Profits With Sports Betting and Speculation - Part Thrice! The Synthetic Intelligence Team

Hello and welcome to the most SIMPLEST intro you'll ever read.

So, in July 2024, I thought to myself "Hmmm. It would be kinda cool to create a horse racing simulator with ChatGPT. So I did, and created this Journey.
Profits with Speculation – Horse Racing AI Model Journey of Money Management

Then, about a year later, I started expanded sports, and even created an "AI Agent Team" of "Sports Betting Agents" modeled after a goofy staff like in the TV show "The Office"
Profits with Speculation: PART DEUX – AI, Sports Simulation, and Analytics

Then, there was Part 3. But I uploaded a zip file without a viruscan
[DELETED]
----------------------------------------
So, in each of those, I said that I WOULDN'T become a "tipster" etc. And well, look who won the poney.

And I can't. However, when I created that "AI Agent Team" it started to work. So, then there's always the idea that if one "builds" or "creates a method" with sports betting, 'why not just it for yourself'

Well, I am

----------------------------

Okay prologue over.

So, because of my background in intelligence and research, I guess, I'm just able to pick up on patterns easily. So I've decide to (hence why I'm here) make money online through building a school community around the framework. I'll simply share my results, since, I do it as lieasure, but also as a way to make money online.

____________

This journey will document how I've used AI to create a synthetic team and the building of that Skool community (Free to paid)

So, I'm just going to share some of my results from the last couple of days. Then I'll go more into the setup.

That was from a few moments ago, actually.








Okay, okay, you get the point.




So, yeah. It's kinda cool

Okay, get ready!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Hello.
Yesterday my record was 7-10, with an ROI of +46.24%

My method has less to do with the w/l record, and more to do with the specific $ invested, dependent on the odds, and game specifics.
So, the above is, absolutely a successful day. Some of the "losses", are actually just combination of the smaller bets.

So, with the "winning parlays" I post them to my Skool community, and will integrate with some other web 2.0s.

Here are some of those wins.
Again, I have no idea who these teams/people are. But, they do make for a good story that pays!


I also combined the bottom with a few other 'winning' legs - so the below compounded.



Next Steps.

Run reports on upcoming games with Structure
Post to Skool
Aim for higher roi% (Higher than 46.24%)

So, will this end up being some kind of service that everyone can use? Or how will it work?

tekitt said:

So, will this end up being some kind of service that everyone can use? Or how will it work?

Click to expand...

Yes, I'm going to make the Skool community membership based for the premium reports and other things.
I would like to make it available for others to use, but, I'm not a programmer, so just teaching the concepts (that I learn) is easier, since they are pretty helpful (searching for survivable/mispriced market lines)

Do you post your losses or only your wins

Inem2 said:

Do you post your losses or only your wins

Click to expand...

In the Skool community that I'm building, I post the wins, but AFTER I post the Forecast/Prediction. I have the AGENT RECORDS posted. So I don't post every win, or every loss - since it's really not about a "picks" service. I do, however, have the records posted, so the more accurate agents/patterns can be seen.

I know there are PLENTY of tipsters, out there, that post only wins. But, I'm not even presenting that type of "picks now!" service.

Speaking of which, here's a few more.
All I've done is used the creative writing aspect and combined it with the technology of AI, to create a synthetic freelance team, that specializes in sports analytics and predictive markets.

It's working.



Now, I won't watch the WNBA, ever - but, I'll def watch any footy team called the Whitecaps!

Now, about the amount - firstly, that's a post on reddit I made. All of my bets are usually $11.11, $5.55 and that sort of thing, so you'll know it's "uniquely" me.
Secondly, it's more about the data being accurate, rather than the amount.

Today I'll be:
-Organizing The Skool Page (There's great data and all, just not very organised)

-Fine-tuning the actual analytical process and translating that into a readable document (that's repeatable 2x-3x in a day).
I want this entire process to be under 3 hours.

-Aquiring new clients using new reddit method (for my other online business)

Thanks so much for reading!!!

edit*
Manual TDLR:
In the past I hired freelancers to help with my online marketing.
One day I thought "Hey I should create AI AGENTS to do that"

Then, I decided to "train" the AI AGENTS to analyze sports odds, outcomes etc.
Now I have a freelance team, that's synthetic - AI - and they just LOVE to analyze Sportbooks - that way, it helps us make money online.

Are you allowed to post your skool community here?

jakol said:

Are you allowed to post your skool community here?

Click to expand...

