Today I looked at this NFL handicapping system and the logic behind it is flawed. First of all picking winners or losers vs the line would be a 50/50 game if one used a random methodology of selection. The Goldman experiment success depends on a majority of the 250 handicappers as not only being ineffectivewhichI define as being between 50% to 52.3% ( equal to or greater than random methodology but below profitability as a result of paying vig) but being very poor handicappers(inability to beat random picking methodology) which over the long run is as difficult to do as picking winners. As far as vegas fading public money is concerned I see vegas more along the line of ignoring public money or moving the line with it, I dont see vegas saying the public money is all over team A -7 points lets fade them and move the line to -6.5. Smart money can move the line as well as lopsided public action.
Goldman Experiment seems flawed
- Thread starter greg
- Start date
The thesis behind the Goldman experiment is very sound, in my opinion. It is not a new theory, but one in which there are many years of successful results going against the grain. In the dozen or so years I've been capping primarily on sentiment-based indicators, there have only been 2 or 3 in which Goldman-type systems haven't shown positive results for the year.
It caputures group sentiment upon which the market price is based, and historically the group sentiment is wrong. However, there are periods in which the crowd is often right. Early season is often one fo those. My fear of excessive reliance presently no such methodoligy RIGHT NOW is twofold: 1. Last year dogs ruled in the NFL, and most big money "sharp" bettors are predominantly dog players. 2. The popularity of chat boards such as this have enabled many of us to fancy ourselves now as wise (or at least wiser) guys. The fact that contrarian handicapping has worked so well, in some ways works against us until the riff-raff is shaken out. That is happening now.
To me the red flag is that the lines on games are now "sharper" than before, adjusted toward the dog to suck in the sharps. A reat example last week was the MIA/JAC game, which had no business being only a 3 point spread. The sentiment among touts and the public was as heavily MIA as any game I've seen in recent years, yet the line held at 3 until a late move to 4 because guys like me played JAC or laid off altogether. Yet, JAC should've covered anyway, so maybe we're just suffering from bad random events. As Nolan recently noted, the lines don't appear to be the problem.
The one that is the trap this week in this mode is KC/OAK. I'm praying all week. "Please don't make me bet Oakland." But I'll probably have to. The ones that make me want to barf are typically the best. Yet, I'll turn tail if I see too much chatter about how the Raiders really are a good fundamental football team.
Bottom line, I'm still a believer in such methods and when the worm turns it will be good again. The crowd is never right for a protracted period of time.
It caputures group sentiment upon which the market price is based, and historically the group sentiment is wrong. However, there are periods in which the crowd is often right. Early season is often one fo those. My fear of excessive reliance presently no such methodoligy RIGHT NOW is twofold: 1. Last year dogs ruled in the NFL, and most big money "sharp" bettors are predominantly dog players. 2. The popularity of chat boards such as this have enabled many of us to fancy ourselves now as wise (or at least wiser) guys. The fact that contrarian handicapping has worked so well, in some ways works against us until the riff-raff is shaken out. That is happening now.
To me the red flag is that the lines on games are now "sharper" than before, adjusted toward the dog to suck in the sharps. A reat example last week was the MIA/JAC game, which had no business being only a 3 point spread. The sentiment among touts and the public was as heavily MIA as any game I've seen in recent years, yet the line held at 3 until a late move to 4 because guys like me played JAC or laid off altogether. Yet, JAC should've covered anyway, so maybe we're just suffering from bad random events. As Nolan recently noted, the lines don't appear to be the problem.
The one that is the trap this week in this mode is KC/OAK. I'm praying all week. "Please don't make me bet Oakland." But I'll probably have to. The ones that make me want to barf are typically the best. Yet, I'll turn tail if I see too much chatter about how the Raiders really are a good fundamental football team.
Bottom line, I'm still a believer in such methods and when the worm turns it will be good again. The crowd is never right for a protracted period of time.
The crowd will not be right enough for a protracted period of time to beat the house however the crowd will not be wrong enough for a protracted period of time to beat the house. I would like to see documented results on simply fading the public(includinghow publc sentiment was measured) would beat the house. I maintain it is just as difficult to pick losers as it is to pick winners vs the line. I am not sure how one can gather what exactly the public sentiment on a game really is. A group of guys from a city or a message board may not correctly measure public sentiment. Line movement can result from public or sharp money or the right touts putting a big play out. Perhaps the nfl linemaker factors in public money on some of the openers maybe .5 pt but not enough to move a line off 3 or 7 and therefore creating a blindly profitable scenerio.
I don't have the kind of documentation you're looking for (Goldman may), but for starters, if your thesis was true, dogs wouldn't consistently cover at a >50% rate season in and season out. Also, not sure about actual documentation, but you would be stunned how often in office pools that 2/3 or more of the players will be lesss than 50% for the season and that whole pools will hit 45-47%. Thus, it IS easier for people to pick losers, but only when they're thinking they're picking winners. And that's why going with the extremes works.
Hey Greg!!!
Hope you are doing. Long time no see....I am looking forward to your college hoop plays this year.
Hope you are doing. Long time no see....I am looking forward to your college hoop plays this year.
Perhaps over the past 5 years nfl dogs may have covered 51% of the time. Never the less random selection will yield 50% success rate. So one can choose to focus on dogs presupposing that the public will continue to bet the favs and the favs will have an inflated number. In anycase there it is a far reach to suggest any group of people will pick losers at a rate over 52.4% over the long run. One may take the goldman groups picks and fade all the teams who they select and who happen to be favorites however does the goldman group a small sample of people represent the complete betting public?is the goldman a select group of habitual losers who bet mostly favorites.? Looking at the infinity nfl contest group I see them up approx. 160 games vs the number with plus 50% vs the number having a 2/1 ratio over less than50% vs the number quite amazing after 6 weeks. I have rarely seen office pools with the poor % you mention. In any case one will do fine fading office pool picks like that . I never thought making money on nfl could be so easy.
