing Indiana drove me crazy last year. They had one of the worst 3pt% in the country, yet we're pressuring them like crazy because that's what we do. Sometimes this defense over-complicates things. Just force them into jumpshots.

And let's not forget the extreme amount of FT's given up. IMO the new rules make this system almost impossible to play without consistently giving up a lot of FTA's.
This quote by Archie Miller is really frustrating for me
 

Attachments

  • Screenshot_20190307-213054.png
    Screenshot_20190307-213054.png
    202.4 KB · Views: 800
#691      
That's a huge advantage in the packline defense. You can cover up lack of athleticism by defending a smaller area and always having help D just a step away. And you can cover up a lack of height by not letting them in the paint nearly as much. In Groce's final year we were 35th in adjD. That was with Maverick Morgan, Leron Black, and Michael Finke. All 3 of those guys played below the rim and were slow as molasses.

I also think the packline is much easier to adapt to each opponent's strengths and weaknesses. Playing Indiana drove me crazy last year. They had one of the worst 3pt% in the country, yet we're pressuring them like crazy because that's what we do. Sometimes this defense over-complicates things. Just force them into jumpshots.

And let's not forget the extreme amount of FT's given up. IMO the new rules make this system almost impossible to play without consistently giving up a lot of FTA's.
On the Groce years, the team definitely was strong on the glass and at keeping the other team off the foul line. But what I always get hung up on is why his two last teams were so different in terms of defensive outcomes even though they were pretty similar in terms of personnel. Maybe Khalid Lewis was really that bad, but I think a major reason why they climbed 100 spots in adjD was that teams hit a threes at a very efficient clip against them in 2016 but didn't in 2017. I'm inclined to think that was much more about the other team just missing or making shots than about anything that was actually under the control of the defense.

I totally agree that fouls are going to be a critical aspect of this year's success or lack thereof, though. It's my hope that with experience and maturity that they'll be able to cut down on the dumb ones and limit the damage. If they don't, it's going to be an uphill battle.
 
#693      
It's more complicated...

TDLR: If you treat a forced turnover as adding your teams average points per possession and subtracting your opponents points per possession, you're doing it wrong.

For example consider a 2 minute game where Team A starts with the ball and the average possession is 20 seconds long if no turnover is forced.
Team A would have 3 score-able possessions at shot clock times 2:00-1:40, 1:20-1:00, and 0:40-0:20.
Team B would have 3 score-able possessions at shot clock times 1:40-1:20, 1:00-0:40, and 0:20-0:00.

If Team B forced a turnover 5 seconds into the game:
Team A would have 3 score-able possessions at shot clock times X2:00-1:55X, 1:35-1:15, 0:55-0:35, and 0:15-0:00.
Team B would have 3 score-able possessions at shot clock times 1:55-1:35, 1:15-0:55, and 0:35-0:15.
So both teams still have equal score-able possessions but one of Team A's duration was reduced by 5 seconds (the time into the possession they were turned over). So unless Team B forces the turnover right at the end of the first 20 second duration, they don't gain an additional scoring opportunity. If you extend this over a 40 minutes game, the average duration of each scoring opportunity is not going to be exactly 20 seconds, so the first turnover may or may not actually change the number of scoring opportunities either team receives or even the duration of time for those scoring opportunities.

Another scenario, If Team B forces a turnover 5 seconds into the game, Team A forces a turnover 5 seconds into Team B's second possession, and Team B forces a turnover 5 seconds into Team A's 4th possession:
Team A would have 3 score-able possessions at shot clock times X2:00-1:55X, 1:35-1:15, 1:10-0:50, X0:30-0:25X, and 0:05-0:00.
Team B would have 3 score-able possessions at shot clock times 1:55-1:35, X1:15-1:10X, 0:50-0:30, and 0:25-0:05.
This is more interesting, if the total duration of the elapsed time before turnover is less than 20 seconds (the average possession duration) as in this scenario, and the delta in turnover is 1 or less, there are no additional scoring opportunities generated. However, you can see that if the time elapsed before turnovers had totaled 20-40 seconds Team B with their additional forced turnover would have 1 more scoring opportunity than Team A (and the duration of that opportunity would be 0-20 seconds depending on the time elapsed prior to those turnovers). If the turnovers added up to 40-60 seconds though, the teams would again have the same number of scoring opportunities.

3 turnovers per 2 minutes is a pretty high clip. That's 1.5 per minute or equivalent to 60 turnovers in a 40 minute game... A typical game has closer to 25 forced turnovers or ~6 turnovers per 10 minutes.

