Page 14 of 14 FirstFirst ... 4121314
Results 261 to 276 of 276
  1. #261
    Join Date
    Dec 2009
    Location
    New York
    Quote Originally Posted by loran16 View Post
    Actually IMO this is where matchups come in. In a long season, your defense (and offense, but I think more on D) is going to face good and bad matchups. But often times teams that make it far into the tourney don't see bad matchups until they're eliminated. So their D seems to improve.
    You are surely correct that the teams that stay alive will see their numbers improve. The phenomenon of defense improving in the postseason is true across multiple sports, though. Playoff baseball, for instance, is a lower run-scoring environment than the regular season. NBA scoring also falls off in the playoffs. (These are two that I know for certain, but it seems anecdotally true elsewhere.) Coaching staffs have increased amounts of time to focus on the opponent, and players probably try harder. Though these traits should apply to both sides of the ball, there does seem to be an ability for defenses to kick it into a new gear in the most important games in a way offenses cannot match.

    Synthesis: I remember a study of soccer tournaments coming out a while back that compared the fortunes of defensive-oriented squads versus offensive-oriented squads. One of the trends they noticed was that an elite defensive unit was more susceptible to an early round upset, but when two elite teams of equal power met in later rounds, the more defensive team was likelier to win. Which supports both your notion that matchups affect defense more (the early round upsets for defensive teams) and the idea that defense can raise its level of play more than offense can (defensive teams defeating offensive teams of equal ability).

  2. #262
    Quote Originally Posted by Des Esseintes View Post
    You are surely correct that the teams that stay alive will see their numbers improve. The phenomenon of defense improving in the postseason is true across multiple sports, though. Playoff baseball, for instance, is a lower run-scoring environment than the regular season. NBA scoring also falls off in the playoffs. (These are two that I know for certain, but it seems anecdotally true elsewhere.) Coaching staffs have increased amounts of time to focus on the opponent, and players probably try harder. Though these traits should apply to both sides of the ball, there does seem to be an ability for defenses to kick it into a new gear in the most important games in a way offenses cannot match.
    Ah but you're misconstruing what we're talking about here. Of course scoring in general should drop in the playoffs, since there will be fewer bad defenses.

    However, what we're talking about is Pomeroy's ADJUSTED defensive and offensive efficiencies, which takes into account the level of offensive and defensive competition a team plays against. Thus the difference in average skill level on O and on D in the tournament SHOULD make no impact on these rankings.
    <devildeac> anyone playing drinking games by now?
    7:49:36<Wander> drink every qb run?
    7:49:38<loran16> umm, drink every time asack rushes?
    7:49:38<wolfybeard> @devildeac: drink when Asack runs a keeper
    7:49:39 PM<CB&B> any time zack runs, drink

    Carolina Delenda Est

  3. #263
    Join Date
    Dec 2009
    Location
    New York
    Quote Originally Posted by loran16 View Post
    Ah but you're misconstruing what we're talking about here. Of course scoring in general should drop in the playoffs, since there will be fewer bad defenses.

    However, what we're talking about is Pomeroy's ADJUSTED defensive and offensive efficiencies, which takes into account the level of offensive and defensive competition a team plays against. Thus the difference in average skill level on O and on D in the tournament SHOULD make no impact on these rankings.
    I don't think I am. In any case, it is not a given that scoring should fall off in the playoffs. There are, after all, just as many fewer bad offenses for your defensive numbers to feast on. In a "perfect" playoff world, better competition would mean adjusted tempo-free stats would remain static, while unadjusted tempo-free stats would converge (i.e. more close games, as a particular team's O scores less and D gives up more).

    If we look at Carolina's tourney run, for example, it makes sense that all those sterling athletes started competing defensively at a higher level. Which meant that a team they might have beaten by 5 in the regular season they beat by 10 in the tournament, and the difference was defensive effort (or increased emphasis by the coaching staff or whatever). That team started 0-2 in conference play and bowed out early in the ACC tournament. It's very hard for me to believe they weren't coasting a bit before responding to the bell in mid-March when the games just so happened to mean more. And the way they responded was not by shooting better or passing better but by defending better.

    Now, Carolina is a single example, and we don't know if things happened for the reasons I described. I do think, though, that it is a reasonable explanation and one that likely applies to a great many teams. Perhaps the best offensive teams give themselves a larger cushion during the regular season, and they don't have to give maximum defensive effort to succeed. When the tournament rolls around, they no longer believe this to be true, their intensity increases, and their defensive rank relative to the other squads--who had to operate closer to maximum defensive capacity all along--rises.

