Ideas are super nice buttttttttt !!!!!!

[lsafak]

Anybody really managed to make money by automation ? I don't think so.
I do am very curious. Whether I tweak or not the most supposedly profitable trades somehow does not trigger and when they do it does not match.
Manual trading may work depending on your patience but lost my belief on automarion for sure.

[Derek27]

No, all the auto guys on here are all losers. That's why people do it full time.

[gazuty]

See the section of guidance headed “ This is just gambling (I don't believe any of you are sucessful)” viewtopic.php?t=23699

Your post is the latest in a long line of scepticism - which is fine - particularly about automation.

Whether it is Shaun, LinusP or many, many others - it’s probably to our advantage that you (and others) don’t believe you can make money automated. Leaves more on the table. Thank you

I recommend you read all of the guidance post.

[ShaunWhite]

Anybody really managed to make money by automation ? I don't think so.

100% yes, but sadly it's not just a case if having an idea and tweaking it. To stand a good chance you have to take it much more seriously. Collect large amounts of data, get your stats maths and data analysis skills up to scratch, and then spend the necessary time on a model or method. Ideally you'd also have a way of replaying the same data on strategy variations before committing cash to the live market.

Trading is very competitive, and the amount of money that can be skimmed off as profit is relatively small compared to the number of participants. It's not just bingo or a fruit machine it's an established industry populated by people with decades of experience so you won't just rock up and compete without at least matching their effort. Although very much the exception, I know 4 or 5 people who regularly exceed £1m a year. You ask if it's possible, well it's not just possible it's the only way to make serious money unless you want to be tied to your desk for 60hrs a week.

You might also want to read this group which is just for people using automation, https://join.slack.com/t/betfairlightwe ... Rjz0ESIJkg or consider attending this for more first hand comments viewtopic.php?f=55&t=20313&start=20. You should also look at videos like this and others from The Betfair AU team https://www.youtube.com/watch?v=zBU5JA4 ... jpy_PBoZEZ.

It's totally possible, but turning your PC into an ATM is every bit as hard as you might expect it to be.

[ShaunWhite]

It's as good a place to put this as anywhere.... I was doing some trend based stuff, forgot the formulas and stumbled across this
https://www.socscistatistics.com/tests/ It's a social sciences resource really but it's got a set of statistics calculators/formulas for a variety of tests of significance, correlation, linear regressions, confidence intervals and the equally loved & hated T-Test, most of which are really useful if you're getting into data (and algos) or even just looking at your results more closely. I liked the way it's all shown step-by-step so it's easy to re-hash in Excel.

btw your "super nice butt" thread title is a real eyecatcher

[gazuty]

https://www.socscistatistics.com/tests/

Great work Shaun.

I happened to be reading this yesterday (as a reminder to myself about how to do it) - https://betterexplained.com/articles/an ... s-theorem/

[ShaunWhite]

I happened to be reading this yesterday (as a reminder to myself about how to do it) - https://betterexplained.com/articles/an ... s-theorem/

"What do you want to do when you leave school?"
"More school"

Screenshot_3.jpg

[Derek27]

It's as good a place to put this as anywhere.... I was doing some trend based stuff, forgot the formulas and stumbled across this
https://www.socscistatistics.com/tests/ It's a social sciences resource really but it's got a set of statistics calculators/formulas for a variety of tests of significance, correlation, linear regressions, confidence intervals and the equally loved & hated T-Test, most of which are really useful if you're getting into data (and algos) or even just looking at your results more closely. I liked the way it's all shown step-by-step so it's easy to re-hash in Excel.

btw your "super nice butt" thread title is a real eyecatcher

What a baptism of fire.

[wearthefoxhat]

Anybody really managed to make money by automation ? I don't think so.
I do am very curious. Whether I tweak or not the most supposedly profitable trades somehow does not trigger and when they do it does not match.
Manual trading may work depending on your patience but lost my belief on automarion for sure.

Those that do well with full automation have been mentioned and answered accordingly. With automation, there are subtle things that can go wrong that are out of your control;

1. Unreliable internet connection - buffering/low speeds.
2. Dodgy API from source. Described as sticky, crashing, not working.
3. Issues with computer set up.
4. Others finding similar edge(s) and front run your automation.

There are stories of bots malfunctioning at times. Even in the high finance markets, algos's go wrong big time, usually for billions at a time.

Overall though, if you can overcome and accept things can mess up at times, then it's definitely the way to go.

