Monday, March 06, 2006
Seven postseason thoughts
1. Ohio State is indeed the best team in the Big Ten. But the gap between the Buckeyes and the second-best team (Illinois) isn't nearly as large as it looked just two weeks ago. Terence Dials is a beast--this blog has said so now for two years. But he's a beast on a team that devotes 40 percent of its shots to threes. And over their last three games OSU hit just 21.2 percent of said threes (14 of 66). They got away with it, of course, because they were playing Michigan, Northwestern, and Purdue. But if that kind of shooting continues, it will end their season within the next 20 days, even with a favorable seed: the Buckeyes' strong interior game is fueled by opponents' fear of the three. (Note: The Big Ten did not shoot threes very well this year. As it was, Ohio State's late-season shooting slump came within a whisker of handing the crown for best three-point shooting team over to Penn State. OSU hit 36.8 percent of their threes in conference play, a notably low number for a conference leader. Last year Illinois led the league with a 41.3 mark. And in 2004 Michigan State hit 43.4 percent (!) of their threes in conference play.)
2. Illinois, all of a sudden, has an offense. The numbers say so and, more importantly, our eyes say so. Suddenly there's ball movement, there's decisiveness, and there's pressure on opposing defenses. Outside shooting helps, of course, and the Illini hit 40 percent of their threes over the last three games. But the inside-outside balance of this team (over the last six games Illinois drained over 58 percent of their twos), while not on a par with Ohio State's, gives a team with a strong defense a much better shot in March than does defense alone. (Note: Shaun Pruitt has come further in one year than I, for one, thought he would in four. If your team doesn't have two strong post defenders, the Illini, with Pruitt and of course James Augustine, will likely expose that state of affairs.)
3. Iowa has the best defense in the Big Ten. While their perimeter D is merely average, the Hawkeyes don't let you bring the ball in the paint and they limit you to one shot. This D is better at home than on the road, it's true, but there are no more such games, of either kind, on the schedule--only neutral courts. So who'd be a tough match up for an offensively challenged and slightly turnover-prone Hawkeye team in the tournament? Maybe someone who can run, force turnovers, get into the Iowa bench, and hit their threes. Like, say, a certain SEC team coached by a former Iowa assistant.
4. Wisconsin--perhaps fatigued, possibly overmatched, definitely thin--hasn't held an opponent to less than one point per possession in almost a month, since playing pre-resignation/resurrection Indiana in Madison. For a team whose offense consists of Alando Tucker and little else, that level of D--and the seed it will give the Badgers in the NCAA tournament--means Bo Ryan will need to be coaching on all cylinders just to get UW past the first weekend.
5. Indiana, as a result of its win against Michigan Saturday in Ann Arbor, will be recognized by the College Basketball Hall of Fame at a ceremony later today for playing the ugliest game ever played by a winning team in over 100 years of organized intercollegiate basketball. (This game featured 44 turnovers, one of which occurred when a pass hit the scoreboard above the court.) In fact, I have it on good authority that Big Ten officials seriously considered crediting both teams with two losses for this affront to the eyes. But, hey, it's a W and the Hoosiers appear to have dancing in their future--who would've thought that a couple weeks ago? Can they do any damage in March? Their negative efficiency margin in conference play says no. (Of course, that's what West Virginia's negative efficiency margin in Big East play said last year--and all the Mountaineers did was come within a free throw of the Final Four.)
6. Michigan State will, we think, get Matt Trannon back for the Big Ten tournament, where a couple wins or even more could quite plausibly happen and would definitely revive the Izzo's-time-of-year talk. One thing to remember, though: State may be assigned a tougher seed than they've seen in a while. (Remember the sweet draw they had as a 1 in the South in 2001?) In fact, a 7- or 8-seed (or a 9- or 10-seed, but that shouldn't happen) would mean they might well see one of the following the first weekend: Connecticut, Villanova, Duke, Memphis, Texas, or Gonzaga. (And even that, of course, assumes a first-round win.)
7. Michigan is about where Indiana was last year heading into the Big Ten tournament: a limbo that's made all the worse for being so ill-defined. In any case, I wouldn't recommend a loss Thursday against Minnesota. So say they win Thursday: Lester Abram is said to be ready to return. But unless Abram can catalyze a defensive 180 all by his lonesome, this team is going to find it very difficult to go anywhere--in Indy or beyond--allowing as many points per possession as does Purdue.
In today's less Wonk-ish venues....
Team stats have been updated and are now FINAL! Get on over to the sidebar and enjoy.
