Hi guys,
I have an idea which may be useful.. Looking for comments...
I was thinking maybe the members of FPA could put forward some ideas on trading statistics to collect that could be sent to FPA directly which might address some of the issues concerning price feeds, requotes, etc.
The idea would be similar to what a trade profiling website (name undisclosed) does - gives you a 'fake EA' that does not execute trades, but reports your trades and some info to their website program which gives you pretty graphs and stats on your trades and allows you to publish them on social media sites.
Perhaps we could band together and get some ideas for what kind of data to collect which could be used to build statistical databases of information on live trading. For instance, it would be fairly easy to develop an algorithm that looks at the difference in price per candles and over a certain threshold, marks it as a potential synthetic price spike, its obvious when synthetic ones occur, the dP/dT is huge. This information could be sent to FPA and collated with other users of the same broker, published on the FPA website, who might report the same thing. You could also vote whether you thinnk it was synthetic or natural.
I think this might be useful as it might weed out the 'false' reports of one or two users having trade difficulties claiming the whole broker is scamming, but also at the same time, if many users are reporting the same issues, and one user says no problems, it may still indicate there's a broker-side issue.
For instance if we collect info like requotes/trade errors automatically from many users, we may be able to generate graphs showing the time of day correlated with requotes. It may show areas where requoting is high (if correlated with extreme market activity). Or it may show that users trading large lots have more requotes, or certain pairs. Or we could collect and dump actual spreads, collect a high, low, and average. Average latency of the user's connection to the broker could be reported as well, so we could weed out people who experience connection issues but this might be because they have a bad internet connection.
This information would give people a realistic (not broker-front page marketing hype) of how the broker performs in an international environment, and users could even filter data from the broker with respect to geographic location, so they could see the 'quality of service' of users in your own country (some brokers may be excellent in most countries but poor in yours, does that make them bad?)
If i were to do 'black hat' thinking it may be possible that brokers can 'DDOS' themselves in a way by making multiple fake trading accounts and altering their statistics by having many fake users that appear to have good stats. However this could be weeded out by checking IP addresses and also linking the FPA EA to the FPA forum account.
Anyone have any comments, good or bad?
edit: I should point out that I collected some statistics on one of my accounts (broker name will not be revealed this is not a scam thread) after experiencing may requotes, and discovered a bad trend, to which when I informed my broker, i was offered money as compensation (my trades were positive, no need to compensate for any loss, in my opinion sounds like 'hush money')... so statistics can be powerful
I have an idea which may be useful.. Looking for comments...
I was thinking maybe the members of FPA could put forward some ideas on trading statistics to collect that could be sent to FPA directly which might address some of the issues concerning price feeds, requotes, etc.
The idea would be similar to what a trade profiling website (name undisclosed) does - gives you a 'fake EA' that does not execute trades, but reports your trades and some info to their website program which gives you pretty graphs and stats on your trades and allows you to publish them on social media sites.
Perhaps we could band together and get some ideas for what kind of data to collect which could be used to build statistical databases of information on live trading. For instance, it would be fairly easy to develop an algorithm that looks at the difference in price per candles and over a certain threshold, marks it as a potential synthetic price spike, its obvious when synthetic ones occur, the dP/dT is huge. This information could be sent to FPA and collated with other users of the same broker, published on the FPA website, who might report the same thing. You could also vote whether you thinnk it was synthetic or natural.
I think this might be useful as it might weed out the 'false' reports of one or two users having trade difficulties claiming the whole broker is scamming, but also at the same time, if many users are reporting the same issues, and one user says no problems, it may still indicate there's a broker-side issue.
For instance if we collect info like requotes/trade errors automatically from many users, we may be able to generate graphs showing the time of day correlated with requotes. It may show areas where requoting is high (if correlated with extreme market activity). Or it may show that users trading large lots have more requotes, or certain pairs. Or we could collect and dump actual spreads, collect a high, low, and average. Average latency of the user's connection to the broker could be reported as well, so we could weed out people who experience connection issues but this might be because they have a bad internet connection.
This information would give people a realistic (not broker-front page marketing hype) of how the broker performs in an international environment, and users could even filter data from the broker with respect to geographic location, so they could see the 'quality of service' of users in your own country (some brokers may be excellent in most countries but poor in yours, does that make them bad?)
If i were to do 'black hat' thinking it may be possible that brokers can 'DDOS' themselves in a way by making multiple fake trading accounts and altering their statistics by having many fake users that appear to have good stats. However this could be weeded out by checking IP addresses and also linking the FPA EA to the FPA forum account.
Anyone have any comments, good or bad?
edit: I should point out that I collected some statistics on one of my accounts (broker name will not be revealed this is not a scam thread) after experiencing may requotes, and discovered a bad trend, to which when I informed my broker, i was offered money as compensation (my trades were positive, no need to compensate for any loss, in my opinion sounds like 'hush money')... so statistics can be powerful
