Trading What I see !?

[goat68]

Just checked in on bot.... only -£5 down ! although it seems it lost over £4 on one race, taken out by what looks like a volatile whipsaw, so i'll let it off on that one, could easily have been +£4

The biggest hurdle in this trading game is being able to mentally endure the continual losing .... for months on end...
I do suspect you can get lucky, ie.you happen upon an "edge" or method early on, giving you a good mental boost.

[Trader Pat]

Luck has nothing to do with it.

Its about testing strategies and then more importantly analysing the data.

Are you analysing the data at the end of the day? Do you know why a losing trade is a losing trade?

[goat68]

Luck has nothing to do with it.

Its about testing strategies and then more importantly analysing the data.

Are you analysing the data at the end of the day? Do you know why a losing trade is a losing trade?

I have LOADS of logging in my bot, so yes. But to answer your question, why a losing trade lost, so the last 2 days mainly because my "stop out" logic was wrong following a tweak. Today so far apart from the whipsaw is better. The whipsaw, bot was on 2 trades, some big bet must have come in at 1min50sec, XM sent one Back trade from 5.5->7.6, the other Lay trade from 6.7->5.5, bot got out, and both actually finished at the close at 7.4 & 5.6, so they didn't mean return either.

[Trader724]

You may have a strategy to fight against everyone, but if you fight with yourself who would win?

[Kai]

The biggest hurdle in this trading game is being able to mentally endure the continual losing .... for months on end...
I do suspect you can get lucky, ie.you happen upon an "edge" or method early on, giving you a good mental boost.

Initial hurdle maybe, but that's the trading game and there's not much you can do about that. You either have some of that ability to dispel the initial illusion of an efficient market relatively quickly, or you have to brute-force it over a very long period.

Depending on your starting position and whether you have any financial trading background etc, I'd say getting there inside a year is relatively quick, but from what I can tell the average timeframe seems to be 2-3 years. If you can beat that average, you've probably done well.

You can get a bit lucky for sure, but unless you understand where that edge is coming from etc, it's likely that it won't last long and that you'll be unable to replace it or build on it further.

[Trader Pat]

I have LOADS of logging in my bot, so yes. But to answer your question, why a losing trade lost, so the last 2 days mainly because my "stop out" logic was wrong following a tweak. Today so far apart from the whipsaw is better. The whipsaw, bot was on 2 trades, some big bet must have come in at 1min50sec, XM sent one Back trade from 5.5->7.6, the other Lay trade from 6.7->5.5, bot got out, and both actually finished at the close at 7.4 & 5.6, so they didn't mean return either.

With the logging you should be able to figure out the losses and tweak as required. Like others have said it doesn't happen overnight so you have to be prepared for lots of ups and downs.

GL

[goat68]

Luck has nothing to do with it.

Its about testing strategies and then more importantly analysing the data.

Are you analysing the data at the end of the day? Do you know why a losing trade is a losing trade?

If I fully understood why the market moved like it does I think i'd be doing better... so "why a losing trade is losing"..? don't really know. I'm trying to define "models" of behaviour that might match markets i've watched, then coding Rules. But as far as I can see there is no reason why a horse goes from price X to price Y at the close, other than Tom Dick and Harry happen to decide to put their money on it or not.........
There is one aspect I would like to "code", but I can't see it is possible in Guardian, that is storing "state" across all markets,eg. recording market behaviour in races 1,2,3 of the day, which races 4,5,6,7,8 can utilize? If I could dump a Stored Value to a file maybe? v1.56 allows loading Stored Values, but i'd like to also go the other way....?

[goat68]

I think my "model" of what market "volume" means maybe wrong... I actually think it means nothing to future movements....but then what does?

[goat68]

bot's doing really badly....the markets seem to have turned against my model.
(i'm going to shut my "eyes")

[goat68]

Going to review my bot logs later, try to make it less complicated...
Simpler

[Tetras]

Luck has nothing to do with it.

Its about testing strategies and then more importantly analysing the data.

Are you analysing the data at the end of the day? Do you know why a losing trade is a losing trade?