The brackets are set!
Here's what we'll see this weekend at the Big Ten tournament in Indy (all times ET)....
Thursday
(8) Penn State vs. (9) Northwestern (ESPN2, noon)
(7) Michigan vs. (10) Minnesota (ESPN2, 2:30pm)
(6) Michigan State vs. (11) Purdue (ESPN2, 6pm)
Friday
(1) Ohio State vs. Penn State-Northwestern winner (ESPN, noon)
(4) Wisconsin vs. (5) Indiana (ESPN, 2:30pm)
(2) Iowa vs. Michigan-Minnesota winner (ESPN Plus, 6:40pm)
(3) Illinois vs. Michigan State-Purdue winner (ESPN Plus, 9:10pm)
Saturday
2-3 bracket semifinal (CBS, 1:40pm)
1-4 bracket semifinal (CBS, 4pm)
Sunday
Championship game (CBS, 3:30pm)
The weekend in Big Ten hoops--yesterday!
Ohio State beat Purdue 76-57 in Columbus to win the Big Ten title outright. The Boilers trailed by just four at the half but, as often was the case this season, saw the game slip away after the intermission. Terence Dials led the Buckeyes with 20 badly needed points--the home team couldn't throw the ball in the ocean from a rowboat yesterday, hitting only 4 of 24 threes. Gary Ware led all scorers with 25 points. (Box score.)
The weekend in Big Ten hoops--Saturday!
Illinois beat Michigan State 75-68 in East Lansing. Midway through the second half the Illini broke a 48-all tie by scoring 12 consecutive points, as two threes from Jamar Smith were sandwiched in between two from Dee Brown. For the game Illinois hit 10 of 20 threes. Paul Davis recorded a 21-14 dub-dub and led all scorers. The Spartans played without both Matt Trannon (still sidelined with a broken jaw) and Marquise Gray (out for the rest of the season after breaking a bone in his foot in Thursday night's win over Wisconsin). (Box score.)
Iowa beat Wisconsin 59-44 in Iowa City. Jeff Horner led all scorers with 22 points and Mike Henderson, oddly enough, recorded six assists. No Badger made more shots than he missed. In the final two games of the season at Carver-Hawkeye, Hawkeye opponents (Penn State and Wisconsin) scored just 82 points on 119 offensive possessions--a rate of just 0.69 points per possession. (Box score.)
Indiana beat Michigan 69-67 in Ann Arbor. One-man statistical explosion--for good and evil--Daniel Horton scored 34 points (including a three at the buzzer and 13-of-13 shooting from the line) but turned the ball over six times, as did Marco Killingsworth (alongside, granted, a 19-13 dub-dub). See thought number 5, above, for more. (Box score.)
Northwestern beat Minnesota 57-53 in Evanston. The Gophers missed five of their 19 free throws and coaxed just eight turnovers out of the Wildcats. Such proved to be the difference in yet another close slow game at Welsh-Ryan. Vedran Vukusic led all scorers with 20. (Box score.)
COMING Wednesday!
Media? Coaches? Bah! What savvy fans truly await is the unveiling of the All-Wonk Team (2.0). I can give you this much of a sneak peek....
Last year consensus reigned supreme. The coaches picked Dee Brown, Luther Head, Deron Williams, Mike Wilkinson, and Vincent Grier for first team All-Big Ten. The writers picked Brown, Head, Williams, Wilkinson, and (it's true--they did this) Bracey Wright. I picked Brown, Head, Williams, Wilkinson, and Alan Anderson. So even with three different teams, everyone involved was agreeing 80 percent of the time.
This year, I can tell already, such consensus will not be the case. See you Wednesday.
Wonk back!
Don't just mutter ineffectually; email me!
If everyone played everyone....
Twice so far this year, I've posted emails from alert reader Ross B., who has, you may recall, outlined the concept of Schedule Independent Wins (SIWs) as follows....
Count me among those who miss the full double round-robin Big Ten schedule.
In 1929, a German mathematician by the name of Ernst Zermelo faced a similar problem when he sought to produce standings for a round-robin chess tournament that was not completed. Zermelo's idea was to assign a numeric rating to each player wherein the probability that Player A might beat Player B is a function of their ratings (and, as later mathematicians modified his system, a "home court advantage" parameter).
The ratings are assigned such that for each player, the sum of the probabilities for each game played equals the number of matches actually won by the player. These ratings summarize both wins and schedule strength as a single number. The higher your rating, the better you've played.
