The -215 Rule... "My" new system

Forum: The Archives Page 2 of 7 «  1 2 3 4 5   »
Author: [The Archives] Topic: The -215 Rule... "My" new system
Matador PM Matador
Joined: Aug 2003
Posts: 18494
 
Matador
Participation Meter
MVP
Posted: 4/25/2006 2:28:44 PM
it would absolutely be irresponsible to say that there is zero correlation between betting patterns (read: line movements) and the eventual outcomes of games.  
 
Of course there will be correlations anywhere you chose to look. That's what all these goofy trend and line movement systems are based on. The real questions is: Are the correlations predictive of anything, or are they random?
KennyWoo PM KennyWoo
Joined: Apr 2005
Posts: 2510
 
KennyWoo
Participation Meter
Veteran
Posted: 4/25/2006 2:30:08 PM
I called my local. I asked him if I could lay -215 even for games that are less than -215. For instance, he let me take Hudson tonight for -215 instead of -130. The way I see it, I could bet practically the whole card this way. I think I could retire by this Friday.  
 
If your local was a Covers' reader there'd be no way he'd let you bet a bunch of games at -215, even if it was on KC or PIT. Your local: playing w/ fire...
vanzack PM vanzack
Joined: Nov 2001
Posts: 52546
 
vanzack
Participation Meter
Legend
Posted: 4/25/2006 2:30:29 PM
I'm not saying I necessarily agree with Panther's system, but without actually doing the research, it would absolutely be irresponsible to say that there is zero correlation between betting patterns (read: line movements) and the eventual outcomes of games  
 
Mod, as the old song goes - "call me irresponsible"....  
 
Do you know that there was a famous study done at Harvard that found out that Harvard had the highest percentage of Cheerios eaters of any major university in the country (THIS IS TRUE - NOT A JOKE). It had researches baffled because it was by a trememndous amount too.  
 
Now, does that mean that if I eat cheerios I have a better chance to get in to Harvard?  
 
CORRELATION NOT CAUSATION  
 
vanzack PM vanzack
Joined: Nov 2001
Posts: 52546
 
vanzack
Participation Meter
Legend
Posted: 4/25/2006 2:31:13 PM
Of course there will be correlations anywhere you chose to look. That's what all these goofy trend and line movement systems are based on. The real questions is: Are the correlations predictive of anything, or are they random?  
 
BINGO!  
 
wintesc PM wintesc
Joined: Apr 2006
Posts: 292
 
wintesc
Participation Meter
Prospect
Posted: 4/25/2006 2:52:36 PM
correlation is a statistical technique which can show whether and how strongly pairs of variables are related. one should NEVER assume that a correlation means that a change in one variable causes a change in another. however we're not trying to make a team win, but rather, trying to find the team that has the better odds of winning.  
 
to answer your question above, obviously it doesn't mean that you have a better chance of getting into harvard. however, if i showed you two students - one from harvard and one from bc - and you knew that student A liked cheerios, what school would you bet that student A attended?
yosef PM yosef
Joined: Nov 2005
Posts: 4253
 
yosef
Participation Meter
Veteran
Posted: 4/25/2006 2:54:13 PM
modthirtyb  
 
hook21 PM hook21
Joined: Dec 2003
Posts: 1056
 
hook21
Participation Meter
Veteran
Posted: 4/25/2006 2:57:52 PM
wasnt the cheerios effect a harvard based physics study which did not look at who ate what and where they went to school? and wouldnt that be a false attribution error, not correlation/causation?  
 
gotta love our useless undergrad degrees.  
hook21 PM hook21
Joined: Dec 2003
Posts: 1056
 
hook21
Participation Meter
Veteran
Posted: 4/25/2006 2:59:10 PM
and we all know causation has no use in mathmatical gambling independent event analysis.  
 
i just feel so smart today.  
 
 
gl to all.  
totalguy PM totalguy
Joined: Oct 2002
Posts: 7662
 
totalguy
Participation Meter
Captain
Posted: 4/25/2006 3:03:09 PM
Independent variables in terms of sports statistics or data are almost never highly correlated. They tend to occur more randomly than you would think. Rarely would you ever see an R-squared as high as even 10% between 2 independent sports variables, meaning that 10% of the variation in one variable is explained by changes in the other, which is pretty worthless.
HutchEmAll PM HutchEmAll
Joined: Jan 2004
Posts: 38673
 
HutchEmAll
Participation Meter
Legend
Posted: 4/25/2006 3:04:03 PM
I know everyone has pretty much beat this to death, but here is one final observation:  
 
loudon points out in his post that 13 of the 18 have covered at home with a line of -215.  
 
