Research and Math Skills - Help to understand how to build a model

[SolidDala]

Hi guys!

Been recent months been more and more interested in handicapping. And two words keeps coming back to me: Hard work and math skills! I now where to find the data and have great math skills and basic programming skills from my college years. Been using basic excel to make sheets over retrieved data and for my plays. I been reading in this corner how to start a model to generate variables to see if a bet has an edge, a value to place a bet. Recently got the SPSS 17.0. In my mind no model - no future in this field, right?

Question is this; Where can I find Tips and guidelines, maybe even tutorial on what to do with all this data and variables. The variables obviously changes depending on what kind of model, but just to get me started

The model's purpose is to find edge for homedogs in upcoming MLB with different scenarios, a linear regression seems the way to go. Think that is enough to start with and get a hang of it??

I am all about hard work, sometimes you just need to stop and ask for help to get on track

[uva3021]

what kind of data do you have or do you plan on working with

with bases everything has to be regressed to the starting pitcher, i have tried to create a formula for such using a variable n-1 (in theory) to predict expected w% as you move down the rotation

before you just concentrate on homedogs though, you must realize that the top 50% of teams always as a collective unit finish considerably high in units earned

good teams win, bad teams lose

[SolidDala]

been collecting a lot to nba from statfox, covers and teamrankings.

For the upcoming MLB season I am about to collect data, are above sites the best sites to collect? New at MLB and been reading to catch up and understanding fundamental principles and what variables to consider. This season is all about learning process for me and see trends and how to use it.

Mentioned homedogs, cause they're winning 44% or similar and heard that backing homedogs is the way to go cause of the price?? As you can see I been taking myself into deep water and need to learn how to swim again

Best way would perhaps to read a book for dummies on handicapping and on witch programs to use, how to use it and how to structure the variables data to alter outcomes. Should perhaps just start with a simple excel sheet for this mlb season and focus on a nba model since thats the sport I have come to know the best and been following all stats from and then after learning both modeling AND about the mlb trends etc start with mlb model

[uva3021]

i researched every season back to 2005 a while back, using statfox, and found teams that finish above .500 earn money

• 2009 – +76.2 Units
• 2008 – +9.33 Units
• 2007 – +36.3 Units
• 2006 – +31.5 Units
• 2005 – +40.6 Units

this is record on just betting every good team basically lol

[Wrecktangle]

Here's a tip refresh your probability and stat, especially Bayesian stat.

Bayes is very counterintuitive, it's what throws most folks in this biz.

[statnerds]

i researched every season back to 2005 a while back, using statfox, and found teams that finish above .500 earn money

• 2009 – +76.2 Units
• 2008 – +9.33 Units
• 2007 – +36.3 Units
• 2006 – +31.5 Units
• 2005 – +40.6 Units

this is record on just betting every good team basically lol

what about when they played each other?

[Meestermike]

Another data site you may wish to consider is http://www.baseball-reference.com/

[SolidDala]

Thanks a lot guys! Gonna check this stuff and site out. Great input on Bayesian stat! Remember that one from one of my statistic and probability courses a few years back, little bit rusty on that part, just been doing mechanical calculations lately, but have a book with a chapter of it, so time to polish

Currently backtracking totals in nba, ppg(hometeam) and ppga(ht) with ppg(awayteam) ppga(at); from this taking a rough expected avg total and check it with the real line. With this comes of course injuries, last games ppg and ppga. Anyone know if this is something to continue work on or if its just on of those 50% outcomes in the long run?

cheers

[durito]

Currently backtracking totals in nba, ppg(hometeam) and ppga(ht) with ppg(awayteam) ppga(at); from this taking a rough expected avg total and check it with the real line. With this comes of course injuries, last games ppg and ppga. Anyone know if this is something to continue work on or if its just on of those 50% outcomes in the long run?

cheers

This.

[suicidekings]

Currently backtracking totals in nba, ppg(hometeam) and ppga(ht) with ppg(awayteam) ppga(at); from this taking a rough expected avg total and check it with the real line. With this comes of course injuries, last games ppg and ppga. Anyone know if this is something to continue work on or if its just on of those 50% outcomes in the long run?

Not the best way of approximating lines, but it's a good starting point. You'll find that some of your lines will be right on with the available lines, and others will be way off, due to adjustments made by the books to balance betting. Sometimes there's a good reason for the adjustment and sometimes not. You need to apply (or at least be aware of) a team's current form and if it deviates from season stats as well to figure out some of the adjustments the books make.

That being said, regardless of how sharp your lines are, the market is a stronger indicator than your lines. Remember, the books have WAY better stats and methods than you.

[vyomguy]

the books have WAY better stats and methods than you.

Not true always.

[suicidekings]

Not true always.

Care to elaborate?

[IrishTim]

Care to elaborate?

This is something I've been thinking a lot about recently. If you have the knowledge to beat the sport, why would you ever go to work for a bookmaker? I don't know how much line makers specializing in Mexican basketball (just an example) make, but presumably you could make a lot more betting it than you could putting up lines for a book.

I think it's you vs. other sharps who specialize in that league, but the way I see it - the competition is for who gets down on the best number first, not vs. the bookmaker's "sharp" who makes the opening line.

In my opinion, Pinnacle doesn't really give a shit about their opening line on Mexican basketball. At $500 limits, it's cheaper for them to hire some shmuck to throw a dart and let the market fix it for them. If you have a good database and a reliable model, you could definitely beat them on the OLs, no?

[SolidDala]

That being said, regardless of how sharp your lines are, the market is a stronger indicator than your lines. Remember, the books have WAY better stats and methods than you.

