|
|
Ken Pomeroy is Full of Beans
|
Ken Pomeroy is Full of Beans
BEVERLY HILLS (MI) -- Mark Twain once famously remarked that there are three kinds of lies:
Lies
****ed lies
Statistics
The biggest star in the college hoops stat-head firmament is Ken Pomeroy, proprietor of www.kenpom.com, considered by many to be Xanadu for CBB numbers freaks.
When you click on Dayton (currently ranked a mighty #104 as this is written) you get Pomeroy’s prediction for the entirety of the 2010-2011 Flyer campaign:
http://www.kenpom.com/team.php?team=Dayton
The lowlights:
-- 17-14 overall
-- 12-7 at home
-- 5-7 on the road
-- 9-6 in OOC games
-- 8-8 in A-10 play
-- A five game losing streak, followed closely by another short-end skein of six in a row
What a total crock of not-good that is right there, Ken.
It’s been said that the only way to fight fire is with fire, so the solitary means available to retort statistics is with more of the same. Regular readers of this space will never accuse me of being a stat freak, preferring to leave them to guys like frequent MB poster Figgie, who seems right at home with numbers. Nevertheless, here goes, thanks to the UD Media Guide:
-- The last time UD lost 14 games in a season was the semi-forgettable 2005-2006 slate; BG’s third at the helm when the team finished 14-17
-- You have to go back to OP’s first year of 1994-1995 to find a season where UD lost 7 games at home
-- 2005-2006 was also the last time the Flyers dropped 6 OOC tilts in one season
-- One must dig all the way back in the archives to 2009-2010 to find a season where UD was .500 in A-10 play
-- That 2005-2006 season was also the last time the Flyers dropped 6 games consecutively (actually 7)
-- Here’s the real kicker for me, tho:
You have to revisit the catacomb of the “dark years” of JOB and the abysmal 4-26 season in 1992-1993 to find a UD team that had two separate losing streaks, one composed of at least 5 and the other of at least 6 games in the same campaign.
So, basically what Mr. Pomeroy is predicting for BG’s Boys is that a perfect sh*t-storm of O’Brien-esque proportions will befall this year’s squad.
Not. Gonna. Happen.
Why not?
Two guys: CJ and CW
Chris Johnson gets it, and BG is challenging him to do more, according to this article from the DDN:
http://www.daytondailynews.com/dayto...n-1034015.html
Chris Wright simply will not allow this to happen, algorithms be ****ed. He didn’t come back for his senior season to go 17-14. If the Western Carolina game showed us anything it is that CW can will this team to victory all by himself.
The beauty of Pomeroy’s little system (for him, at least) is that he gets to adjust his guesstimates game-by-game as the season goes on so in essence he’s never wrong.
You gotta believe that Kenny boy will be doing more than his fair share of adjustin’ this year when it comes to the Flyers.
That’s it “From the Swamp.”
You can email me at: swampy@udpride.com
|
|
Article Tools |
|
|
|
|
The larger the sample size, the more accurate the prediction.
If, going into the tournament, I was to place money down on games based on predictions by the RPI, KenPom, or a college basketball fanatic, I'd go with KenPom every time.
You may not like what KenPom has to say at this point in the season about the Flyers, but really, the team hasn't given him anything, numbers-wise, to predict any differently.
|
|
|
And so the logorithm backpedaling corrections have begun:
UD up to #101 (from #104) and projected to be 18-13 (from 17-14). It'll jump again when they beat SHU on the road.
The problem as I see it is that Kenny is projecting UD to play the rest of the entire schedule exactly like the team that lost by 34 on the road to UC and that they will never, ever improve. And don't tell me that it's a game-by-game statistical projection and you can't add them up. If that's the case then why publish them all?
Great follow up story on ESPN the other night on the Eagles comeback vs. the NY Giants. When they were down by 21, their chances of winning were 1%. Scored a TD, up to 9%; scored another, went up again. Tied the game, up to something like 40%. (Why not 50% they had just as much of a chance to win, right?) Punter fails to kick OB, Eagles win.
Stick that in your logorithm and smoke it, Ken.
|
|
|
Swamp,
I'm not sure why the hate for Ken. He doesn't improve or factor in improvement for UD? So what, he doesn't do that for any other team, theres no way to accurately predict how mucy, if at all any team will improve, or worsen thru the course of the season.
|
|
|
Medford:
I guess I've always hated math and I'm taking it out on the teacher.
|
|
|
Originally Posted by Swampy Meadows
And so the logorithm backpedaling corrections have begun:
UD up to #101 (from #104) and projected to be 18-13 (from 17-14). It'll jump again when they beat SHU on the road.
The problem as I see it is that Kenny is projecting UD to play the rest of the entire schedule exactly like the team that lost by 34 on the road to UC and that they will never, ever improve. And don't tell me that it's a game-by-game statistical projection and you can't add them up. If that's the case then why publish them all?
