What if you understand a particular sport at a deep and nuanced level, do your homework, and then use that to build a better model using machine learning????????????//

I'm thinking we should just take the online course right from the source

https://lagunita.stanford.edu/courses/HumanitiesSciences/StatLearning/Winter2016/about

Hastie and Tibshirani literally wrote the book on ML so why not learn it right from them. Since we're all coming from different backgrounds, we can then recommend places to circle back to learn more about programming, calculus (lol), and matrix operations. That seems more productive than getting bogged down in months of pre-req prep before even watching one lecture.

Next step, pwning Stanford douchebags at their local bars.



"You spent $150,000 on an education you could have got for free watching Hastie and Tibshirani's lectures online."

Yet you are 100% wrong on all counts.

I used to think this too and then I read The Logic of Sports Betting which explains it all in a very clear easy to read manner. Give it a look and let us know what you think.

Until the sportsbooks start using Deck Prism Sports, then we are all out of a job

how do i make a resume

I love when these meek eggheads come out of the woodwork and tell us all the REAL SCOOP behind sports betting. Listen guys, understand a sport on a DEEP and NUANCED level and DO YOUR HOMEWORK (lol). Oh and heres an old and outdated book that would barely have helped you years ago, that you wont even read cuz youre all so dumb! And heres some more wrong things too!

Mallachi you are delusional if you think opening lines have insane levels of precision and hard work by the books.

anything aside big american sports originated by a crew at cris and maybe Pin doesnt have much prep into it

Whatever guys. I have better things to do than argue with people on the internet. Best of luck.

Also, just to clarify, I am not saying that sports books cannot be beaten. All I am saying is that I am skeptical that someone who, by his own admission, really doesn't have much of a math background, can build a model using Excel or R or whatever that will give him an edge over sports books (or, for that matter other bettors who have a better background.) I think there is a big difference between a model that the Ed Millers of the world are capable of building and someone who took a business calculus class.

My other assumption is that sports books employ some people with strong quant backgrounds to help them set lines, particularly for sports like baseball and basketball. If this isn't the case, I would be highly surprised, but I don't know for a fact that they do.

So, whatever. Like I said, I have no interest in getting into a flame war, I bet a lot on MMA and do well and I occasionally bet baseball futures and football, but I'm not a professional sports bettor and if some of you are and have more insight on this then I do, then more power to you. That's it. If my original post came off as condescending and douchey, then I apologize.

lol the Ed Millers of the world. The people who won't bet +145 YES RSIF when the total is 8.5?

or the guy who in his book said the opposite of what you're saying, that books use old charts and not modern tech to set some of these lines

His book is pretty trash.

Look at SSB. In part of the book the author talks about push charts, goes through how to calculate them, idea of windowing vs no windowing, then has relevant examples of how this is useful (the classic basic strategy teasers). His book takes that approach where he gives you knowledge and through his examples you feel he understands how to use this knowledge.

Similar set up in King Yao's book which is heavily focused on using databases and he goes through relevant examples.

Ed Miller's book has random ideas, structured oddly, and then his examples don't leverage any of the ideas and then he basically throws the ideas to the wayside when discussing examples and takes very square approach in the analysis.(which is why I think it's appealing to "gambling twitter" because it pretends to be quantitative but his examples boil down to the "feeling" approach that appeals to many novice gamblers).

It's clearly a ploy to try to position himself as an expert in the field so he can sell his line service that makes unrealistic promises.

I reject the notion that one needs a glorified $160,000 piece of paper to have quant skills. I happen to hold a degree in math, so I know from experience how little it means (at the undergrad level, at least). I've learned way more about math on my own (and with the internet and books) than in college. And that's not counting the fact that most college math courses are similar to you learning on your own (minus the element of curiosity). The prof throws a bunch of stuff on the board at a rate that you don't have time to think about it; all you can do is transcribe it to your notebook. And then it's your job to go home and figure out what it meant. So then what the hell is the point of the lecture? Just send me a pdf of what you're writing on the board.

Acing a college course mostly means you didn't have distractions and you knew some things when it was time to take a test. Actually understanding a topic is something else. I'll take someone who's passionate about learning on their own time any day over some preppy with a degree who only learns when a test is coming up.

What a math degree mostly says about someone is that they were a person who decided to major in math. There is a correlation between that and being cut out for math, but the actual courses aren't what do it. Mostly what matters is the interest in math, which for some people doesn't come until after the age of 18.

I'm not saying every Joe Schmoe can watch some Khan Academy videos and start beating the lines, but the whole, "Pffff do you even MIT, brah?" is ridiculous.

DefNotRsig's anecdote about the Stats MSc who clicks buttons without actually knowing wtf is going on doesn't surprise me.