If we calculate this Zermelo rating for each team, we can simulate a 20-game Big Ten schedule by adding up the probability of winning home and away vs. each opponent. This number, which we'll call Schedule Independent Wins, reflects the number of games a team would expect to win over a full 20-game schedule if it continues to play at its current level.
Schedule Independent Wins are to traditional standings what tempo-free stats are to traditional statistics.
Got it? Good. Here are the final numbers from Ross, scaled down from the 20-game round-robin to estimate number of wins in an across-the-board (theoretically) neutral 16-game schedule....
1. Ohio State (11.83)
2. Illinois (11.35)
3. Iowa (10.84)
4. Indiana (9.02)
5. Wisconsin (8.90)
6. Michigan State (8.83)
7. Michigan (8.00)
8. Penn State (5.83)
9. Northwestern (5.83)
10. Minnesota (4.84)
11. Purdue (2.73)
Ross adds:
By subtracting Schedule Independent Wins from actual wins, we can now measure how much each team benefited from the unbalanced schedule: 0 represents a neutral schedule, positive numbers an "easy" schedule:
1. Purdue (0.27)
2. Ohio State (0.17)
3. Northwestern (0.17)
4. Penn State (0.17)
5. Iowa (0.16)
6. Minnesota (0.16)
7. Wisconsin (0.10)
8. Michigan (0.00)
9. Indiana (-0.03)
10. Illinois (-0.35)
11. Michigan State (-0.83)
Indiana and Michigan played essentially neutral schedules. Everyone else received a small boost built on the backs of Illinois and, especially, Michigan State. The Spartans played a total of four games against the bottom four teams, the worst possible draw. Still, the league was so competitive that the worst case schedule cost MSU less than one full game.
The Big Ten's tiebreaking system has much room for improvement. Wisconsin gets the 4-seed over Indiana for a home win over the Hoosiers, without the risk of a return game. Similarly, Iowa gets the 2-seed for a home win over Ohio State, compared to a road loss to the same by Illinois. Instead of focusing on one game, where home advantage often makes the difference, the Big Ten would do better to first look at overall schedule strength.
Fortunately, the title race was not affected by the unbalanced schedule this year. Congratulations to the Ohio State Buckeyes, undisputed Big Ten champions.
Ross B.
Well done, Ross. Thanks. ¶ 6:44 AM
1. Ohio State is indeed the best team in the Big Ten. But the gap between the Buckeyes and the second-best team (Illinois) isn't nearly as large as it looked just two weeks ago. Terence Dials is a beast--this blog has said so now for two years. But he's a beast on a team that devotes 40 percent of its shots to threes. And over their last three games OSU hit just 21.2 percent of said threes (14 of 66). They got away with it, of course, because they were playing Michigan, Northwestern, and Purdue. But if that kind of shooting continues, it will end their season within the next 20 days, even with a favorable seed: the Buckeyes' strong interior game is fueled by opponents' fear of the three. (Note: The Big Ten did not shoot threes very well this year. As it was, Ohio State's late-season shooting slump came within a whisker of handing the crown for best three-point shooting team over to Penn State. OSU hit 36.8 percent of their threes in conference play, a notably low number for a conference leader. Last year Illinois led the league with a 41.3 mark. And in 2004 Michigan State hit 43.4 percent (!) of their threes in conference play.)
2. Illinois, all of a sudden, has an offense. The numbers say so and, more importantly, our eyes say so. Suddenly there's ball movement, there's decisiveness, and there's pressure on opposing defenses. Outside shooting helps, of course, and the Illini hit 40 percent of their threes over the last three games. But the inside-outside balance of this team (over the last six games Illinois drained over 58 percent of their twos), while not on a par with Ohio State's, gives a team with a strong defense a much better shot in March than does defense alone. (Note: Shaun Pruitt has come further in one year than I, for one, thought he would in four. If your team doesn't have two strong post defenders, the Illini, with Pruitt and of course James Augustine, will likely expose that state of affairs.)
3. Iowa has the best defense in the Big Ten. While their perimeter D is merely average, the Hawkeyes don't let you bring the ball in the paint and they limit you to one shot. This D is better at home than on the road, it's true, but there are no more such games, of either kind, on the schedule--only neutral courts. So who'd be a tough match up for an offensively challenged and slightly turnover-prone Hawkeye team in the tournament? Maybe someone who can run, force turnovers, get into the Iowa bench, and hit their threes. Like, say, a certain SEC team coached by a former Iowa assistant.
