Trading in excel

[kiwi_kid]

While I appreciate that many of you hold your cards close to your chest with regards to trading strategies, as a relative newbie to triggered betting, does anyone have any suggestions to where to start in building a trading sheet from within excel?

[PeteB]

I would say learn to trade manually first, until you have an approach that is reproduceable and has an edge. It doesn't need much of an edge if you can repeat it a lot. Use the power of compounding to grow your bank, rather than searching for the holy grail of systems.
(mind you it took me about a year and a half to realise the truth of this)

Once you know what your system is, then it's much easier to think about how to automate it.

Look at the tutorials on the sites for some of the other products - there are often videos of trading in progress - which of these approaches could be automated?

[kiwi_kid]

Thanks Pete,

I have been manually trading for a while now, with some success, although rightly or wrongly, my strategy is very much a combination of watching for trends, WOM and some gut feel, which when combined seem to work for me. WOM is one which I can write a spreadsheet for, no problems there, although as we all know WOM trading can be a bit hit and miss (am I right?), trends have a nasty habit of reversing themselves and my computer just doesn't understand my gut feelings like I do. Perhaps WOM trading could be a good place to start, and if so, how much Weight??

[PeteB]

Yeah the gut feel bit is tricky.... but there must be something that gives you that gut feel

As for how much weight - not very helpful , but you need to backtest - the good thing about BA is you can monitor lots of events at once at a good refresh rate - so sort out a sheet that can record an event (for example, if you trade single selections, then just copy the row for the selection to a separate area of the worksheet every time it changes. Record lots of events (and then lots more). Now try your weight on all the saved histories and see what works.
Also helps with the gut feel - look at where your first attempt at automation would place bets, and if it places one when your gut feel says not to, look at that in detail, and see if you can work out exactly what it is that your gut feel is picking up. Add that to the model and re-run - first see if it fixes the case you were looking at, and then see if it improves some other saved histories or makes them worse. I.e. are you fitting your approach to a particular situation, or making it more generic so it will work better in all situations.

As a guideline, the fewer rules your system has, the better - if you find you have rules that are only triggered once or twice in your saved histories, is difficult to draw any conclusions as to whether that rule is helping your system or not.
Obviously you also need some way of working out whether your bets would be matched when running against a saved history - if BA exported the traded volume per price (as per the betfair graph screen) this would be easy - as it is you have to be a bit cunning to do this.

(but yes I'd agree that trading purely on WOM is not using all the information that's available to you)

[kiwi_kid]

Thanks Pete, will give that a go