If I fully understood why the market moved like it does I think i'd be doing better... so "why a losing trade is losing"..? don't really know. I'm trying to define "models" of behaviour that might match markets i've watched, then coding Rules. But as far as I can see there is no reason why a horse goes from price X to price Y at the close, other than Tom Dick and Harry happen to decide to put their money on it or not.........
There is one aspect I would like to "code", but I can't see it is possible in Guardian, that is storing "state" across all markets,eg. recording market behaviour in races 1,2,3 of the day, which races 4,5,6,7,8 can utilize? If I could dump a Stored Value to a file maybe? v1.56 allows loading Stored Values, but i'd like to also go the other way....?

I don't think it matters why the market moves, you just want predictable patterns, at least in terms of their profit vs risk potential. In your examples in the quote below, where's the edge in these trades exactly: how are you controlling risk, what is your bot trying to achieve? Like, I said before, I don't automate, but I can't see what predictable pattern you're taking advantage of there. If you've got a solid pattern based on patterns and characteristics for the market, then even the whipsaws are often in your favour, or even if they're not you can absorb them over time from the profits. I don't know exactly what you're trying to achieve, so perhaps you do have solid reasons for those trades.

"I have LOADS of logging in my bot, so yes. But to answer your question, why a losing trade lost, so the last 2 days mainly because my "stop out" logic was wrong following a tweak. Today so far apart from the whipsaw is better. The whipsaw, bot was on 2 trades, some big bet must have come in at 1min50sec, XM sent one Back trade from 5.5->7.6, the other Lay trade from 6.7->5.5, bot got out, and both actually finished at the close at 7.4 & 5.6, so they didn't mean return either."

[goat68]

Luck has nothing to do with it.

Its about testing strategies and then more importantly analysing the data.

Are you analysing the data at the end of the day? Do you know why a losing trade is a losing trade?

If I fully understood why the market moved like it does I think i'd be doing better... so "why a losing trade is losing"..? don't really know. I'm trying to define "models" of behaviour that might match markets i've watched, then coding Rules. But as far as I can see there is no reason why a horse goes from price X to price Y at the close, other than Tom Dick and Harry happen to decide to put their money on it or not.........
There is one aspect I would like to "code", but I can't see it is possible in Guardian, that is storing "state" across all markets,eg. recording market behaviour in races 1,2,3 of the day, which races 4,5,6,7,8 can utilize? If I could dump a Stored Value to a file maybe? v1.56 allows loading Stored Values, but i'd like to also go the other way....?

I don't think it matters why the market moves, you just want predictable patterns, at least in terms of their profit vs risk potential. In your examples in the quote below, where's the edge in these trades exactly: how are you controlling risk, what is your bot trying to achieve? Like, I said before, I don't automate, but I can't see what predictable pattern you're taking advantage of there. If you've got a solid pattern based on patterns and characteristics for the market, then even the whipsaws are often in your favour, or even if they're not you can absorb them over time from the profits. I don't know exactly what you're trying to achieve, so perhaps you do have solid reasons for those trades.

"I have LOADS of logging in my bot, so yes. But to answer your question, why a losing trade lost, so the last 2 days mainly because my "stop out" logic was wrong following a tweak. Today so far apart from the whipsaw is better. The whipsaw, bot was on 2 trades, some big bet must have come in at 1min50sec, XM sent one Back trade from 5.5->7.6, the other Lay trade from 6.7->5.5, bot got out, and both actually finished at the close at 7.4 & 5.6, so they didn't mean return either."

I do have a reason for the trades, without saying exactly, but based on patterns I think i've seen and behaviour modelled on how I think market behaviour happens... However, it's not working very well at the moment..! But in answer to your question, I do have a plan...just not a very good one!

[goat68]

That analysis took a while! But im feeling good about the simplified model and some good patterns i think I found.

[Anbell]

If I could dump a Stored Value to a file maybe? v1.56 allows loading Stored Values, but i'd like to also go the other way....?

You can do this.

[goat68]

If I could dump a Stored Value to a file maybe? v1.56 allows loading Stored Values, but i'd like to also go the other way....?

You can do this.

Ah can you? I'll have to download 1.56 and try. I only saw loading SVs in the release notes.
Thanks