4. Wisconsin--perhaps fatigued, possibly overmatched, definitely thin--hasn't held an opponent to less than one point per possession in almost a month, since playing pre-resignation/resurrection Indiana in Madison. For a team whose offense consists of Alando Tucker and little else, that level of D--and the seed it will give the Badgers in the NCAA tournament--means Bo Ryan will need to be coaching on all cylinders just to get UW past the first weekend.
5. Indiana, as a result of its win against Michigan Saturday in Ann Arbor, will be recognized by the College Basketball Hall of Fame at a ceremony later today for playing the ugliest game ever played by a winning team in over 100 years of organized intercollegiate basketball. (This game featured 44 turnovers, one of which occurred when a pass hit the scoreboard above the court.) In fact, I have it on good authority that Big Ten officials seriously considered crediting both teams with two losses for this affront to the eyes. But, hey, it's a W and the Hoosiers appear to have dancing in their future--who would've thought that a couple weeks ago? Can they do any damage in March? Their negative efficiency margin in conference play says no. (Of course, that's what West Virginia's negative efficiency margin in Big East play said last year--and all the Mountaineers did was come within a free throw of the Final Four.)
6. Michigan State will, we think, get Matt Trannon back for the Big Ten tournament, where a couple wins or even more could quite plausibly happen and would definitely revive the Izzo's-time-of-year talk. One thing to remember, though: State may be assigned a tougher seed than they've seen in a while. (Remember the sweet draw they had as a 1 in the South in 2001?) In fact, a 7- or 8-seed (or a 9- or 10-seed, but that shouldn't happen) would mean they might well see one of the following the first weekend: Connecticut, Villanova, Duke, Memphis, Texas, or Gonzaga. (And even that, of course, assumes a first-round win.)
7. Michigan is about where Indiana was last year heading into the Big Ten tournament: a limbo that's made all the worse for being so ill-defined. In any case, I wouldn't recommend a loss Thursday against Minnesota. So say they win Thursday: Lester Abram is said to be ready to return. But unless Abram can catalyze a defensive 180 all by his lonesome, this team is going to find it very difficult to go anywhere--in Indy or beyond--allowing as many points per possession as does Purdue.
In today's less Wonk-ish venues....
Team stats have been updated and are now FINAL! Get on over to the sidebar and enjoy.
The brackets are set!
Here's what we'll see this weekend at the Big Ten tournament in Indy (all times ET)....
Thursday
(8) Penn State vs. (9) Northwestern (ESPN2, noon)
(7) Michigan vs. (10) Minnesota (ESPN2, 2:30pm)
(6) Michigan State vs. (11) Purdue (ESPN2, 6pm)
Friday
(1) Ohio State vs. Penn State-Northwestern winner (ESPN, noon)
(4) Wisconsin vs. (5) Indiana (ESPN, 2:30pm)
(2) Iowa vs. Michigan-Minnesota winner (ESPN Plus, 6:40pm)
(3) Illinois vs. Michigan State-Purdue winner (ESPN Plus, 9:10pm)
Saturday
2-3 bracket semifinal (CBS, 1:40pm)
1-4 bracket semifinal (CBS, 4pm)
Sunday
Championship game (CBS, 3:30pm)
The weekend in Big Ten hoops--yesterday!
Ohio State beat Purdue 76-57 in Columbus to win the Big Ten title outright. The Boilers trailed by just four at the half but, as often was the case this season, saw the game slip away after the intermission. Terence Dials led the Buckeyes with 20 badly needed points--the home team couldn't throw the ball in the ocean from a rowboat yesterday, hitting only 4 of 24 threes. Gary Ware led all scorers with 25 points. (Box score.)
The weekend in Big Ten hoops--Saturday!
Illinois beat Michigan State 75-68 in East Lansing. Midway through the second half the Illini broke a 48-all tie by scoring 12 consecutive points, as two threes from Jamar Smith were sandwiched in between two from Dee Brown. For the game Illinois hit 10 of 20 threes. Paul Davis recorded a 21-14 dub-dub and led all scorers. The Spartans played without both Matt Trannon (still sidelined with a broken jaw) and Marquise Gray (out for the rest of the season after breaking a bone in his foot in Thursday night's win over Wisconsin). (Box score.)
Iowa beat Wisconsin 59-44 in Iowa City. Jeff Horner led all scorers with 22 points and Mike Henderson, oddly enough, recorded six assists. No Badger made more shots than he missed. In the final two games of the season at Carver-Hawkeye, Hawkeye opponents (Penn State and Wisconsin) scored just 82 points on 119 offensive possessions--a rate of just 0.69 points per possession. (Box score.)
