Trading What I see !?

[goat68]

It was a classic Large Lay order triggering a large "drift" trend, at least bot got £9 on it !

[goat68]

I might try and figure out the traded/queued volume at 7.2 later, to see if that gives an indication of the case...
(blimey 1100 posts, I am prolific aren't I!)

[ShaunWhite]

I'll try and look at it later if I get time.

"classic Large Lay order triggering a large "drift" trend" ... sounds like you've got a contender for a 2nd strategy, if you can prove that's actually a thing.

If you do have spin-off ideas it's best to keep them seperate. If you try and make a single bot that's a grand unified theory it will just be over complicated and impossible to see what element is causing which profit/loss.

[goat68]

So more discrepancies on trading a market today. My backtest shows no match, but live trading matched. However examing the log closely, and the price was stable so it was fairly easy to check, I could see over the 55second fill&kill period, only £30 traded at the order price, and the order was £60 and there was approx 70 ahead in the queue, so I can't see how the live trading got it matched.
My thinking now is there's no real reliable way of backtesting bet fill, as the Betfair matching engine is very arbitrary and non-logical. Which throws my whole reliance on backtest of a strategy into jeopardy, or at least to be very wary...
Hmm, not good

[goat68]

mmm not going well.... Anbell, Goat might be slipping..........

[goat68]

So this week's increase staking has been very interesting so far, only down 1/3 of this week's Sainsburys order so far, so at this rate i'm going to be hungry next week!
It has shown issues with backtest accuracy of larger stakes on more illiquid market intervals, basically you can't!, trying to model for example Laying £30 on a 3rd fav at odds of 7.0, where there's currently £70 queued and £531 traded, who knows is 50ms later the £70 ahead of you gets cancelled, or 50ms before your order hits someone else's £200 gets in...and also the Betfair matched volume at a price doesn't seem to tally very well in such a market...

The best I can do is treat it as a model, for strategy comparison, but in real terms the only way of knowing is live testing with half reasonable stakes.
The question is do you go "bust" before finding that elusive profitable strategy...........

[goat68]

So just done a rough calculation of my investment in this since I started trading pre-race last July, currently comes to roughly:
- trading losses = -£350
- BetAngel subscription to now about = -£180
- various education costs = -£147
So current total about = -£680
I suppose not huge in the scheme of things for 9months

[goat68]

There's a huge variations of skills on this forum, but there's definitely a large number of profitable traders, just judging by the number of different usernames I see every week talking about there level of PC (doesn't annoy me one bit). Yet you also see newbies that are struggling to get a basic BA automation script to run, that took me about 5mins based on my background, yet 9months and i'm still digging a debt hole... does that extrapolate a newbie having difficulty to 9years..? probably not.

[ShaunWhite]

My thinking now is there's no real reliable way of backtesting bet fill, as the Betfair matching engine is very arbitrary and non-logical. Which throws my whole reliance on backtest of a strategy into jeopardy, or at least to be very wary...
Hmm, not good

It's not arbitary, the one you described today as not being matched on your system but matched on Betfair might well have been down to cross-matching.

[PapaShango]

So just done a rough calculation of my investment in this since I started trading pre-race last July, currently comes to roughly:
- trading losses = -£350
- BetAngel subscription to now about = -£180
- various education costs = -£147
So current total about = -£680
I suppose not huge in the scheme of things for 9months

Hi goat,

Have you ever thought about writing your own custom software ? You seem to know your way around a computer and I believe there are other forums and betfair groups that can help and assist you.

From what I can gather, most of the genuinely profitable automated players all seem to use their own software.

I think I’m correct in saying that ShaunWhite uses his own custom built software now too, and if you look at the inspirational graph that he posted on the credibility thread, his profits have reversed sharply and have been heading to the moon ever since.

Maybe something to consider ?

[wearthefoxhat]

So just done a rough calculation of my investment in this since I started trading pre-race last July, currently comes to roughly:
- trading losses = -£350
- BetAngel subscription to now about = -£180
- various education costs = -£147
So current total about = -£680
I suppose not huge in the scheme of things for 9months

"They" say a business can take 2/3 years before it begins to turn a profit.

third.png

On a business plan, the set up costs are usually weighted towards the 1st year. Most cashflows would then reflect less expense in year 2. Your post reflects that, especially the "learning" cycle and the course costs incurred.

The skillset of the CEO (you) will also determine which direction things go. You've clearly demonstrated you have a strong analytical approach, strong software knowledge. In business there are no such thing as failures, just lessons learned, same applies to trading.

One thing that worked for me, was to take some time out. What actually happens, is the mind resets as the intensity subsides. It doesn't have to be too long, but enough to help recharge, and go again.

[napshnap]

...
Have you ever thought about writing your own custom software ?
...

Good advice.
Investigating your own soft logs may be sometimes Colombo's level job, then others soft is even more so.

[goat68]

So just done a rough calculation of my investment in this since I started trading pre-race last July, currently comes to roughly:
- trading losses = -£350
- BetAngel subscription to now about = -£180
- various education costs = -£147
So current total about = -£680
I suppose not huge in the scheme of things for 9months

Hi goat,

Have you ever thought about writing your own custom software ? You seem to know your way around a computer and I believe there are other forums and betfair groups that can help and assist you.

From what I can gather, most of the genuinely profitable automated players all seem to use their own software.

I think I’m correct in saying that ShaunWhite uses his own custom built software now too, and if you look at the inspirational graph that he posted on the credibility thread, his profits have reversed sharply and have been heading to the moon ever since.

Maybe something to consider ?

So sort of yes, however I would most likely pickup betfairlightweight and flumine project as a starting point.
There is also the £300 BF API investment, which is small in the scheme of things.
I was trying to prove BA Guardian could be a way forward, but I am more and more seeing its limitations, and see it really seems just an automation aid to manual trading for most people.

[goat68]

Giving Peter credit, all I am doing through Guardian automation is trying to apply known sort of triggers/strategies to the right "shape" of market, and through analysis & backtest finding the right strategy & market shape that seems to work, basically along the lines of all Peter's videos

[Trader724]

You will get there, you just need time. It took me a year to find the solution to the partially matched bets.