Parameters Explained?

[Emtaxx]

Hi all, I'm hoping some one can just confirm If i have the correct interpretation of the parameters...

Offset = Ticks away from your original bet

Stop Trigger??? = Is this where you set a tick mark i.e. 4 ticks away from ur original position as a signal to then activate the 'Place at (Ticks)'? This is the one I dont understand...

So.. 'Stop Trigger (Ticks) = 4
Place at (Ticks) = 10?

Fill or kill is just if it doesnt get matched it dissapears
and offset batches is just the amount divided in groups to the offset?

[Emtaxx]

Also to go with this. Why do my offset and stop loss bets dissapear? and not stay in the market?

[ShaunWhite]

Orders lapse when the market turns in play unless you specifically set them to persist or to take SP.

[Emtaxx]

Hi Shaun its in play. I belive whats causing the issue is the offset bets are also tied into the fill or kill bets...

TBH this automation stuff is really quite tricky lol... How long does it take on average to get something consistantly profitable !?

[Emtaxx]

I really hope someone could explain to me why 'offset bet with greening and trailing stop' doesn't actually green up? Or if it's meant to, why isn't it greening up for me?

I was under the impression that, as there is 'greening' in the script, that it should green up once one of the offsets has been met.

I am set up with offset ticks at 10, stop trigger ticks at 5 (and place ticks at 6), and fill or kill at 900 secs. In this scenario, if the stop trigger kicks in for the loss, why does it not 'green up' across the board?

I am at my wits end trying to figure this one out. Surely the whole point is for the stop loss to kick in and the system green up automatically? It appears as though the stop loss kicks in, but it doesn't green up and the liability on the (in this case horse racing) horse keeps going up or down. It's waiting for me to go green up the race. But why?

I hope I have made myself clear with this little query; I'm not very good at explaining things, lol.
Top
User avatar
andrejt
Posts: 153
Location: Slovenia
Contact: Contact andrejt
Sun Apr 12, 2015 7:24 pm

My problem was similar, so here's the explanation from support :

First the videos of my problem :

1) Manual betting, stop loss hit, greening works
https://www.youtube.com/watch?v=pnyQbLO4kkI

2) Auto betting, lay bet hit, greening works
https://www.youtube.com/watch?v=ZQrdSGBnPz0

3) Auto betting, stop loss hit, NO greening
https://www.youtube.com/watch?v=zvMajgFM5nI


And here's the explanation :

"OK those videos have helped highlight what you are seeing. With offset betting with greening and stop, the stake is set to green-up evenly if the trade is successful (i.e. it's stake is set based on the price of the offset rather than the stop)..... the unmatched bet can't be left in the market with two different stakes and betfair don't support a one-cancels-other order, so this is the nearest we can get to the ideal solution. If the stake had been set based on the price of the stop loss then it wouldn't properly green-up when the offset is successfully filled.

In the case of the manual stop loss used with the right-mouse click on the ladder, this works based on the current net stake and doesn't have an unmatched bet on offer to the market, so although it has the advantage of redding-up exactly when it is triggered, it requires the manual exit of a successful trade.

An alternative approach is to place all trades without greening so the same stake is used to enter and exit the trade regardless of the price and then green all selections just before the market is expected to turn inplay (perhaps five seconds before the official race start time)."

Hope this helps you a bit

I Suppose this gives me clues as to why it not been working............

I've been going round in loops

[Dallas]

On the 'Staking' tab of your main settings you have the option of choosing where the stop loss bet is calculated from (offset, trigger or placement) price, I'd suggest setting that to 'Placement'
https://www.betangel.com/user-guide/sta ... ility.html

Also remember Betfair will still match it at a better price if it's available when the order reaches the exchange so even when calculating it from the placement price if you get matched at a better price you won't have a fully even figure across all selections

[Emtaxx]

Thanks Dallas, This is key technical information I've been missing. I did not know there was a function for this. Hopefully it will help pull me closer into the greens.

[ShaunWhite]

TBH this automation stuff is really quite tricky lol... How long does it take on average to get something consistantly profitable !?

A bit longer than the time it takes to learn to trade manually, somewhere between a few months and never?

But the problem is there's a load of forum advice for manual traders and none about the process of automating. Lots of Guardian setup help but nothing about what goes before that, or how to gain information from the orders and matches it creates. Just looking at market level pl tells you almost nothing.

Insufficient sample size before making changes is the most common cause of failure.

Sample size :
Yes, I remember the general idea we discussed — it was something like this:

"The smaller the edge (ROI) you're trying to detect, the larger the sample size you need.

That rule of thumb comes from the logic that:

Standard deviation of ROI decreases with more bets (since variance scales with ).

A margin (edge) must be significantly larger than the noise to be statistically reliable.

To detect an edge at a useful confidence level (say 95%), the mean ROI must be several standard errors away from zero — which translates into large samples for small ROIs.

In simpler terms:

1% edge → 10k+ bets
0.5% edge → 40k+ bets
0.2% edge → hundreds of thousands, or you're in random noise territory

We also discussed that these numbers assume reasonably consistent bet sizing and variance. If you’re working with varying stakes or extreme variance, even more samples are needed.

Want me to re-derive the exact math behind that again for a specific ROI and confidence level?

It's this that makes finding a strategy tough if you're doing it in real time, and why data and sims are popular. They require a considerable amount of effort but while the manual guys are doing their apprentice year, you need to be doing something productive.

[Emtaxx]

insightful stuff as always Shaun.

I find you may have an profitable idea, however the tricky part is quantifying where the profit lies within that idea.

I'm aiming for low risk / high volume & High Risk / High volume scenarios to help me push over the variance in PnL's

[ShaunWhite]

I find you may have an profitable idea, however the tricky part is quantifying where the profit lies within that idea.

Totally agree — spotting a good idea’s one thing, but the real edge is in isolating where the actual value lives. That’s where most fall short.