Let's crunch some numbers: That's 13-5. If you bet $215 to win $100 on these games, then you profit $225, or 2.25 units. Over 18 games having risked a total of $3,870. Hmmmmm.  
 
NOW, let's take a look at the numbers if the next 18 games results are different by JUST ONE GAME. In other words, the home favs of -215 go 12-6 instead of 13-5.  
 
In this case, you LOSE $90. Just ONE game difference. Now sure, it's possible that the next 18 games go 14-4....or 15-3....but is it likely? I doubt it.  
 
To me, that's an AWFUL lot of $ to risk (almost $4K) to win MAYBE 2 units on something that is absolutely unrelated to the game that is played on the field (the EXACT line).  
 
Lastly, it shouldn't surpise anyone that HOME favorites of -215 have a nice record overall. They SHOULD.  
 
 
About the only player that I can think of that would think a -215 line would have a special significance over a line of say -190 or -260 would be Pete Rose.  
 
Trends are great. We all use 'em. We all love 'em (probably a bit too much). You should respect them, but at the same time be wary of them. And also be wary of when a trend carries some weight and when it's completely born out of randomness (props to Matador).  
 
And with that, I'll shut up.  
 
van-  
 
wintesc PM wintesc
Joined: Apr 2006
Posts: 292
 
wintesc
Participation Meter
Prospect
Posted: 4/25/2006 3:07:34 PM
loudon, to answer your last comment, i doubt it would be worth the time/money for any "vegas" service to perfrom a statistical review along these lines. what vegas wants more than anything is balanced action. in a traditional -110 betting scenario, best case is 50% on each side and they take the vig. zero risk
wintesc PM wintesc
Joined: Apr 2006
Posts: 292
 
wintesc
Participation Meter
Prospect
Posted: 4/25/2006 3:09:49 PM
last comments, then i need to get back to work.  
 
let me first say that i really appreciate the intelligent conversation rather than the caps and exclamation points you see all over this forum
HutchEmAll PM HutchEmAll
Joined: Jan 2004
Posts: 38673
 
HutchEmAll
Participation Meter
Legend
Posted: 4/25/2006 3:10:25 PM
to answer your question above, obviously it doesn't mean that you have a better chance of getting into harvard. however, if i showed you two students - one from harvard and one from bc - and you knew that student A liked cheerios, what school would you bet that student A attended?  
 
mod-  
 
This is an excellent point. It would be like saying, "if I told you this team was a -215 home favorite, would you likely tell me that they won the game?"  
 
The answer, of course, is "yes." But in the cherios example, you're not paying $2.15 to win $1.00. Your "guess" as to what school they attended based on what they ate may top 50% and that is great. But in loudon's example you need to consistently be in the 70% range in order to make a profit.
vanzack PM vanzack
Joined: Nov 2001
Posts: 52546
 
vanzack
Participation Meter
Legend
Posted: 4/25/2006 3:10:50 PM
mod you seem like a well spoken smart guy. The discussion is always good.  
 
yosef PM yosef
Joined: Nov 2005
Posts: 4253
 
yosef
Participation Meter
Veteran
Posted: 4/25/2006 3:10:54 PM
hutch em all  
 
totalguy PM totalguy
Joined: Oct 2002
Posts: 7662
 
totalguy
Participation Meter
Captain
Posted: 4/25/2006 3:11:41 PM
I guarantee you that however complex a predictive model you create to try and spit out winners, PhD-level statisticians working for the books have a much much more complex and all-inclusive model and have had it in place for years if not decades.
yosef PM yosef
Joined: Nov 2005
Posts: 4253
 
yosef
Participation Meter
Veteran
Posted: 4/25/2006 3:11:58 PM
the smart guys thread  
 
Toples PM Toples
Joined: Mar 2005
Posts: 614
 
Toples
Participation Meter
Rookie
Posted: 4/25/2006 3:13:35 PM
bla bla  
 
my little brother Luka (he's 8) says that -215 rule is OK  
 
for a
vanzack PM vanzack
Joined: Nov 2001
Posts: 52546
 
vanzack
Participation Meter
Legend
Posted: 4/25/2006 3:15:12 PM
I guarantee you that however complex a predictive model you create to try and spit out winners, PhD-level statisticians working for the books have a much much more complex and all-inclusive model and have had it in place for years if not decades.  
 
