Kolek planning to go pro
Kenpom had Indiana beating OSU today 67-65. This is a perfect example...I would ask why?? Ohio State is a better team. IDC where Indiana is playing them. Is OSU much better?? Probably not, but still better and for me, if it's close like that I will always go wit hthe better team. Tomorrow night he has TCU beating Kansas 75-74 @ TCU. Kansas wins that game..period. Anyone think Kansas isn't better then TCU even with all their issues right now?? That's two examples of why I have issues with "predictive" models.
Maybe..but not necessarily because of the anomolies i spoke of..what if the tean that lost by 40 was down 3 starters for whatever reason and had to play walk ons significant minutes? As I said there are so many things that can affect an outcome..I just hate "rewarding" a team for only losing by 1 as opposed to 40..because thats basically saying "good job you tried your best". But that kind of fits the world we live in now, whereas we reward teams/individuals for their great effort(a participation trophy/ribbon) even if they lost. Different world then I grew up in..i know that much..at the end of the day..a loss is a loss(or should be) and an ugly win should always be worth more than a "pretty" loss.
TAMUI do know, Newsie is right on you knowing ball.
Maybe? Wouldn't it just be easier (and more accurate) to say that it is probable that if Team A loses to Duke by 1 and Team B loses to Duke by 40 that Team A is the better team? Now it is just one game, so that is why systems take all the games into account, so if the team that lost by 40 was an anomaly game that will only be one data point in a season of 30-40 data points. Also, aren't you the guy who says that injuries are excuses and teams should still win in spite of them?As to the bolded, an ugly win is always worth more than a pretty loss. There is never a situation where it is better to lose than it is to win a game. These systems just reward "pretty" wins more than "ugly" wins and punish "ugly" losses more than "pretty" losses.
Predictive model was damn close. Said TCU by 1. Game just went to OT, in which case anything can happen, but predictive model was essentially off by 1 point.
Close enough isn't good enough...I said emphatically without a doubt KU would win..soooo...yup. It's easy...go with them ore talented team, and you will always be right more often then you are wrong. Not sure why any of you fall all over Pomeroy?? His "model" doesn't account for the eyeball test, which as I said, I trust my eyes.
I would take pomeroy over your eye test about 100% of the time if I was betting. Just saying.
So...someone was emphatic that KU would beat TCU even though the god Pomeroy(that everyone here worships) said TCU would win. Hmmm...maybe Guru is a little smarter than most of you think. Folks, watch enough basketball, and if you go with the better team, I can assure you you will be right much more than you will be wrong. That one was like taking candy from a baby...as an added bonus, i give you the next flaw pomeroy has..UW beating MSU. Michigan State wins that game. Period. You're welcome.
That's your prerogative..I'm trying to point out to alot of you that worship Pomeroy..his model isnt the be all end all. I can predict or use #'s for almost anything..but you can make #'s say anything you want them to say.
People seem to think #'s and model's outweigh talent. They don't and never will.
If you are able to judge talent(which I have a unique ability to do), then you will be further ahead of any predictive model there is.
Weren’t you the one that said St Johns stinks. What happened to your eye test when we lost to them twice?
So let me get this straight. The model predicted that at the end of 40 minutes, the margin of victory would be 1. ample of why the model doesn't work?[/b]His model said TCU would win..his model was wrong. I said Kansas would win(which btw many ppl scoffed at because of how bad Kansas was on the road..yadda yadda yadda). Talent rises to the top. Kansas is a more talented team than TCU..I will take the more talented team every time regardless of where its playedNo one has said KenPom is the end all, be all. It is one system with flaws. And no you can't make #s say anything you want to. KenPom has published his methodology for all to see. He doesn't manipulate the numbers to say certain things. The numbers cannot lie, they can only tell you what they are programmed to tell you. You just have to be able to understand what they are telling you.Numbers don't outweigh talent. Talent is measured by numbers.Prove it. Make your own predictions and compare them to other predictive models. If your unique ability is as good as you say it is, you will make a lot of money. I'd guess that a lot of MLB managers thought the same thing before all that moneyball nonsense.
the general consensus is Duke is the favorite to win it all..and thats NOT because some #'s say they should..it's simply because they are the most talented team in the country..period. Talent trumps everything else..period. To think otherwise is foolish and naive
His model said TCU would win..his model was wrong.
Your eye test had them needing OT to win a game you said they would win....period?Uhm, ok.
Sigh. No it didn't. The model said TCU had a 51% of winning. Which is basically a coin flip. KenPom doesn't pick winners of games, it gives probabilities on game.
Who won the game? I emphatically said KU would win the game..and they did. Yet you want to credit Pomeroy because his predictive model was one point off whoop-dee-doo he didn't pick the right team to win and that's what matters. Like last week Creighton took Nova to overtime.. did anyone honestly believe that Nova wasn't going to win that game of course they were going to they are better than Creighton.
K whatever..his MODEL gave TCU a better chance of winning that game than Kansas..his model was WRONG. There should never be a model that gives the Lesser talented team regardless of venue a better chance of winning then the more talented team ever that's what I'm saying. Numbers don't judge talent..case in point..Houston..there #'s are all pretty good..now maybe some of you are guillable enough to believe that "well the #'s say they are the 10th most talented team in the country..so they must be". All you have to do is watch them play and you will see that yes they are a good basketball team are they as talented as their numbers make them out to be hell no!
And the numbers say the should because they are the most talented team in the country. These numbers that you keep going on about are just one way of measuring talent.
Who do you think is more talented, Marquette or Villanova?If Villanova then what happened on Saturday?If Marquette then I am assuming you will be taking a loan out from the bank to bet on us when we play in Philly since location doesn’t matter. Please confirm. You will get a great return as we will undoubtably be underdogs
Who won the game? I emphatically said KU would win the game..and they did.