Indiana beat Michigan 69-67 in Ann Arbor. One-man statistical explosion--for good and evil--Daniel Horton scored 34 points (including a three at the buzzer and 13-of-13 shooting from the line) but turned the ball over six times, as did Marco Killingsworth (alongside, granted, a 19-13 dub-dub). See thought number 5, above, for more. (Box score.)
Northwestern beat Minnesota 57-53 in Evanston. The Gophers missed five of their 19 free throws and coaxed just eight turnovers out of the Wildcats. Such proved to be the difference in yet another close slow game at Welsh-Ryan. Vedran Vukusic led all scorers with 20. (Box score.)
COMING Wednesday!
Media? Coaches? Bah! What savvy fans truly await is the unveiling of the All-Wonk Team (2.0). I can give you this much of a sneak peek....
Last year consensus reigned supreme. The coaches picked Dee Brown, Luther Head, Deron Williams, Mike Wilkinson, and Vincent Grier for first team All-Big Ten. The writers picked Brown, Head, Williams, Wilkinson, and (it's true--they did this) Bracey Wright. I picked Brown, Head, Williams, Wilkinson, and Alan Anderson. So even with three different teams, everyone involved was agreeing 80 percent of the time.
This year, I can tell already, such consensus will not be the case. See you Wednesday.
Wonk back!
Don't just mutter ineffectually; email me!
If everyone played everyone....
Twice so far this year, I've posted emails from alert reader Ross B., who has, you may recall, outlined the concept of Schedule Independent Wins (SIWs) as follows....
Count me among those who miss the full double round-robin Big Ten schedule.
In 1929, a German mathematician by the name of Ernst Zermelo faced a similar problem when he sought to produce standings for a round-robin chess tournament that was not completed. Zermelo's idea was to assign a numeric rating to each player wherein the probability that Player A might beat Player B is a function of their ratings (and, as later mathematicians modified his system, a "home court advantage" parameter).
The ratings are assigned such that for each player, the sum of the probabilities for each game played equals the number of matches actually won by the player. These ratings summarize both wins and schedule strength as a single number. The higher your rating, the better you've played.
If we calculate this Zermelo rating for each team, we can simulate a 20-game Big Ten schedule by adding up the probability of winning home and away vs. each opponent. This number, which we'll call Schedule Independent Wins, reflects the number of games a team would expect to win over a full 20-game schedule if it continues to play at its current level.
Schedule Independent Wins are to traditional standings what tempo-free stats are to traditional statistics.
Got it? Good. Here are the final numbers from Ross, scaled down from the 20-game round-robin to estimate number of wins in an across-the-board (theoretically) neutral 16-game schedule....
1. Ohio State (11.83)
2. Illinois (11.35)
3. Iowa (10.84)
4. Indiana (9.02)
5. Wisconsin (8.90)
6. Michigan State (8.83)
7. Michigan (8.00)
8. Penn State (5.83)
9. Northwestern (5.83)
10. Minnesota (4.84)
11. Purdue (2.73)
Ross adds:
By subtracting Schedule Independent Wins from actual wins, we can now measure how much each team benefited from the unbalanced schedule: 0 represents a neutral schedule, positive numbers an "easy" schedule:
1. Purdue (0.27)
2. Ohio State (0.17)
3. Northwestern (0.17)
4. Penn State (0.17)
5. Iowa (0.16)
6. Minnesota (0.16)
7. Wisconsin (0.10)
8. Michigan (0.00)
9. Indiana (-0.03)
10. Illinois (-0.35)
11. Michigan State (-0.83)
Indiana and Michigan played essentially neutral schedules. Everyone else received a small boost built on the backs of Illinois and, especially, Michigan State. The Spartans played a total of four games against the bottom four teams, the worst possible draw. Still, the league was so competitive that the worst case schedule cost MSU less than one full game.
The Big Ten's tiebreaking system has much room for improvement. Wisconsin gets the 4-seed over Indiana for a home win over the Hoosiers, without the risk of a return game. Similarly, Iowa gets the 2-seed for a home win over Ohio State, compared to a road loss to the same by Illinois. Instead of focusing on one game, where home advantage often makes the difference, the Big Ten would do better to first look at overall schedule strength.
Fortunately, the title race was not affected by the unbalanced schedule this year. Congratulations to the Ohio State Buckeyes, undisputed Big Ten champions.
Ross B.
Well done, Ross. Thanks. ¶ 6:44 AM
