Forget the post just above this one (ahem). In fact, I teamed up with Interbank so that they boycott their system so that you can concentrate on analysing the data I'll be sending you shortly hehehhehehehhe(there's alotta datta in derr)
!Beering
Gnat... it is 22:36 here!
: )!
Yah beering here too and it's not even 17h yet (but it's friday and I deserve it and it tastes good and I worked hard all week and I have a headache and alot of testing ahead and ............ and I don't even NEED a reason
BTW, I just had a BIG light bulb flash madly on top of my head about 2 minutes ago. Since I now have 2 weeks worth of "leave the bots alone to do their thing" data (or shortly put "forward test data"), well by golly, I can backtest those 2 weeks right? Hehehhe so I'll be able to evaluate to what extent the backtests are (a) exactly the same (b) more than actual or (c) less than actual. Of course, the "correct" answer is (c) since it will mean, as cryten implied some time ago, that .......... MA MA MIA........... the forwards are superior to the backwards comprende? OK well, I need to TEST that before I get all excited hehehhhehehe. Shall post away these results as fast as I can
It was a tough 2nd week for the bots. Particulary eurjpy which played ping pong with the blue and pink dots almost every day. Will need to discipline that one.
Here are the balances:
Autobots: 5,002.98
Humanbots: 5,707.41
My hamburger bots are down the toilet pipe, not yet flushed but not in a healthy position/condition. But again I played with the units (0.1/1.0) and the parameters, shut them down, reopened ahhhhhhhh wasn't a good week. But we'll start fresh in 2 days (I need to eat see, so them grocery bots need to perform better!!!)
For a summary of the week's activity, see the zippies.
EDIT: nb: those aren't the zippies I'm talking about in my earlier post, just this week's summary of both accounts
well will you at least keep your fingers crossed? or in this case (since yer on da phone), your toes????
Are you talking about my "Edit" note (if so LOL) or about the results themselves?
These are extremely preliminary results. (But they're good) nb I'm talking here about the longer timeframes.
Although they had a so-so performance in the last 2 weeks, my forward bots "outdid" the backtests by (balances):
Forward tests: 5002.98$
Back tests: 4592.17$
by 410.81$
Now if I take that and divide it by 5000 (the initial balance) that's 8.22% better for the forwards (anyone feel free to correct my math here).
Soooooooooooooooo if this can be applied to the overall backtests I've done which cover the last 8 months and render (lemme go check my numbers, I've done some updating on those, actually downsized what I posted here a bit, but not much) ok so it's 2037 pips per month.
Let's see, 2037 + 8.22% = 2204 pips/month
You may want to further the agony by doing this:
2204 * 0.87 (mini lot) = 1,917$.48/month
or
2204 * 8.7 (standard lot) = 19,174.80$/month
and if we apply the same rules to the shorter time frames - which I also fine-tuned, assuming the same behaviour (which I'll check later, but for the sake of having agonizing fun here)
1393 + 8.22% = 1508 pips/month
1508 * 0.87 (mini lot) =1,311.96$/month
or
1508 * 8.7 (standard lot) = 13,119.60$/month
Assuming sufficient capital and MM (and all that) either:
mini lots
1,917.48$ + 1,311.96$ = 3,229.44$/month (I'll live with that)
or
standard lots
19,174.80$ + 13,119.60$ = 32,294.44$/month (I'll die with that)
By preliminary I mean I'll need to understand why the forwards have better results. First thing which comes to mind is my backtests are mostly 90% accurate, with NZDJPY at 85% and CADJPY at n/a%. So the 8.22% + kinda makes sense.
But I'll check/compare everything in detail tomorrow and post what I find.
Now, my eyes are really starting to look like the bloody ones cryten posted, so I'll take a BREAK and go have something to eat before my eyes rolls on the floor like marbles with me trying to catch them by sound
Gnat,
I forgot to mention it a thousand times today but I love your new fly/man av, verrrrry cool
... I hope that Plutonite is not pissed, as his avatar is the original Da Vinci man and I am using a violated version. oh dear! I would offer to join in but my record is about 55% quality with MT4
Hey Pipsy!
As you are online already... did you send me the Zip file at all? I need to send you something in return and I cannot as I am lacking your EMailadress!!
Did you fall out off bed?
Gmatt bzz... bzzz!!
I just pm'ed you and cry about that + my email addy. You'd need to be Einstein to figure my email address otherwise, right.......
(I was so bad in maths back in school ... only learned it later on U of Hamburg)