IMO, I prefer semi-automation. Allow a bot to enter the market based on variables..etc and then manage the exit dependent on what I see or hear. That way I have the decision making firmly in my hands and can manage profits/losses accordingly. Not everyone's cup of tea, but that's my personal preference.

[goat68]

Anybody really managed to make money by automation ? I don't think so.
I do am very curious. Whether I tweak or not the most supposedly profitable trades somehow does not trigger and when they do it does not match.
Manual trading may work depending on your patience but lost my belief on automarion for sure.

Automation "is hard"! I've been told by some of the guys Shaun refers to who make £1M... It has to be hard otherwise there would not be these profitable opportunities, anything "easy" gets snatched up and is not profitable long term... As Shaun says the overall margin is small, so to get something worthwhile it has to be hard to find...

I'm still plugging away, but the best takeaway I would say is "it has to be hard for a reason" ! (i've not found it yet )

[decomez6]

The suggestions in this thread is very helpfull and if you are looking for an introductory path this is how I would suggest you go ( I am a beginner )

Start here:
Quick intro to statistics
https://youtube.com/playlist?list=PLuj ... 1EKl4FM6

If you like what you see head here,
What to expect in sports modelling ,

https://youtube.com/c/SportsBettingTruth

And read this .
Sports modelling in excel by Andrew mark

https://www.google.co.uk/aclk?sa=L&ai= ... x6BAgWEAE

NB : a good grasp is of excel is essential. Plenty of recourses out there.


My take :

BA has a lot of recourses especially the servants section I find some gems that are well structured. So I suggest you go through the automation and truely exhaust your options before diversification ,all depending on the sports you are trading.
The more advanced players like shaunwhite , use other softwares and have a deapth of knowledge that requires more than a hobby approach to the science of data analytics , my suggestion is to keep your head in one place and be good at it . Take one branch and be as good as the top crop .
Eg . If it’s horse racing , settle on one thing eg handicaps when only a certain condition is met ,then investigate it with a statistical approach .
Do not ignore the DOGGIES ! They are good for testing ( more regular and can be profitable)

There are no short cuts to manual trading , it will make you a better algo trader .
A mixture of both worlds of manual and algo will allow you to keep track of the rare outliers and pounce on them when they come your way as well as give you a good chance to work with human psychology which is hard to automate.

[Euler]

I'm always confused and delighted with sceptics in equal proportion.

Confused because it's been proven beyond all doubt that sports trading is possible. Manual, automated, a bit of both. There is gold in them that hills.

But delighted as the fact that people can't see this or be prepared to put the effort in, proves why.

[lsafak]

Anybody really managed to make money by automation ? I don't think so.

100% yes, but sadly it's not just a case if having an idea and tweaking it. To stand a good chance you have to take it much more seriously. Collect large amounts of data, get your stats maths and data analysis skills up to scratch, and then spend the necessary time on a model or method. Ideally you'd also have a way of replaying the same data on strategy variations before committing cash to the live market.

Trading is very competitive, and the amount of money that can be skimmed off as profit is relatively small compared to the number of participants. It's not just bingo or a fruit machine it's an established industry populated by people with decades of experience so you won't just rock up and compete without at least matching their effort. Although very much the exception, I know 4 or 5 people who regularly exceed £1m a year. You ask if it's possible, well it's not just possible it's the only way to make serious money unless you want to be tied to your desk for 60hrs a week.

You might also want to read this group which is just for people using automation, https://join.slack.com/t/betfairlightwe ... Rjz0ESIJkg or consider attending this for more first hand comments viewtopic.php?f=55&t=20313&start=20. You should also look at videos like this and others from The Betfair AU team https://www.youtube.com/watch?v=zBU5JA4 ... jpy_PBoZEZ.

It's totally possible, but turning your PC into an ATM is every bit as hard as you might expect it to be.

Thanks a lot. Algo trading FX and crypto more than 10 years. Will reiterate my server and automation setup and keep on trying.

[ShaunWhite]

Algo trading FX and crypto more than 10 years.

Yet you thought it wasn't possible on sports? Perhaps that was just frustration. Every market everywhere from pork bellies to the grand national behaves along pretty much the same lines so look for the similarities rather than the differences.

[lsafak]

Algo trading FX and crypto more than 10 years.

Yet you thought it wasn't possible on sports? Perhaps that was just frustration. Every market everywhere from pork bellies to the grand national behaves along pretty much the same lines so look for the similarities rather than the differences.

Matching is the problem over here. Not much liquidity is frustrating. I think that's why also this market is kinda hard, honestly I still could not manage automation probably will let it go and stick to back/lay over tournament winners as usual.