Great follow up story on ESPN the other night on the Eagles comeback vs. the NY Giants. When they were down by 21, their chances of winning were 1%. Scored a TD, up to 9%; scored another, went up again. Tied the game, up to something like 40%. (Why not 50% they had just as much of a chance to win, right?) Punter fails to kick OB, Eagles win.
Stick that in your logorithm and smoke it, Ken.
|
No, when the Eagles tied it up, they were still the road team AND, more importantly, the Giants were going to have possession. The points is that of all of the previous NFL games played, teams in the exact situation the Eagles were in won 40% of the time.
Statistics are a beautiful thing if you understand them and what they're telling you.
|
|
|
Originally Posted by Swampy Meadows
And so the logorithm backpedaling corrections have begun:
UD up to #101 (from #104) and projected to be 18-13 (from 17-14). It'll jump again when they beat SHU on the road.
The problem as I see it is that Kenny is projecting UD to play the rest of the entire schedule exactly like the team that lost by 34 on the road to UC and that they will never, ever improve. And don't tell me that it's a game-by-game statistical projection and you can't add them up. If that's the case then why publish them all?
Great follow up story on ESPN the other night on the Eagles comeback vs. the NY Giants. When they were down by 21, their chances of winning were 1%. Scored a TD, up to 9%; scored another, went up again. Tied the game, up to something like 40%. (Why not 50% they had just as much of a chance to win, right?) Punter fails to kick OB, Eagles win.
Stick that in your logorithm and smoke it, Ken.
|
First of all, I think you mean algorithm, not logorithm. Secondly, when the Eagles were down 21 with 8 min left, their chance of winning was 1%. The fact that they ended up winning the game doesn't mean that the original prediction was wrong, it just means that only 1 out of 100 times a team will win given that situation, and the Eagles happened to be that 1. Also, of course the % will increase as they score and close the deficit. All of those numbers you cited above are purely situational. The fact that the Eagles ended up winning does not discredit those percentages in any way.
I guess what I'm trying to say with all this is I don't understand your argument against Pomeroy. He is simply using the raw numbers from what has happened so far this season. Is he somehow supposed to determine the potential for improvement for every team and then give it a numerical value in terms of wins and losses? From what I've seen in the past, his predictions are often pretty darn close. His system is a completely objective way of making predictions, and is one of the best out there. No one here is saying that the Flyers can't or won't improve as the season progresses, but from the play thus far his prediction probably isn't too far off.
|
|
|
I'd be interested in what his system would have our record at using the stats he has right now for all teams we have played and going back in time and 'replaying' the games we have already played. I would bet we'd be far from 9-3 actual record we have and would be more like 6-6. If so, you can take his future projections with the same grain of salt.
|
|
|
Originally Posted by SHQCKEY
I think you mean algorithm, not logorithm.
|
Logorithm...algorithm...I ain't got no rithm.
|
Last edited by Swampy Meadows; 12-22-2010 at 10:05 AM..
|
|
Originally Posted by Marysville Flyer
I'd be interested in what his system would have our record at using the stats he has right now for all teams we have played and going back in time and 'replaying' the games we have already played. I would bet we'd be far from 9-3 actual record we have and would be more like 6-6. If so, you can take his future projections with the same grain of salt.
|
That is roughly what his "LUCK" category quantifies. Actual vs expected results and UD is in the top 25 - meaning they have won more games than they statistical model would predict.
|
|
|
Now up to #93 with a projected record of 19-12. The one game that changed since the last revision? The home game vs. X was a loss before; now it's a W. The algorithm still has us losing to GM and NMU. I just don't see both of those happening.
|
|
|
Kenny boy continues to pedal backwards, as UD is now:
-- currently ranked #88
-- predicted to go 20-11
-- now seen as finishing 9-7 in A-10 play
The backpedaling will continue after Saturday afternoon's game vs. NMU and I'll be there to see it happen.
|
|
|
Is he backpeddling, or is UD just starting to gel after struggling mightly early in the year vs sub-mediocre squads? I think its more of the latter, and his numbers are reflecting the improvement in UD's play, I doubt he's going in and revising UD's numbers personally because the early data doesnt' match up with the later data.
|
|
|
Aww, c'mon Medford--you're ruining all my fun. I love to watch stat-heads go down in flames. Thank God they still have to play the games, instead of suffering a pre-determined fate at the hands of the almighty algorithm or logorithm or whatever the heck it is.
|
|
|
Kenny P update:
-- UD is now #83; he still doesn't get it
-- Overall record now projected to be 22-9
-- A-10 record to be 10-6
-- Game at UMass is now a W
He's not done backpedaling, trust me.
I exchanged PMs with WH of the A-10 MB the other day. He now sees UD as going 12-4.
Me?
I'm sticking with 14-2.
|
|
|
Article Tools |
|
Display Modes |
Linear Mode
|
Posting Rules
|
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
HTML code is Off
|
|
|
|