Uhhhh yeah, but do they know about the 215 rule?  
 
wintesc PM wintesc
Joined: Apr 2006
Posts: 292
 
wintesc
Participation Meter
Prospect
Posted: 4/25/2006 3:18:41 PM
totalguy, the books don't need that kind of research. the books don't want to bet against you. they want you to bet against someone else. that's why the lines move, because in certain instances, it makes more sense for them to take a cent or two hit to make the "other" side of the bet more attractive
wintesc PM wintesc
Joined: Apr 2006
Posts: 292
 
wintesc
Participation Meter
Prospect
Posted: 4/25/2006 3:21:10 PM
you are on the right path thinking about PhD-level statisticians researching this stuff though. it's a specific field of economics that's gaining more and more steam, especially with the recent success of Steven Levitt's book Freakonomics
totalguy PM totalguy
Joined: Oct 2002
Posts: 7662
 
totalguy
Participation Meter
Captain
Posted: 4/25/2006 3:22:17 PM
They do need that kind of research in terms of knowing what the true ML is. Smart line makers know the true ML in practically every instance and bury it in between either side of their opening lines.
yosef PM yosef
Joined: Nov 2005
Posts: 4253
 
yosef
Participation Meter
Veteran
Posted: 4/25/2006 3:24:03 PM
that's what i say modthirtyb because it just makes too much sense for vegas to take a steady profit risk-free. think about it...would't you like your bankroll to go up 5 or 10% every day no matter what the outcomes were? but soon enough handy and others will be around to talk about how the books set trap lines to draw lopsided action. isn't is just possible that that goes on only very rarely if at all (keep ya on your toes), and that the books just want your vig day in and day out? it's a no-brainer.
totalguy PM totalguy
Joined: Oct 2002
Posts: 7662
 
totalguy
Participation Meter
Captain
Posted: 4/25/2006 3:34:26 PM
Michael Perry BetWWTS bookmaker:  
 
"As an oddsmaker, I occasionally hear football bettors proudly proclaim that they’re not numbers’ guys, meaning they don’t put a lot weight on stats.  
 
I hear similar statements made by football pundits all the time on the radio and television.  
 
I’m not sure why guys make these sorts of comments. Perhaps it’s because they failed math in high school and are afraid of numbers. Or maybe it’s because they feel they’re so knowledgeable about the game that they believe they can handicap a team or game just by watching tape.  
 
Whatever the case, those who foolishly snub their noses at the numbers are good for the bookmaking business, believe me.  
 
But the reality is, the numbers don’t lie. In fact, one influential linesmaker once confessed to me that he had stopped watching games and crafted his odds strictly through statistical analysis. And this is the same guy often credited with releasing the opening line in Las Vegas!  
 
The wise guys and the betting syndicates that give bookmakers panic attacks every week are also numbers’ guys. Typically they input stats into complex computer programs that spit out picks, which all too often win.  
 
While few have access to these computer programs, most have access to in-depth NFL statistics and trends that are freely available on the Internet. Thorough analysis of these numbers will arm even novice bettors with more educated picks."  
vanzack PM vanzack
Joined: Nov 2001
Posts: 52546
 
vanzack
Participation Meter
Legend
Posted: 4/25/2006 3:38:11 PM
Totalguy - thanks for posting that.  
 
The key, once again, is finding what stats mean causation. Thats it in a nutshell.  
 
Some are important, but all to often the fatal flaw of the human mind is to make correlations that are unimportant and irrellevant.  
 
The key is knowing the difference.  
 
GL  
 
 
 
PS - I think the guy in the article was directly referring to the 215 rule. You need to read between the lines, he knows this rule getting out to the massses would be devastating to the books.  
 
Forum: The Archives Page 2 of 7 «  1 2 3 4 5   »
You have entered the forum as a GUEST. 
You must login/register to post or reply.