  4. #264
    Join Date
    Feb 2007
    Location
    Durham, NC - Since 1985

    Dork poll update

    Duke is back on top again today in KenPom by a hair over THE Ohio State University.
    Duke is 3rd in Defensive efficiency and 5th in Offensive efficiency.
    Only three teams in top 10 in both categories: Duke, OSU and Kansas.

    Sagarin has Duke at 3 in both Overall and Predictor categories behind OSU and Kansas.

    In the RPI, which everyone is bashing these days as not particluary relevent, Duke is 6th behind Kansas, San Diego St., BYU, OSU and Georgetown...


    And BTW, Nolan is up to 2nd in the KPOY rankings behind The Jimmer.

  5. #265
    Quote Originally Posted by loran16 View Post
    Of course scoring in general should drop in the playoffs, since there will be fewer bad defenses.
    In the playoffs you face teams with above average defenses, but also above average offenses. So its not obvious to me that unadjusted scoring would always go down in playoffs.

    I think there is something to be said for the fact that many players don't always give 100% on both ends of the court, and they usually try harder on offense than defense. So when tourney time rolls around and players up their effort level, defense benefits disproportionately.

    In the NBA, the offensive efficiency of teams in the closing minutes of a game is significantly lower than average--I think this is at least partially because of increased effort on the defensive end.

  6. #266
    Join Date
    Feb 2007
    Location
    Washington, D.C.
    Quote Originally Posted by Delaware View Post
    Duke is back on top again today in KenPom by a hair over THE Ohio State University.
    Duke is 3rd in Defensive efficiency and 5th in Offensive efficiency.
    Only three teams in top 10 in both categories: Duke, OSU and Kansas.

    Sagarin has Duke at 3 in both Overall and Predictor categories behind OSU and Kansas.

    In the RPI, which everyone is bashing these days as not particluary relevent, Duke is 6th behind Kansas, San Diego St., BYU, OSU and Georgetown...


    And BTW, Nolan is up to 2nd in the KPOY rankings behind The Jimmer.
    Once again, looking at Sagarin and Pomeroy you see that there are three teams that are a step better than the rest (Duke, Kansas, and Ohio State) and that you could throw a hat over the three of them, they are so close.

    Things could still change, of course.

  7. #267
    Join Date
    Dec 2009
    Location
    New York
    Quote Originally Posted by Zeb View Post
    In the NBA, the offensive efficiency of teams in the closing minutes of a game is significantly lower than average--I think this is at least partially because of increased effort on the defensive end.
    I saw that study, too. Someone commenting on it, perhaps Zach Lowe at SI, speculated that not only was it a function of defenses trying harder but of offenses try too hard. By that he meant more isolation sets and one-on-one action, which is a far more predictable attack for the defense to try to counter. This feels correct to me, that the best offense is the one you've been running all day. When everyone knows your best player has the ball and is going to try to make a play, the defense has less to worry about. For all his late-game Black Mamba rep, Kobe is more efficient within the confines of the triangle than he is when he breaks them.

  8. #268

    rpi

    Quote Originally Posted by Delaware View Post
    Duke is back on top again today in KenPom by a hair over THE Ohio State University.
    Duke is 3rd in Defensive efficiency and 5th in Offensive efficiency.
    Only three teams in top 10 in both categories: Duke, OSU and Kansas.

    Sagarin has Duke at 3 in both Overall and Predictor categories behind OSU and Kansas.

    In the RPI, which everyone is bashing these days as not particluary relevent, Duke is 6th behind Kansas, San Diego St., BYU, OSU and Georgetown...


    And BTW, Nolan is up to 2nd in the KPOY rankings behind The Jimmer.
    Just checked Inside RPI Daily and Duke is actually up to No. 5 in the RPI (and while we may bash it, I keep saying, it's the one dork poll that the committee uses most extensively). Duke is ahead of fellow No. 1 seed contenders Pitt (No. 7) and Texas (No. 9).

    interesting note -- Duke's 15 wins against top 100 teams are the second best on the list. Only Kansas (with 16 top 100 wins) has more. Duke's seven top 50 wins are also among the best totals ... although Duke's one top 25 win (over UNC) is pretty pathetic.