I think so. Usually, at least with journeys, it's the writer themselves who don't wish to share their exact stuff, because they don't want to "spoon feed"

However, I don't care. Firstly, there's a guy named GringoMonkey who spoon fed how to do Amazon Kindle and amounted for several years of awesome profit for me thanks to that.

Secondly, this really can't be replicated just from reading because it requires something unique (custom AI assisted research).

But all of that to say, yes, I'll share it within the next couple of updates as I get everything just a bit more tightened up.

Again, it started off just for fun, for me organizing my creative writing + analysis + digitalized investment + betting niche corner over here.

Since I'm here:



So. Once again, ignore the amount bet. Or don't.
The point is, that, the key is that the DATA is good. Whether that's an $11.11 bet, or an $111.11 investment is how the person choose to play the game.

Sometimes when I post on Reddit, they mock the small investment. Which is GREAT. That means my method is safe. You see, the key here, to a certain extent is quantity - but also quality - Let's be clear - I'm not making 100 bets a day.

About 10-14, at a time - and mostly 2-3 legs that are PROBABILITY BASED along with other factors NOT associated with public narrative. And, then I'm just sharing it on Skool combined with CREATIVE WRITING (comic strips, other ways to convey the message).

I currently have TWO more "pending" parlays (submitted by the same AGENT, I believe). One has no legs completed (scheduled for today), and the other, below, is awaiting one game completion:



Finally, I know "it's been done before", or "others do it this way" etc etc - but when I do things COMPLETELY DIFFERENTLY from everyone else, things seem to be much more profitable that way.

Hey, look at that UNDER 197, at -350 for the WNBA game.
My Agent (or model) actually suggested something like Under 184, which was something like +110.
See what I mean?
Maybe Ill watch an WNBA game one day. Well, probably not.

Now, I set the next slate of TARGET GAMES, for the SPORTS FORECASTING AGENTS, that were "trained"/"created" about a year or so ago.

Hi friends,
So, in the above post, I showed the example of that WNBA PROP, the UNDER 197, that CONVERTED, at a pretty lame -350, where you're "paying" the sportsbook to be right. Such a weird culture.

Anyway, I said something like "oh the model actually picked an under 187 or something...."

I want to show the proof. So, I decide to just whip the files into backtesting really quickly, just to see EXACTLY what that was:


There it is. As you can see the "AGENTS" are built in completely unique way.
Again. I know they are not real agents.
We know they are not making predictions.
We know that "they" are not really "them" - it's just LLMs, codes, 0s and 1s.

With that said, it works.

Hey, welcome! Wishing you all the best with your work.

Adult-HTML said:

Hey, welcome! Wishing you all the best with your work.

Click to expand...

Many thanks.
Fun fact. I've started this journey within the same month three years in a row (see first post).

I'll give the link to the skool, in my next update that will be a bit more structured and detailed. This will be a very interesting endeavour, because i've finally figured out Skool's very strange architecture!

Hello all.
So, I've organized the main 'pages' of my 'sports betting forecasting' page on Skool. The real interesting thing is that, using this platform allows me to organise this stuff in a way that I've been trying to do with Trello, for a while now. Plus, it allows me to get my creative writing out.

More details about the Synthetic Freelance Team That's Trained to Research Sports Analytics, Odds, Public Sentiment and Probabilities.

That's my raw definition of what this is, again. You know, like, how in wrestling, when they are "in wrestling", they pretend it's real. It's called kayfabe.
Well that's what you'll see me do, here. I'll refer to the agents as "the team" etc., him/her. Just know, again, I see these as LLM's that's able to string 0s and 1s together using code and, well, artificial intelligence.

But it's hella helpful.

So, let's take a look at what've I've done.

The Team
It's a team of about 20 agents. I made them last year using the framework used in old 80s and 90s business, character, and world building simulation games. You know, the VERY early days of AI. I created them, then started adding in more, and more, and more into their profiles. What this does is it prevents 'basic' halluciations. So, when I ask for a sports analysis - I'm not saying John Madden or Beckham has been resurrected through my work - but it's a huge difference from opening up an LLM and typing; "gimeee an eIght lEg ParLay with +4543 odds to win!"

So, because I'm the kind of person that like's that kinda thing, I created a Managing Director, then a (loose) structure of staff 'Agents' and such, all uniquely 'programmed'.



So, this is a screenshot of a pdf created with ChatGPT. In the "prompt" used to generated this 4 page pdf, I also had it emulate writing style, so it's completely not AI-content talk, filled with a tapestry of woven repeated no fluff fluff.

