1. #1
    BuddyBear
    Update your status
    BuddyBear's Avatar Become A Pro!
    Join Date: 08-10-05
    Posts: 7,233
    Betpoints: 4805

    Data Set Help!!!

    I know there are a lot of professionals and knowledge people on this board so I am hoping for some insights and good feedback. For this football season, I was thinking of using a data set. Nothing fancy, maybe Excell or even SPSS (which is a stats program I am familiar with) but when I thought about it, I realized I have no idea what variables I would use in my data sheets. And it seems like a lot of information I would place in my data sheets can easily be retrieved online as it is so I said why bother. But then I realized, to become a successful gambler you need to go above and beyond what the "recreational Joe" does and put some real effort.

    While I recognize that, I need help from those who you data sets in their gambling analysis. What are your variables? Do you use any statistics (i.e. regression), Do you gain any advantage that is not normally availabe by keeping your own data sets on teams, How do I best use data sets for football?

    No need to answer every question, or any for that matter, just thought I'd throw them out there. Any insights or observations would be great. Thanks!!!

  2. #2
    jjgold
    jjgold's Avatar Become A Pro!
    Join Date: 07-20-05
    Posts: 388,190
    Betpoints: 10

    I am no pro on this but it does seem the real advanced player uses homegrown numbers and stats and not web stats.

    All I do is track every type of wager I make on a server/web page and it does help.

    GL Guy

  3. #3
    SBR_John
    Wisky
    SBR_John's Avatar Become A Pro!
    Join Date: 07-12-05
    Posts: 16,471
    Betpoints: 42225

    Track Mudcat down and ask him. Someone told me if he had a choice of looking at beautiful women or his gambling Excel spreadsheets he'd be -340 to go for the Excel spreadsheets.

  4. #4
    Mudcat
    Mudcat's Avatar Become A Pro!
    Join Date: 07-21-05
    Posts: 9,287

    Quote Originally Posted by SBR_John
    Track Mudcat down and ask him. Someone told me if he had a choice of looking at beautiful women or his gambling Excel spreadsheets he'd be -340 to go for the Excel spreadsheets.
    I feel like I should be able to generate a good comeback for that but no luck so far. :confused: But don't get too comfortable John; there may be a great zinger coming your way any minute now.


    I saw this post and was actually a bit confused by it. I track all my wagering, research, banking etc. on Excel spreadsheets. I don't mind talking about that but I'm not sure that's on target for your question.

    Bear, are you talking about tracking a bunch of player/team statistics that you use towards your capping? That kind of thing?

  5. #5
    BuddyBear
    Update your status
    BuddyBear's Avatar Become A Pro!
    Join Date: 08-10-05
    Posts: 7,233
    Betpoints: 4805

    I track all my wagering, research, banking etc.

    Well it's a loaded question and a bit vague. I was hoping you could talk about your use of data sets in regards to your research. I keep track of my wagering history as well as banking details but it's the research part I am most concerned about. Specifically...what variables do you keep track off and then how do you decide on which teams to play.

    Specifically Mudcat, I know you've mentioned you play underdog money lines. How do you know which ones have the most value? I mean basically could you describe your research methods and process. I don't want you to go into great detail but i am interested in how you do it and what recommendations you have. Thanks.

  6. #6
    Mudcat
    Mudcat's Avatar Become A Pro!
    Join Date: 07-21-05
    Posts: 9,287

    Quote Originally Posted by BuddyBear
    Specifically Mudcat, I know you've mentioned you play underdog money lines. How do you know which ones have the most value? I mean basically could you describe your research methods and process.
    Sure.

    My methods begin with my basic knowledge of how sports betting works and then go to straight mathematics.

    The knowledge is that bettors like favorites (and, to a lesser extent OVERS). That is reflected in the lines. Lines don't reflect the expected outcome of games, they reflect bettor psychology. The books want to balance their action and they know that the psychology of bettors causes them to lean more towards big favorites than they really should. That often leads to lines being skewed and leaving good value on the underdog side.

    That basic fact is fundamental to almost everything I do.

    When I want to do a research project, I start with a theory. An example might be, I believe I would make money if I bet every moneyline dog of +200 or greater on every NCAAF game?

    I start a spreadsheet (I just use a basic Excel spreadsheet). For something like that it would be a very simple set-up. I just need to record the date, team (distinguishing HOME vs away) and the best line I can find (line shopping between strategic books is extremely important). After the games I record if the bet would have won or lost and I track the units.

    The important thing is to collect a lot of data. If a system is making money after 14 plays, that's not good enough. It's statistically insignificant. That's why I just shake my head when I see a capper talking about being 7-2 for their last 9 picks or something like that. Those kinds of statistics are meaningless. Something needs to work for at least a couple hundred picks before it can be considered a genuine pattern. So that's what I need in a research project. It has to be profitable after at least a couple hundred picks and then I can start putting real money behind it.