    (Sorry, no link -- it's an ESPN Insider)

    I was trying to find the conference rankings. Both Sagarin and Pomeroy now have the ACC at No. 4 (behind the Big 10, Big East and Big 12 ... ahead of the Mountain West, Pac 10 and SEC). The RPI still has the ACC at No. 5, behind the Mountain West.

  9. #269

    Wasn't Temple #24 when we played them?

    Quote Originally Posted by Olympic Fan View Post
    Just checked Inside RPI Daily and Duke is actually up to No. 5 in the RPI (and while we may bash it, I keep saying, it's the one dork poll that the committee uses most extensively). Duke is ahead of fellow No. 1 seed contenders Pitt (No. 7) and Texas (No. 9).

    interesting note -- Duke's 15 wins against top 100 teams are the second best on the list. Only Kansas (with 16 top 100 wins) has more. Duke's seven top 50 wins are also among the best totals ... although Duke's one top 25 win (over UNC) is pretty pathetic.

    (Sorry, no link -- it's an ESPN Insider)

    I was trying to find the conference rankings. Both Sagarin and Pomeroy now have the ACC at No. 4 (behind the Big 10, Big East and Big 12 ... ahead of the Mountain West, Pac 10 and SEC). The RPI still has the ACC at No. 5, behind the Mountain West.
    I thought Temple was #24 when we played them. That helps us a little.

  10. #270
    Join Date
    Mar 2007
    Location
    Mount Kisco, NY
    I am not sure I saw this posted anywhere so I'll add it here:

    A few weeks ago, Seth Davis debunked the strength of schedule and non conference argument for title winners.
    http://sportsillustrated.cnn.com/201...bag/index.html

    "The average strength of schedule rank of the last 17 NCAA champions was 28.1, ranging from 3 to 66. The average nonconference strength of schedule rank was 128.4, ranging from 35 to 253. Both averages were lower than I expected, but the range is what really disproves the theory. Some teams have won taking on all comers, while others have sailed through a regular season virtually unchallenged and still ended up hoisting the big trophy."

    I think it's better to look at a fact that SI's Luke Winn touts, that a great predictor of who can win the title comes from teams that are Top 15 in KenPom's adjusted offensive and defensive efficiency. As of today, those teams are:

    Duke - 5th in Adjusted offensive, 3rd in adjusted defensive
    Ohio State - 2nd and 10th
    Kansas - 3rd and 7th
    Purdue - 11th and 6th

    That's it. Texas' offense (29th) holds it back, Pitt is 20th on defense, BYU is just outside because they are 16th on defense, and those are the only one close to being top 15 in both.

  11. #271
    Quote Originally Posted by Billy Dat View Post
    I am not sure I saw this posted anywhere so I'll add it here:

    A few weeks ago, Seth Davis debunked the strength of schedule and non conference argument for title winners.
    http://sportsillustrated.cnn.com/201...bag/index.html

    "The average strength of schedule rank of the last 17 NCAA champions was 28.1, ranging from 3 to 66. The average nonconference strength of schedule rank was 128.4, ranging from 35 to 253. Both averages were lower than I expected, but the range is what really disproves the theory. Some teams have won taking on all comers, while others have sailed through a regular season virtually unchallenged and still ended up hoisting the big trophy."

    I think it's better to look at a fact that SI's Luke Winn touts, that a great predictor of who can win the title comes from teams that are Top 15 in KenPom's adjusted offensive and defensive efficiency. As of today, those teams are:

    Duke - 5th in Adjusted offensive, 3rd in adjusted defensive
    Ohio State - 2nd and 10th
    Kansas - 3rd and 7th
    Purdue - 11th and 6th

    That's it. Texas' offense (29th) holds it back, Pitt is 20th on defense, BYU is just outside because they are 16th on defense, and those are the only one close to being top 15 in both.
    Unfortunately, as we discussed previously, I think both Davis and Winn are looking at post-tournament Pomeroy numbers. According to Pomeroy's pre-tournament numbers, UNC was 35th in defense going into the NCAAT (tied for 15th after winning it). I wish I had pre-tourney numbers prior to 2009, but UNC's 2009 success suggests you shouldn't discount teams like Pitt or BYU, probably Texas, and maybe others.

  12. #272
    Join Date
    Feb 2007
    Location
    Washington DC
    Quote Originally Posted by Billy Dat View Post
    I am not sure I saw this posted anywhere so I'll add it here:

    A few weeks ago, Seth Davis debunked the strength of schedule and non conference argument for title winners.
    http://sportsillustrated.cnn.com/201...bag/index.html

    "The average strength of schedule rank of the last 17 NCAA champions was 28.1, ranging from 3 to 66. The average nonconference strength of schedule rank was 128.4, ranging from 35 to 253. Both averages were lower than I expected, but the range is what really disproves the theory. Some teams have won taking on all comers, while others have sailed through a regular season virtually unchallenged and still ended up hoisting the big trophy."