Lol this subforum has so much leveling per square inch and now that I'm able to recognize it, I love it

Just bough the ed miller book

i feel im going places now,after this purchase

https://soundcloud.com/pinnacle-podcast

Pretty positive in one of these with the Trading Director(you know the head guy of lines at Pinnacle) he virtually disproves/addresses everything you believe to be true Malachii. Also if books were that accurate then opening limits would be drastically higher and soft books wouldn't copy Pinny/BM once lines have more or less settled.
Also most quants that are as smart as you are assuming likely A. Make more money betting vs Pinny then actually working for them. B. Make more money betting for wallstreet than working for Pinny.

as usual we turn to ed miller



it's why i always encourage people to get started building a projection system even if it's just with a pen and paper (alan boston still uses a spiral bound notebook for his power ratings). i also warn people they probably won't win with it. the goal isn't to win out the gate but to enter the arena (what i call the lab), watching betting markets, and tinkering with projections. most of the time i'm posting plays here it's to motivate me to work on new stuff and watch the markets while i'm still developing stuff. the goal is to be working on and thinking about stuff and hope to achieve one of those key ed miller insights. you won't be able to do that unless you're actively working on it. many sports bettors start out like spanky chasing steam and playing middles then move onto betting their own stuff. although the truth is most winning bettors still chase steam because it works, it's relatively simple, building your own stuff is hard, and after all that work it's likely it will lose anyway. so you gotta really love sports analysis stuff and be hungry. having a good handle on math and stats will help but it's not a pre-req since it's developing that key insight that will drive success more than just ensambling random machine learning methods.

here is a great project to get started on.

read up on tom tango's marcel the monkey forecasting system and use it to create a player projection system for an obscure sport. this is a relatively simple projection system that still does amazingly well compared to projection systems that are an order of magnitude more complex. it will work for any sport although the specifics will vary. then use those player projections you created with an established game modeling framework (four factors for nba for example, again the specifics will vary this is how you learn) and boom you have a model that should beat polish backetbell. win at that for a bit and move upward and onward onto handball.

or stay focused on polish basketball. develop a network of pph shops to bet with and go from microstakes level volume to slightly more volume. over time you realize you can improve your player projection system in some ways specific to the sport that will double your roi and you are now the world's foremost gambling expert on polish basketball. the point is there are so many ways to go but nothing is possible until you start. sitting around saying, "bookmakers are too smart, i can never win" is a limiting belief and once you get rid of those you open up a whole world of possibilities.

last post. please don't mistake me for some super smart or great gambler. i am a "sucks player" as my four year old often reminds me. i lose all the time, i regularly want to quit, i don't model/bet full game stuff because my stuff isn't good enough to beat anything other than tiny markets. i bet low hanging fruit in player and team props and even then my results don't align with my projected edge because i can't build accurate models. but somehow i still win (sometimes) and you know what i enjoy this **** even though it's a frustrating waste of time.

THE ED MILLERS OF THE WORLD

start with the small tomg and build up

think what are important for baseball games

starting pitchers
hitters
bullpen

ok what is needed for starting pitchers

projected # of outs (pinny offers a prop on this now)
proj # of hits/walks they give up (prop on this too)
proj K's (prop)

hitters just the inverse (and props on all these too)

try to get close to market on those

then try to figure out bullpen (its impossible) and now you got a full game line

break it down into manageable chunks and put them together

my suggestion is start with basketball. it's the easiest to predict and more consistent (thats why in DFS you see like 10 points separating 1st and out of the money even with a full game slate). the hardest part of basketball is predicting minutes

or just open a bookmaker, how is this legal

Rafael Devers to have 1+ hits and BOS to Win -133

Rafael Devers to have 3+ hits + runs + RBIs and BOS to Win -109

Rafael Devers o2.5 HRR -204

can't bet under or no on anything lol

or, find a few sharps online. check them out as best as you can. get them PPH accounts through friends/networking. Get 25% of their weekly action, up or down. Get 25% of their losses from the bookie. Make money every week no matter if they win or lose. After two bad weeks in a row and after you pay the bookie no problem, ask for more credit and a % off their losses. He agrees to 5k more and 10% a week off their losses since he think he has a whale. (10% off means if they lose 1000 they only owe 900 but if they win 1000 they get 1000). Keep the 10% yourself or split it with your sharps. So now youre printing money every week. Have multiple deals like this with different people. Have sharps sharing accounts and get as much credit on each account as possible. Everytime its maxed out ask for more. Tie up money in futures and ask for it back in credit. If you still have it when the season ends youll get "free credit."

When the agent finally shuts you down, you throw one last hail mary. This can be dicey and is situation dependent, but you tell the agent the truth. Say its been professionals betting the whole time and do you see how good they can do? These guys are part of a syndicate (or some other bs.)You get the agent to turn dirty. If you can pull this off youre a true hustler. (I did this with the head of security for Godsmack a long time ago who looks exactly howd you expect him to look. That was one of the only times I was nervous going to meet someone). The agent can get almost unlimited accounts...he gets you a bunch and you cut him in. He'll want a ridiculous percentage cuz they always have ridiculous deals with their bosses (thats how bad most people are at betting). Youll have to explain to them that theyre making money on both sides and the sharps need a big piece for their "group" etc and you work out a deal. Now hes doing a little bit of a dangerous game...hes actually hoping his book loses money. If his boss finds out he could be in some trouble.

Now you have 10 plus accounts with sometimes up to 4 or 5 people with their hands in each one. Always be on the look out for anyone taking action and put out the word to anyone you know that youll give them cash for a bookie referral. Play lots of poker, especially home games, and ask around. Create spreadsheets for easy weekly tracking and be good with meeting up. This with a fun little part time job could make you good six figures a year.

do any of the accounts offer nhl shots on goal though?

that's why i started modeling NHL props, the goal is to bet full games next year.

the only sport i never tried to beat. it's on my bucket list before i die

Pinnacle copies 100% of the openers on their site from whatever literally any other book has first. I'm sure this requires many quants making hundreds of thousands of dollars.

no

Are you sure this person isn't spending sleepless nights modeling all the sports Pinnacle offers?