It's a very LONG prompt that I generated in 2023 when I was focused on beating AI content detectors.



Here's the skool page. It's free to join, and just mention that you're from BHW and I'll give you the premium access (free) so you can see these things that upload and such.
Sirius Sports Intelligence Agency

It's going to be very unique.

Additional

Medium.com - I have a member account there. I also have a newsletter, and a profile with very few followers. I may bring this account alive. I already have a related (same brand) newsletter set up

Monaids Ads - I'm spelling that wrong. But I may use them in the future.

Reddit - I post the more sizeable or "clever" win images. I also have a subbreddit which I believe I got, literally 1 member, from. I've not been very aggressive here, but I plan to.

Pikkit - I have a pikkit account. My stats have been pretty amazing since this system suddenly clicked into place. I don't check it much because I don't like the adverts and the many ways they try to suck you into making decisions they want you to make. But, as my stats and roi% increase, I'm going to use them to drive traffic to the Skool page.
_____________________________________________________________--

"Parlay Submissions"​

I "run" the agency like a sales call centre, where the reps must make "new account/sales" submissions that convert.

Here, the analogy is that, my Forecasting Team must make "parlay submissions" that convert.
Each agent record is track. This is part of backtesting/training. I've learned that the key isn't the AMOUNT of information that you give the AI (with whatever you're doing - but ESPECIALLY in analysis/research work), but it's the TYPE of data.

So, you'll notice the Skool page shall represent more of that flow, or structure, in concept, or creative presentation. It'll be pretty weird. But awesome.

I currently have 3-5 pending "bets", or SUBMISSIONS, with 1 currently having 2 out of 3 legs cleared at +118 with that game happening tomorrow at 2pm.
____________________________________________________________________

Good day
i said good day

Hello there.
So, continuing along this framework, here's how it works.
First, like yesterday,I have the AI AGENTS analyze a bunch of games. 24 hours before like. If need be, we'll do more analyses closer to game times. However, in my experience, the results are usually the same, whether research is done 24 hours, or 24 minutes before an event.

So after it's done, we have a workbook that was created that separates all researched games, separated by AGENT and other notes.


Now, we have a SLEW of reserached props/market lines, separated by AGENTS and other fields.

As a reminder, these AI AGENTS are 'trained' to analyze sports through simulation and other very abstract methods and references that I baked into the model to get some amazing results.

So, for one, this serves as:
How to (properly) do research with ChatGPT
How to make money with ChatGPT
How to use CHAGPT TO help with sports betting -
or my personal favourite:

How to turn CHATGPT into a complete synthetic freelance staff, trained in sports analytics, predictive markets, physics, geometry, research, and a whole bunch of other stuff, that frankly, it's not common for people to make agents out of.

What do I mean. Well, take AGENT Pollux Prime. My Agent. Created it about a year ago. Before it was to help me improve chess. So, I gave "him" "qualities" of a chess grandmaster.

Now, those qualities, SOMEHOW, SOMEWAY, make this AGENT / MODEL very good at creating sports analytics structures.

So, there.
Okay, that's step one.

The next step here is the fun part - it's the part that I do with a bowl of popcorn and a glass of mighty fine water.
I go through the data. See what's found, uncovered, exploits etc.

Then, we're going to take this data - or the data that I find, specifically, and turn it into a 4 page bulletin/newsletter type of deal with the key metrics, highlights etc. The above spreadsheet / screenshot is the result of these LOOOOOOOOONNNGGG, but very structured reports on each game/event.

Content creation part next.

I almost left out the best part of the document.
Now, in 2024, during the really cool GPT updates, I spent some time in a VERY VERY VERY RURAL area. So, even though I was busy, I still had ALOT of free time. So, I really "built" some deep models, agents, and research tools.

So, I say that to say, my model, or my 'agents', here in this screenshot is the SIRIUS MODEL. When that percentage is really high, that's a good deal that that market/prop is going to cash. Don't ask my why, but data doesn't lie.



Now, you gotta understand something. When I look at numbers like this, I get all excited. My mind sees in patterns. It's weird. But pretty awesome for stuff like this. Anyway, so, that's one reason why I'm aiming at providing this information to others who can really use it. The super positive intention of mine is to reduce gambling harm by providing some information people really into this stuff can use.

And, essentially, this is an extension of something I used to do for HOURS and HOURS and HOURS on end, during the ages of about 9-18. So this is like, an 'updated' version of my 1990s version of AI. "If this...than that." -*mind blown*