Econometrics is simply the application of statistical methods to financial time series. Things like dealing with autoregressive processes, box-jenkins forcasting, etc.
I'm not sure what the canonical econometrics approaches offer above standard statistical techniques like regression -- most of its complexity really is devoted to handling data with temporal dependency, and for most handicapping I think one is better of treating games as genuinely independent events.
I'm not sure what the canonical econometrics approaches offer above standard statistical techniques like regression -- most of its complexity really is devoted to handling data with temporal dependency, and for most handicapping I think one is better of treating games as genuinely independent events.
The analysis of sports forecasting : modeling parallels between sports gambling and financial markets /, Mallios, William S. 1935-
It's published by an academic press and written by a dutch guy. His english is OK but the style of writing is a little different and hard to follow sometimes.
It's published by an academic press and written by a dutch guy. His english is OK but the style of writing is a little different and hard to follow sometimes.
Sorry to take so long in answering but I don't get to this site very often. I could but since I started day and swing trading I just haven't bothered. I'll be back for the football season though. I'm really not into the wnba. I think the game is great but it interests me on par with arena and european football. I'll be back in a few days.
My guess is it would be modestly successful at best. What the problem is in sports is that you have teams going up and down constantly, versus the spread. Teams go on nice runs and the stats will show it and make you overvalue them, they regress to the mean and you get burned. If you want to do something like this first use your model to study which teams play closest to how you expect them to based on the model's predictions. Then using only those teams you can go ahead and try to beat the vig, but once again if you don't take into account historical numbers and create some semblance of knowing if a team is just better than they were or if they are just playing over their heads, well you won't do well. Computer models can help with this very subject, but you have to be asking them to tell you this, not predict the score of a particular game.
I think Wildbill is right - a general theory might be moderately successful with a lot of work, but your best bet would be to specialize in a specific segment of the field - isn't that what successful financial analysts tend to do - specialize in market segments and focus on specific niches in those particular markets?
In my experience, the more specific I am in applying a theory to sports betting, the higher my win percentage tends to be; it's when I get greedy and broaden my data base from specific to general that I lose.
Good luck!
In my experience, the more specific I am in applying a theory to sports betting, the higher my win percentage tends to be; it's when I get greedy and broaden my data base from specific to general that I lose.
Good luck!
I would say that most of the useful items from econometrics are already part of a decent gambler's common sense. Eg., Wong's sportsbetting book (with no pretense of novelty) talks about long you have to wait to see whether a trend is indeed a trend or just a fluke, the importance of using your model (gamblers call it a "system") to forecast beyond the historical sample on which it was based, etc.
picked it up the other day. Have no idea what this book is about, but am checking out a bunch of different books.
Anyone read it?
I'll probably pick through it, but would like to know if any parts I should esspecially pay attention to while reading. I am an econ major and have taken econometrics and found it pretty interesting, so I figure I'll like this no matter.