I hear you, and agree completely with it at least with my knowledge. Had to start somewhere right, this way I learn more and understand outer variables inflections better and what to look for in the future. This game is some hard shit to master

[durito]

This is something I've been thinking a lot about recently. If you have the knowledge to beat the sport, why would you ever go to work for a bookmaker? I don't know how much line makers specializing in Mexican basketball (just an example) make, but presumably you could make a lot more betting it than you could putting up lines for a book.

I think it's you vs. other sharps who specialize in that league, but the way I see it - the competition is for who gets down on the best number first, not vs. the bookmaker's "sharp" who makes the opening line.

In my opinion, Pinnacle doesn't really give a shit about their opening line on Mexican basketball. At $500 limits, it's cheaper for them to hire some shmuck to throw a dart and let the market fix it for them. If you have a good database and a reliable model, you could definitely beat them on the OLs, no?

Pinny's mexican baseball limits are always circled down to $50.

[suicidekings]

This is something I've been thinking a lot about recently. If you have the knowledge to beat the sport, why would you ever go to work for a bookmaker? I don't know how much line makers specializing in Mexican basketball (just an example) make, but presumably you could make a lot more betting it than you could putting up lines for a book.

I think it's you vs. other sharps who specialize in that league, but the way I see it - the competition is for who gets down on the best number first, not vs. the bookmaker's "sharp" who makes the opening line.

In my opinion, Pinnacle doesn't really give a shit about their opening line on Mexican basketball. At $500 limits, it's cheaper for them to hire some shmuck to throw a dart and let the market fix it for them. If you have a good database and a reliable model, you could definitely beat them on the OLs, no?

If we're talking about Mexican basketball and other low limit or otherwise exotic games, then sure. The books offer them because there's always going to be someone that will put money down on them. Same reason online sportsbooks offer wagers as small as $1 = to draw a larger share of the total market. The lines are probably going to be sloppier than higher profile sports and a bettor that specializes could have a solid edge, but again, at lower limits. It's not worth it for the books to put a lot of resources into them. However, these games represent a small percentage of the books total action. They make their biggest money on NFL, NBA & MLB.

If you're one of the big, high-limit books, taking potentially hundreds of thousands (or millions) of dollars of action on a single game, the accuracy of the opening line is more important for exactly the reasons you listed. An opening line that balances market activity means less line movement, and less risk undertaken by the books. The expense of the highest quality information and more experienced linemakers is much more worthwhile to ensure their profit over the long term for these sports.

[MonkeyF0cker]

The model's purpose is to find edge for homedogs in upcoming MLB with different scenarios, a linear regression seems the way to go. Think that is enough to start with and get a hang of it??

Why/how would your model only find edges for home dogs? You need to do a LOT more research into regression modeling.

[Wrecktangle]

Why/how would your model only find edges for home dogs? You need to do a LOT more research into regression modeling.

I'm with Monkey on this one; why not just work your regression for all the games and see how each category (HF, HD, AF, AD) is doing?

[SolidDala]

It's a Deal! Will work on the regression part to see how both teams doing!

[IrishTim]

Pinny's mexican baseball limits are always circled down to $50.

If we're talking about Mexican basketball and other low limit or otherwise exotic games, then sure. The books offer them because there's always going to be someone that will put money down on them. Same reason online sportsbooks offer wagers as small as $1 = to draw a larger share of the total market. The lines are probably going to be sloppier than higher profile sports and a bettor that specializes could have a solid edge, but again, at lower limits. It's not worth it for the books to put a lot of resources into them. However, these games represent a small percentage of the books total action. They make their biggest money on NFL, NBA & MLB.

If you're one of the big, high-limit books, taking potentially hundreds of thousands (or millions) of dollars of action on a single game, the accuracy of the opening line is more important for exactly the reasons you listed. An opening line that balances market activity means less line movement, and less risk undertaken by the books. The expense of the highest quality information and more experienced linemakers is much more worthwhile to ensure their profit over the long term for these sports.

I was just using Mexican basketball as an example. At what point then (measured by volume, limits, whatever) do the small markets turn into the big ones and their opening lines become not beatable?

My initial assertion is still the same - if you have the knowledge to set sharp lines, you can make more betting them than whatever salary the books will pay you to set their lines. If it really is the sharpest of the sharps setting the opening lines, well then there isn't much hope for any of us to make money at this.

[durito]

The opening lines are certainly beatable in every sport, otherwise they wouldn't move. Your assertion is correct. They don't need a super accurate opening line, that's why they open with low limits.

[IrishTim]

Thanks durito, that's what I was trying to get at. I was using Mexican hoops as an extreme example, but it's cheaper for most of the books to open with low limits and make the market fix it for them to pay someone who is capable of creating sharp lines a salary attractive enough to bring him to the other side of the counter.

[Flying Dutchman]

Openers and closers are beatable. Unfortunately, if the books see someone beating the market with greater frequency, they'll react against it. That's why it's a market.

[suicidekings]

I was just using Mexican basketball as an example. At what point then (measured by volume, limits, whatever) do the small markets turn into the big ones and their opening lines become not beatable?

My initial assertion is still the same - if you have the knowledge to set sharp lines, you can make more betting them than whatever salary the books will pay you to set their lines. If it really is the sharpest of the sharps setting the opening lines, well then there isn't much hope for any of us to make money at this.

I do agree with this. You can attack the opening lines and succeed. I just think given the opportunity, I would much rather be the guy setting the lines than playing them.

As for the quality of the opening lines, I never said they weren't beatable in the bigger markets. Lower early limits are there for good reason. However, there's so much variability and uncertainty in player performance that even the sharpest lines are way off on a fairly regular basis.

[uva3021]

what about when they played each other?

noted,