    I think it's better to look at a fact that SI's Luke Winn touts, that a great predictor of who can win the title comes from teams that are Top 15 in KenPom's adjusted offensive and defensive efficiency. As of today, those teams are:

    Duke - 5th in Adjusted offensive, 3rd in adjusted defensive
    Ohio State - 2nd and 10th
    Kansas - 3rd and 7th
    Purdue - 11th and 6th

    That's it. Texas' offense (29th) holds it back, Pitt is 20th on defense, BYU is just outside because they are 16th on defense, and those are the only one close to being top 15 in both.

    Just to clairfy, is the adjustment for pace? To even out the # of possessions so it's all apples? Gracias.

  13. #273

    top 25

    Quote Originally Posted by Saratoga2 View Post
    I thought Temple was #24 when we played them. That helps us a little.
    Duke is actually 4-0 against teams that were ranked in the top 25 at the time we played them -- No. 4 Kansas State, No. 6 Michigan State, No. 20 UNC and No. 24 Temple.

    But that's a poor way to rank it -- both K-State and Michigan State have fallen far out of the top 25.

    When I said that Duke is 1-1 against the top 25 (small 't') I'm using the RPI metric -- Duke has played two teams that are currently ranked in the top 25 of the RPI, beating No. 11 UNC and losing to No. 14 St. John's. If K-State keeps winning, they could get in soon (they are tied for No. 27 at the moment), which would give Duke two top 25 wins).

    Beating Temple did give Duke a bit of a bump -- the Owls are No. 33 in the RPI. But it was not a top 25 win.

  14. #274
    Join Date
    Feb 2007
    Location
    Orlando, FL
    Quote Originally Posted by superdave View Post
    Just to clairfy, is the adjustment for pace? To even out the # of possessions so it's all apples? Gracias.
    No, the adjustment is for strength of schedule. He used to publish raw and adjusted efficiency numbers until this year when he stopped publishing the raw ones.

    All four factors which make up offensive and defensive efficiency (eFG%, OR%, TO%, and FTA/FGA) are independent of pace.

    Edit: I did some digging on the KPom site and you can still find the raw OE and DE here and compare them to the adjusted values

  15. #275
    Join Date
    Feb 2007
    Location
    Washington DC
    Quote Originally Posted by tbyers11 View Post
    No, the adjustment is for strength of schedule. He used to publish raw and adjusted efficiency numbers until this year when he stopped publishing the raw ones.

    All four factors which make up offensive and defensive efficiency (eFG%, OR%, TO%, and FTA/FGA) are independent of pace.

    Edit: I did some digging on the KPom site and you can still find the raw OE and DE here and compare them to the adjusted values
    Thanks - I appreciate the explanation. I need to dig through KenPom some before the NCAAs start. I like where we are sitting right now!

  16. #276
    Quote Originally Posted by superdave View Post
    Thanks - I appreciate the explanation. I need to dig through KenPom some before the NCAAs start. I like where we are sitting right now!
    Not sure if this is what you were asking, but there's no need to adjust Pomeroy rankings for pace. Those numbers are PER possession, aka tempo-free, so pace is already taken into account in the numbers and SOS #s.
    <devildeac> anyone playing drinking games by now?
    7:49:36<Wander> drink every qb run?
    7:49:38<loran16> umm, drink every time asack rushes?
    7:49:38<wolfybeard> @devildeac: drink when Asack runs a keeper
    7:49:39 PM<CB&B> any time zack runs, drink

    Carolina Delenda Est

Similar Threads

  1. Duke Gets Point in USA Today Poll and #38 Sagarin
    By III in forum Elizabeth King Forum
    Replies: 15
    Last Post: 09-30-2008, 10:48 AM
  2. Dork polls: #1 Sagarin, #3 Pomeroy
    By hurleyfor3 in forum Elizabeth King Forum
    Replies: 7
    Last Post: 02-08-2008, 08:49 AM
  3. Quick Look at RPI, Sagarin and Pomeroy Ratings
    By gw67 in forum Elizabeth King Forum
    Replies: 2
    Last Post: 01-16-2008, 03:52 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •