If i had to look for one MINOR negative on a great night, the sloppy play the last couple of minutes kept Marquette from betting in the top 100 at www.kenpom.com. The problem with the system is all points, so beating Arizona State by 7 at home dropped them from 67 to 74th, while it improved MU from 111th to 103rd.
By comparison, when Georgia ran up a 31-point win on 244th rank, they moved from 69th to 50th. Keeping the margin around 15 would have pushed MU up well into the top 100.
Michigan is ranked #60 in kenpom, we are 103, and Sam houston is 105. All of us have 4 losses. Tell me how sam houston is only 2 spots from us but we are 43 from Michigan. And Georgia Tech is 93. Kenpom rankings are garbage.
Here is a link to the schedule for UC Santa Barbara, a 4-5 team that is ranked 89 on Kenpom (14 spots ahead of us).
http://espn.go.com/mens-college-basketball/team/_/id/2540/uc-santa-barbara-gauchos
Quote from: KipsBayEagle on December 17, 2014, 03:14:16 AM
Michigan is ranked #60 in kenpom, we are 103, and Sam houston is 105. All of us have 4 losses. Tell me how sam houston is only 2 spots from us but we are 43 from Michigan. And Georgia Tech is 93. Kenpom rankings are garbage.
It is probably only modesty that prevented you from posting the link to your own ratings system. I assure you we would all like to look at it and learn from you.
Quote from: KipsBayEagle on December 17, 2014, 03:22:22 AM
Here is a link to the schedule for UC Santa Barbara, a 4-5 team that is ranked 89 on Kenpom (14 spots ahead of us).
http://espn.go.com/mens-college-basketball/team/_/id/2540/uc-santa-barbara-gauchos
Hey, my daughter goes there! Leave them out of it :)
Quote from: KipsBayEagle on December 17, 2014, 03:14:16 AM
Michigan is ranked #60 in kenpom, we are 103, and Sam houston is 105. All of us have 4 losses. Tell me how sam houston is only 2 spots from us but we are 43 from Michigan. And Georgia Tech is 93. Kenpom rankings are garbage.
I agree. I think basketball is a bad game to use point margin in the ranking system. Free throws in garbage time have an unfair impact on that, and should not be considered. This is why I wouldn't bet on a line in a basketball game (I'd be more comfortable with betting on football, for example). Wins and losses, strength of schedule, and home court advantage is all that should be considered. This is what RPI takes into account.
No system is perfect, but I'll take RPI over KenPom any day.
We are 80 in the RPI, and that is before our win last night. It will be updated later today.
www.warrennolan.com has MU RPI at 60 with the win. If we can rattle off a 5-6 winning streak over the next couple weeks and find ourselves in the RPI top 50 heading into the meat of BE play I like our chances at sneaking into the tourney. Feed the Fisch!!!
Quote from: PaceArrow02 on December 17, 2014, 09:24:18 AM
www.warrennolan.com has MU RPI at 60 with the win. If we can rattle off a 5-6 winning streak over the next couple weeks and find ourselves in the RPI top 50 heading into the meat of BE play I like our chances at sneaking into the tourney. Feed the Fisch!!!
Agreed! If we just hold serve we will be 8-4 heading into the BEAST season, and we are in a decent position. Get 11 or 12 wins in the conference season and we are in the tournament discussion.
Quote from: KipsBayEagle on December 17, 2014, 03:14:16 AM
Michigan is ranked #60 in kenpom, we are 103, and Sam houston is 105. All of us have 4 losses. Tell me how sam houston is only 2 spots from us but we are 43 from Michigan. And Georgia Tech is 93. Kenpom rankings are garbage.
Umm, no they are not. Go ahead and check the Vegas lines and Kenpom estimates. I do it routinely. They are within a point or point and a half 90%-95% of the time. The over/unders are extremely close as well. Doesn't account for injuries, players like Fischer being eligible, etc. Marquette was not a good team without Fischer, now they look decent.
Vegas knows what they're doing, and they use similar statistics to set lines and over/under. KenPom is awesome, my friend and I use it to bet against Vegas. We look for 4 or 5 point differences, and then tease to move 4 more points. It works, I recommend it to sport betters. The 4 or 5 point differences are hard to come by, it does take some time to research.
Quote from: KipsBayEagle on December 17, 2014, 03:14:16 AM
Michigan is ranked #60 in kenpom, we are 103, and Sam houston is 105. All of us have 4 losses. Tell me how sam houston is only 2 spots from us but we are 43 from Michigan. And Georgia Tech is 93. Kenpom rankings are garbage.
Especially at this time of year. Wait until we are 20 games in and in conference. If we keep improving and winning, that ranking will take care of itself. Wojo still is feeling where he must distribute minutes, although this game should help him solidify that. The best lineup:
C--Luke and Taylor back up
PF-Anderson and Taylor
PG-Carlino
2G-Du. Wilson
Wing-JJJ
The above is also most conducive to running fast breaks, with Luke patrolling the interior for rebounds to start the run.
Quote from: willie warrior on December 17, 2014, 09:41:49 AM
Especially at this time of year. Wait until we are 20 games in and in conference. If we keep improving and winning, that ranking will take care of itself. Wojo still is feeling where he must distribute minutes, although this game should help him solidify that. The best lineup:
C--Luke and Taylor back up
PF-Anderson and Taylor
PG-Carlino
2G-Du. Wilson
Wing-JJJ
The above is also most conducive to running fast breaks, with Luke patrolling the interior for rebounds to start the run.
I agree with this .
Quote from: KipsBayEagle on December 17, 2014, 03:14:16 AM
Michigan is ranked #60 in kenpom, we are 103, and Sam houston is 105. All of us have 4 losses. Tell me how sam houston is only 2 spots from us but we are 43 from Michigan. And Georgia Tech is 93. Kenpom rankings are garbage.
They're predictive ratings that use past performance from an efficiency perspective and adjust for strength of opponent played. It has nothing to do with wins and losses, as that has not been shown to be the best predictor over time. And at this point of the season still have some preseason expectations built in (those will disappear proportionately with each subsequent game played, and will not factor in at all soon). Plus, in some of the games we DID play this year, we were hot garbage against some bad teams.
HOWEVER - those stats for the first 8 games do not account for Luke, and thus I believe we will be underrated via this system this year.
But that does not mean KenPom's ratings are "garbage." You just have to know how they're calculated and make adjustments that may be necessary.
Do individual ratings factor into KenPom rankings, or is it simply overall team performance (with individual ratings being a separate system)?
Where I'm going with this is whether Luke's ratings could be retroactively applied to MU's season pre-ASU in order to generate something of a metric as to where MU is ranked as the team stands right now, not where they're ranked based on how the team has performed so far.
In other words, given that his is a predictive model, how does KenPom adjust for things like injuries, suspensions, residency, etc.
Quote from: 79Warrior on December 17, 2014, 09:03:33 AM
Hey, my daughter goes there! Leave them out of it :)
Pic?
Quote from: Benny B on December 17, 2014, 11:26:16 AM
Do individual ratings factor into KenPom rankings, or is it simply overall team performance (with individual ratings being a separate system)?
Where I'm going with this is whether Luke's ratings could be retroactively applied to MU's season pre-ASU in order to generate something of a metric as to where MU is ranked as the team stands right now, not where they're ranked based on how the team has performed so far.
In other words, given that his is a predictive model, how does KenPom adjust for things like injuries, suspensions, residency, etc.
As far as I know, individual ratings do not factor in to overall team ratings. However, I'm sure Bama or Henry Sugar can and will do some sort of retroactive analysis of Luke's hypothetical impact to the first 8 games once enough data is available. Cuz they do awesome stuff like that.
(http://express.ece.ucsb.edu/pictures/ucsb_aerial2.jpg)
Here you go, Jay Bee. She's a beaut.
Quote from: Warrior Code on December 17, 2014, 11:52:40 AM
(http://express.ece.ucsb.edu/pictures/ucsb_aerial2.jpg)
Here you go, Jay Bee. She's a beaut.
Well played to protect the honor of 79warriors daughter! I believe we need a new rule for the blog that says asking for photos of someone's daughter is off limits!
Quote from: Benny B on December 17, 2014, 11:26:16 AM
Do individual ratings factor into KenPom rankings, or is it simply overall team performance (with individual ratings being a separate system)?
Where I'm going with this is whether Luke's ratings could be retroactively applied to MU's season pre-ASU in order to generate something of a metric as to where MU is ranked as the team stands right now, not where they're ranked based on how the team has performed so far.
In other words, given that his is a predictive model, how does KenPom adjust for things like injuries, suspensions, residency, etc.
They do not, and that is what Dean Oliver at ESPN has tried to develop with their BPI - a rating that accounts for if you beat a team when they (or you) were missing a key player or - as in our case - how good will we be beginning with Luke being active. One reason I set up Value Add is I wanted to figure out how much a team changed with one player missing, since by my calculations a guy with a Value Add of 4.0 means your team if four points better with him than without him. However, Projected Value Add before they take the court is a guess - for example going by that Arizona State was getting a better player back than we were last night, and obviously the opposite was true.
I respect the opinion of KipsBayEagle and Bleuteaux, but I don't go as far as either of you.
I really like the ratings at kenpom, and the gamblers basically start with them and then make their adjustments to get very accurate spreads, so they are a good baseline but you have to make common sense adjustments.
I don't agree that RPI is better, because the problem there is if you beat the 300th best team by 1 point, it is the same as beating them by 40. Ken Pomeroy has said he played around with accounting for running up margins and just has not found something that worked, but it is a real problem when scoring 10 more points to win by 30 points instead of 20 points gets you an extra +10, whereas driving for a key bucket at the buzzer to chance a 1 point loss to a 1 point win is just worth +2.
I've played around with using the square root of the scoring margin, so that the points that win games with both team's starters on the court are much more important than what the scrubs do late in a blowout.
If you recorded square root margins, then you would get a +2 swing for that last second game-winning buck (changes -1 to +1 since the square root of 1 is 1), and you would only get 1 extra point for scoring 9 points in the last minute to go from a 16 point win (square root 4) to a 25 points win (square foot 5).
Quote from: Warrior Code on December 17, 2014, 11:52:40 AM
(http://express.ece.ucsb.edu/pictures/ucsb_aerial2.jpg)
Here you go, Jay Bee. She's a beaut.
My cousin went there. Now he's a New England surf dude.
Quote from: bamamarquettefan on December 17, 2014, 02:23:05 AM
If i had to look for one MINOR negative on a great night, the sloppy play the last couple of minutes kept Marquette from betting in the top 100 at www.kenpom.com. The problem with the system is all points, so beating Arizona State by 7 at home dropped them from 67 to 74th, while it improved MU from 111th to 103rd.
By comparison, when Georgia ran up a 31-point win on 244th rank, they moved from 69th to 50th. Keeping the margin around 15 would have pushed MU up well into the top 100.
Keeping the kempom down while winning is difficult to do, and I'm all for it. Let's sandbag opponents as much as possible. As long as you keep winning, you won't be able to keep the kenpom down for long.
Aww shucks. We aren't ranked anywhere in the top 1/3 of the NCAA's, but who really cares? We're not into conference play yet and nobody even makes it down to 100 while checking out these lists.
Quote from: MU Fan in Connecticut on December 17, 2014, 12:50:24 PM
My cousin went there. Now he's a New England surf dude.
I can see why. It's one of the schools I visited, along with UC San Diego, Irvine and I think Davis. Even though some of those campuses are like resorts, I chose to stay closer to home and stick with Glorious Alma Mater and never looked back.
Quote from: bamamarquettefan on December 17, 2014, 12:29:05 PM
Well played to protect the honor of 79warriors daughter! I believe we need a new rule for the blog that says asking for photos of someone's daughter is off limits!
yet we keep posting photos of opponents alumni porn stars--who are also daughters of somebody. We can't have it both ways.
Question for anybody: Are there any MU alumni that are porn stars? And please, you male MU alumni (you know who you are), don't respond by saying you are a porn star, unless you can verify it with live video action.
Quote from: seakm4 on December 17, 2014, 02:36:29 PM
Aww shucks. We aren't ranked anywhere in the top 1/3 of the NCAA's, but who really cares? We're not into conference play yet and nobody even makes it down to 100 while checking out these lists.
Actually we are in the top 1/3 of the NCAA. There are 352 Divsion 1 teams.
Quote from: bamamarquettefan on December 17, 2014, 02:23:05 AM
By comparison, when Georgia ran up a 31-point win on 244th rank, they moved from 69th to 50th. Keeping the margin around 15 would have pushed MU up well into the top 100.
Hey Bama,
What are you thinking when you say "well into the top 100"? Throwing some numbers around quickly, I can see maybe ~97th with 8 more points scored on offense last night... but didn't assume anything for a recency factor.. curious what Pyth you'd forecast with another 8 points...