Fairbot Automation Conditions Setting Tips

For those of you who use Fairbot and in particular the new automation facility, I thought I'd bring your attention to some important considerations when configuring a rule.

You must be clear about what you are trying to do and how the market behaves and the way your rule will react to changes in that market. Specific care must be taken when creating any 'conditions' that will trigger your rule to place a bet or trade in the market.

For example, imagine you wanted to back the first runner that fell below odds of 5.0 say. If you refer to the image below you'll see that there's one of three options you can select which, when met, will trigger the rule.

Care must be taken when selecting either 'Back Price', 'Lay Price' or 'Last Traded Price'. This is particularly true if your new rule is going to be used in-running.

Given the large gaps that can rapidly appear and disappear in a fraction of a second, it is quite possible that the money on an outsider may disappear from the market momentarily thus bringing the first available back price below the trigger point of 5.0. If you've selected 'Back Price' as part of the condition's details then the rule will fire and your bet/trade will be placed on the outsider. This may not be what you intended.

In this particular case, it would be better to use 'Last Traded Price' or, though perhaps counter intuitive, 'Lay Price'. I prefer the latter as it guarantees that the runner is definitely trading below your trigger point rather than it being triggered by a large gap in the prices or a momentary spike/dip.

So, make sure you understand your market and be aware which conditional setting would be the better option.

Care must be taken though when testing these in simulation mode. The software can never know if your stake would have been taken in the real market and therefore it has to make some assumptions. In addition, the refresh rate in simulation mode appears to be throttled. If you don't believe me, switch simulation mode off and on close to the off with the refresh rate set to 0.2 secs and compare the rate of change of the numbers in the grid. This will also have an impact on your rule when testing in simulation mode.

After testing, make sure you test in real mode with small stakes until you are satisfied that your rule is behaving as you expected it to.

6 comments:

John Smith said...

Thank you I am testing the greenup and red up function at mo and this also is causing me problems with un matched bets in play 😁 keep up the good work

Alistair said...

It doesn't surprise me that you have unmatched bets in-play. There's a number of things that may be happening that's causing this and, I imagine, applies to any automation software not just Fairbot.

In simulation mode, the market MUST go through your 'bet' in order for the software to assume it has been matched. Even then, I've seen in simulation mode and not just when testing the automation feature, prices being traded in the market that are past my pretend money sitting on the ladder. This is especially true when, as I mentioned in my post, simulation mode refresh rates are apparently throttled.

Once you've tested the basic concept in simulation mode, you must carry out some live testing of this.

You may still have a problem in real mode that results in bets being unmatched. When the green up all function is activated either through a rule or you manually clicking on the button, the software submits the corresponding trade out bets on each runner, but those bets are still subject to a 1 second bet submission delay. It is quite feasible that the market has already gone past that trade out point on one or two runners resulting in unmatched bets.

To help mitigate against that, make sure you green up at 'current available price' and not 'best price'.

Also watch out if you a laying a number of runners with a view to trading out in-running.

Let's I lay 2-3 runners with a view to trading out in-running. I lay each for a liability of £50 which means if I let the the race run to it's conclusion, the most I could lose is LESS than £50. If one of the selections falls early in the race, one would think that is a good thing. However, the green up all will now show a red figure with a LARGER liability than if I let the race run. This shouldn't be as I'm guaranteed a successful lay on the fallen runner.

In this situation I've clicked on the green up all button when in simulation mode and sure enough, this much larger red is spread across all runners. For obvious reasons, I've not tried this in real mode.

When I've done this in simulation mode using automation, it appears that the automated green up all function doesn't fire but it needs further investigation to confirm.

I presume this phenomenon occurs because there's no price available on the other side of the market causing the green up all function to miscalculate the trade out totals. I don't know why the automated function wouldn't fire in this situation - though I don't discount some other setting I'd specified resulted in it's failure. I've yet to report this to Binteko so I have no idea if this is an issue for other users.

To get back to your original point though, yes, you will get the possibility of unmatched bets when trading either manually or automatically. Make sure you understand the settings within the automation feature to help mitigate against this. However, if you are trading just the one runner, I wouldn't use the green up all option at all. Instead, I'd advise you place a closing trade at the price that will give your desired profit. That way, your closing trade will be sitting in the market waiting to be matched rather than the green up all function trying to chase the price. This is much more difficult to do when trading multiple runners of course.

Alistair said...

Just to let readers know I've sent Binteko an email referencing this post.

John Smith said...

Thank you for all your help

Gee Dee said...

Hi Alistair.. Rather than assume could I ask you to clarify your comment regarding automatic green up...
(To help mitigate against that, make sure you green up at 'current available price' and not 'best price'.)
My version of Fairbot gives the following options 1. Use available price .. 2. Use most profitable price. I not sure how the variance has occurred in my version, but have they changed how the function operates or just the wording.
Finally in Automation Rules menu 'GENERAL' tab & sub menu EXECUTION, what is the significance of EXECUTION COUNT & WAIT TIME AFTER EXECUTION. If you are familiar with the feature. Many Thanks

Alistair said...

Gee Dee, it's just a difference in wording that I made in response to a previous comment. I wasn't looking at the Fairbot at the time to get the exact wording on the dialogbox.

EXECUTION COUNT simply tells the rule you've created to run that number of times with a delay of WAIT TIME AFTER EXECUTION between each time the rule fires.

If you have specific questions regarding the operation of the software I suggest you contact the developers directly at binteko.com. They are very helpful and usually quick to respond.