    That is my basic methodology. I collect a large sample of data. I break it down in some different ways (i.e. are HOME DOGS doing any differently than away dogs? Is it working better using lines of +400 or more rather than +200 or more). I look for what's making money long term.

    I hope to bring a lot of specifics about this stuff to this forum over the coming months. I'm testing some stuff right now on MLB alternate runlines that I plan to start transferring to the Sports forum soon.

    And I've got some solid stuff on NFL and NCAAF when the time comes.

    So keep your eyes open around here. It would be great to get your thoughts on some of the tests. I also intend to post real picks on theories that are already well-tested.

  7. #7
    Blackroc78
    Blackroc78's Avatar Become A Pro!
    Join Date: 08-15-11
    Posts: 1,189


  8. #8
    paranoyd androyd
    paranoyd androyd's Avatar SBR PRO
    Join Date: 10-01-11
    Posts: 6,459
    Betpoints: 134523

    very easy to do. just set up your excel sheet with separate columns for all raw data that you would input daily: dates, each team in contest, final score, spread, and ML. and then create separate formula-driven columns where the ML winner, spread winner, and total points winner are calculated from the raw data columns - just use simple IF statements here to avoid data entry errors.

    from there, you can run numerous pivot tables from this same data set capturing each 'angle' you want to examine and create betting trends from. these tables can be driven from team-based data, or any of the column data that you want to be the base/focal point.
    Last edited by paranoyd androyd; 02-18-12 at 09:17 PM.

  9. #9
    brewers7
    brewers7's Avatar Become A Pro!
    Join Date: 03-11-06
    Posts: 298
    Betpoints: 4441

    I agree with Mudcat...I've been doing this for a long time and am in the midst of adding data back to the 1981-82 season in the NBA...But the point is, I constantly ask myself, "Why did this team cover in this spot?"...Or "Why didn't they?"...

    I have come up with dozens upon hundreds, if not thousands of situational scenarios over the past 3 decades and started my own database, which has grown exponentially the last 7 years...

    I have attached a sample from my upgraded database (1998-99 season), but keep in mind the data is not accurate and most of it is blank, I just haven't gotten around to filling in the blanks for that season yet as I have most of the data, but haven't gotten around to filing in every season yet...But I have a template set up for every season and you are looking at a template in that sample...

    The NBA is the most situational sport of any, by far, bar none, to bet on, and knowing which situations are successful, that stand the test of time, is critical...I have a boatload of info including mileage travelled between games, the game where a new coach coached his first game during a midseason coaching change, games where starters or superstars did not play a game (for any reason) or their first game back from injury...Among a plethora of other data...

    I have accumulated this data because I, like Mudcat, have dozens, no, hundreds, if not thousands of theories I want to back-test and I want to add variables to the point where I can break a situation down to dozens of levels...

    Just a rough example off the top of my head...I would like to know:

    What happens when Team A is playing:

    A back-end of a back-to-back
    4th game in 5 nights
    On the Road
    Travelling more than 4000 miles during those 4 games
    with a .650 winning % or higher on the season
    With a .500 winning % or higher on the road for the season
    coming off a 10+ point loss
    shooting less than 35% from the field their last game
    who is in the top 5 in the league in FG% on offense
    who hasn't covered 3 in a row

    vs. Team B, who is playing:

    with 2 days rest
    At home
    who has travelled zero miles their previous 3 games
    with a winning % under .500 on the season
    with a home winning % of less than .550
    coming off a 20+ point win
    shooting over 55% their previous game
    who is in the bottom 5 in FG% allowed
    who have covered 3 straight

    I want to know what happened in this situational spot (not to mention hundreds of other situations) over the last 31 seasons...

    Anyway...

    I have theories...Many theories...And I want to test them...And I am hoping my data will be complete enough by the end of the summer to start back-testing as I have 1Q-2Q-3Q-4Q-1H-2H lines and totals going back 31 seasons (including this season)...Lots to test...

    So like Mudcat says: "start with a theory"...

    And then do your homework...And then come up with another theory, and another, etc...

    And then keep doing your homework until you have answers for your theories...

    I have answers to some of my theories, but not all of them and I am on a quest to have answers for ALL of them...

    GL...
    Attached Files
    Last edited by brewers7; 02-21-12 at 02:59 AM.

  10. #10
    probettor1
    probettor1's Avatar Become A Pro!
    Join Date: 04-22-11
    Posts: 1,985

    Do you really think that any one winning will give any help? No they won't. You can read all the books and threads out there and you will find nothing. Once you give away one tip to the public the bookies will change the lines every time the situation appears. That is why pro gamblers has to re-adapt their game constantly. Yes you can learn a lot with the analisys of the handicappers, which are free the day after the game but be aware that all those handicappers are overall losers and sell picks because they cant make a living betting.

Top