If Team B forces turnovers the first 4 times that Team A holds the ball, and Team A forces a turnover only the 2nd time Team B has it (all turnovers taking place 5 seconds into the possession) and we extend the game clock to 10 minutes:
Team A would have 13 score-able possessions at shot clock times X10:00-9:55X, X9:35-9:30X, X9:25-9:20X, X9:00-8:55X, 8:35-8:15, 7:55-7:35, 7:15-6:55, 6:35-6:15, 5:55-5:35, 5:15-4:55, 4:35-4:15, 3:55-3:35, 3:15-2:55, 2:35-2:15, 1:55-1:35, 1:15-0:55, 0:35-0:15.
Team B would have 16 score-able possessions at shot clock times 9:55-9:35, X9:30-9:25X, 9:20-9:00, 8:55-8:35, 8:15-7:55, 7:35-7:15, 6:55-6:35, 6:15-5:55, 5:35-5:15, 4:55-4:35, 4:15-3:55, 3:35-3:15, 2:55-2:35, 2:15-1:55, 1:35-1:15, 0:55-0:35, 0:15-0:00.
Obviously turning your opponent over 4 times more often is great but in this scenario it only resulted in 3 additional score-able possessions over your opponent (which is roughly equal to the margin of forced turnovers you get, but not exactly since if the shot clock were 15 seconds shorter in this example, it would have only been an extra 2 possessions as explained by the earlier examples).

TDLR: If you treat a forced turnover as adding your teams average points per possession and subtracting your opponents points per possession, you're doing it wrong.

Thank you for the incredibly well-scenarioed response and I apologize for whatever i said in my original question that hinted that i did not understand the concepts of 'possessions' or 'time'. I also appreciate your twice using bold print to say i am doing it wrong. Now i KNOW i am doing it wrong.
I also now know that 3 turnovers per 2 minutes is 1.5 per minute or equivalent to 60 turnovers in a 40 minute game (golly i ran out of fingers and toes), but i feel like i keep missing the part where you explain the right way to calculate the point value of a turnover from a box score/stats page. Please enlighten me, oh Chrono-master!
TLDR*: Please stop assuming we are idiots.
 
#694      

Tevo

Wilmette, IL
Really believe this year is more about the coaching than the players. His defense works very well against teams that are not familiar with it so it is good in tournaments. However, when coaches become familiar with it, they make adjustments. The real answer is in changing the defense within the game to keep the opponents off balance. With a more experienced team, making adjustments becomes practical. I expect to see several variations this year beyond the pressure and the zone they played last year. Do like the constant ball pressure to make the opponents uncomfortable all the time.

I know it's hard enough to get your team to play one defense really well, but one thing I loved that Bill Self used to do (not sure if he still does because I no longer watch his games) is switch from man to zone occasionally during a single possession, especially as the clock runs down. Just when a guy is about to go Hero on your and try driving past or using a pick to get open, we switch to zone. It's also useful coming out of timeouts or on inbounds plays, which Underwood does, too. But being able to suddenly throw a new look at a team can really keep them off-balance. Like everything, thats easier to do with a veteran team who's been able to work on it for a few successive seasons.
 
#695      
Jeff Goodman came out with his list of the top 102 College Basketball players for 2019-20: https://watchstadium.com/news/jeff-...ge-basketball-players-for-2019-20-10-30-2019/

Two Illini, Ayo and Giorgi, made the list. Ayo at #9 and Giorgi at #96. Here's what he had to say about each of them:

9. Ayo Dosunmu, PG, So., Illinois – He’s got good size for the point guard position and is also blessed with athleticism. Look for Dosunmu to be one of the better players at his position in the country, and expect for his team to take a significant jump — especially if he can make shots from the perimeter.
2018-19 Stats: 13.8 ppg, 4.0 rpg, 3.3 apg

96. Giorgi Bezhanishvili, PF, So., Illinois – The native of Georgia finished strong, averaging 15.7 points over the last 11 games, and I fully expect him to take another step forward this season. He’ll expand his range, improve on the glass, and he brings no shortage of intangibles to the table.
2018-19 Stats: 12.5 ppg, 5.2 rpg

Other B1G players who made the list:
1. Cassius Winston, Michigan State
12. Jalen Smith, Maryland
16. Lamar Stevens, Penn State
21. Kaleb Wesson, Ohio State
40. Anthony Cowan, Maryland
44. Zavier Simpson, Michigan
47. Joe Wieskamp, Iowa
63. Xavier Tillman, Michigan State
90. Nojel Eastern, Purdue
99. Daniel Oturu, Minnesota
 
Status
Not open for further replies.