Setting a bot to follow the money pre-race

[Follow the Money]

Hi, I just started using Bet Angel and for the life of me can't seem to get my bot to follow betting trends.
Any help with this would be greatly appreciated.

Basically, I wanted the bot to trigger 5 mins before the race. Run a max of 8 times and trigger every 90 sec to check the price of the top 3 favorites to see if the price has moved in X amount of ticks from the previous cycle.

Now to try and set this up, in the General Tab I put Applies to selection (in position 1 (sorted by favortism) and in the rules Conditions. I put a Relative Odds Condition of:
Last Traded Price is Equal To the price of In Position 1 (Sorted In Order of Favortism) minus 4 Ticks

I then redid this rule for the 2nd line of betting and 3rd and 4th

The only other Conditions I put in were The Market is not in Play and on the 2nd-4th line of favorites I put a Fixed Odds condition that the position 1 favorite's Back Price is greater than 2.

I tried experimenting with the Last Traded Price Condition by changing the variables but couldn't get it to fire.

Am I setting this up wrong? Is there a way to set it so It checks the differnce in ticks from the last cycle and if that selection has come in X amount of ticks it triggers the bot.

If any one can explain to me how to set this up properly or why my automation isn't firing I would be much obliged.
Thanks Dave.

[mugsgame]

If you can achieve this you will be the richest individual to have ever struck a bet. It cannot be done. The main reason being the variables involved in pre race trading are too complex and in some cases random. You may have a limited amount of success a bit further out, but in the 5 minutes before a race? No Chance.

Sorry to bring you bad news, I am sure the other pros will agree

[LinusP]

You would have to use excel, a macro that records the price every set number of seconds and work off that. Not too hard to do if you have some experience with excel but I agree with Mugs in terms of it being unprofitable.

[PeterLe]

I think it would be extremely challenging to create a PreRace trading BOT from the automation rules in guardian. IMO you will have a much better chance using Excel. Be prepared for lots of sleepiness nights! and wear a helmet because you will be pulling your hair out!
I once spent so long in the study doing this, when I came out the dog barked at me because she didn't recognise me!

[EyePeaSea]

If you can achieve this you will be the richest individual to have ever struck a bet.

That's what I'm doing now, except I run this in-play, which is even more time critical and volatile. I do this with code behind Excel, storing the data in SQL for later analysis. However, unless I'm mistaken, that's a 5 year old Volvo outside my house, not a new Maserati. So I have to say that this level of complexity is achievable, even by people not making any real money.


Dave,

I've spent a lot of time looking for trends in BF. The key question that I'm trying to answer with my data analysis is "what is a trend?". Or more accurately "I've found a trend, but is it consistent enough to make more money that it loses?".

If you're suggesting betting on trends, then personally, I think that's a good approach to investigate. It's absolutely opposed to horse form analysis and relies in trust (of your system and code) over gut instinct and sometimes logic!

Coding a system to trade based on a trend is the (relatively) simple part. Identifying the parameters that constitute a trend is the hard part.

Not wanting to be negative, but realistically the number of permutations from a relatively simple formula can be huge.

The formula might be:

Odds change by X ticks
Odds range is ? (e.g. high, medium, low)
Duration of the change is >= X seconds <= Y seconds
Only interested in trading when the odds are >= X and <= Y
Only start to trade X seconds into in-play (or X seconds before in-play)

Those 5 parameters could equal 52,500,000 permutations (off the top of my head/calc.exe). Each permutation will result in a different set of trades, and a different P&L.

So, say you pick some figures as parameters for that formula. How do you test? How many days results do you track before you start using real money?

The point I'm trying to make is that your post may be a valid approach. But it's also possible that you could spend the next 10 years trying different permutations on the same theme, and never hit the right one. This is my dilema! But I'm here for the journey. If the final destination is nice, then that's just a bonus.


I've never looked at automation in BA, having always used my own code. But I struggle to believe that BA automation can give you the flexibility needed for this sort of approach.

Regards


Ian

[Follow the Money]

Hey guys thanks for your replies.
I'm not good with excel formulas but I wouldn't have thought it would be so difficult for someone who knows how to programme in excel to set it up. This is what I think it should look like.

Activate X mins pre start time until X mins after start time with a "Not in Play" rule.
On Activation take the value in cell A(Back Price) and place in Cell B.
In X(Re-arming time) amount of time take value in cell A(Back Price) and Place in Cell C.
If Value in Cell C is more than X% less than Value in Cell B, then activate a Back Bet at Best Market Price.
If Value in Cell C is not more than X% less than Value in Cell B then Place Value C in Cell B and Re-trigger Rule.

Thoughts?

[EyePeaSea]

Yup. Your formula or VBA just needs to watch a combination of G1 and F4 (that will tell you how long it is before the race or how far into the race you are).

As for the rest, yes, that's do-able. Off the top of my head, I can't think how to keep Shuffling the values ('temp' Cell B to 'temp' Cell C) with just a formula. Might be possible (as in, almost anything is possible)...

But it would be straight forward with VBA. Particular things to watch for are how you invoke the trade. You can get VBA to put the Odds, Value and command in, or, have those cells populated with a formula and get VBA to put the required values into other 'temp' cells. Shouldn't make a difference, but I find the second method more reliable.

Ian