Author Topic: How Reliable is TimeMachine and MarketFeeder Test Mode?  (Read 5088 times)

Tags:
  • All members
  • Posts: 118
Hello,

I'M making some interesting results in live test mode as well with Time Machine. But exactly how reliable are these tests? I'm asking that, because I repeated the test 3 times in the same day and it gave different P/L the 3 times.

  • All members
  • Posts: 169
Re: How Reliable is TimeMachine and MarketFeeder Test Mode?
« Reply #1 on: 31 May 2014, 22:12 »
I think are very reliable. For me its not important if the green up its a little diferent. I Suppose you are testing in play, so lime the odd move so quicly sometimes catch differents odds, so i think you must not worried about that

  • All members
  • Posts: 905
Re: How Reliable is TimeMachine and MarketFeeder Test Mode?
« Reply #2 on: 01 Jun 2014, 09:43 »
Time machine is more accurate than test mode as you are using real market data.

  • All members
  • Posts: 118
Re: How Reliable is TimeMachine and MarketFeeder Test Mode?
« Reply #3 on: 01 Jun 2014, 13:34 »
But then again, how the results can differ so much when I repeat the tests with the exact same triggers and settings?

  • All members
  • Posts: 905
Re: How Reliable is TimeMachine and MarketFeeder Test Mode?
« Reply #4 on: 01 Jun 2014, 14:05 »
Try using the same refresh rate at which the data are recorded - 0.5 sec

  • All members
  • Posts: 118
Re: How Reliable is TimeMachine and MarketFeeder Test Mode?
« Reply #5 on: 01 Jun 2014, 17:03 »
OK, now we are getting somewhere. One more thing... Is it normal to get differente results for the same race but testing with diferent speeds?

  • All members
  • Posts: 905
Re: How Reliable is TimeMachine and MarketFeeder Test Mode?
« Reply #6 on: 01 Jun 2014, 18:17 »
I haven't compared different refresh rates in TM, but I have always had the same results at 0.5 sec.

  • All members
  • Posts: 118
Re: How Reliable is TimeMachine and MarketFeeder Test Mode?
« Reply #7 on: 01 Jun 2014, 18:26 »
I don't mean that, I mean the speed control in time machine.

  • All members
  • Posts: 905
Re: How Reliable is TimeMachine and MarketFeeder Test Mode?
« Reply #8 on: 01 Jun 2014, 18:45 »
Don't know, I've always used the fastest.

  • All members
  • Posts: 118
Re: How Reliable is TimeMachine and MarketFeeder Test Mode?
« Reply #9 on: 02 Jun 2014, 12:16 »
I'm bummed out... exact same trigger, exact same refresh rate, exact same markets and the results differ...

  • Administrator
  • Posts: 8821
  • Gender: Female
*
Re: How Reliable is TimeMachine and MarketFeeder Test Mode?
« Reply #10 on: 02 Jun 2014, 12:29 »
I can explain the different results with the slightly different moments at which the market is refreshed when you restart it over and over again in Time Machine. Because it cannot be reproduced to a millisecond, as a CPU cannot be relied upon with that precision, you query the market's prices at different moments each time.

Like in real-time mode, two users betting in the same market with the same strategy can get different results, as it is impossible to get prices and post bets at exactly the same time.

Hope that makes sense.
Always try your triggers in Test Mode before switching to real money!

Follow us on Twitter.

Join our WhatsApp chat!

Присоединяйтесь к официальному Telegram-каналу!

  • All members
  • Posts: 118
Re: How Reliable is TimeMachine and MarketFeeder Test Mode?
« Reply #11 on: 02 Jun 2014, 16:53 »
That does make sense. So you recommend to run my tests when anything else is running on my computer? Also, does it improve the results if I set the priority of feeder7.exe t high in windows task manager?

  • All members
  • Posts: 905
Re: How Reliable is TimeMachine and MarketFeeder Test Mode?
« Reply #12 on: 02 Jun 2014, 18:12 »
Results will average out over an extended period of testing.  Don't be concerned about the repeatability over a few markets.

 

Please note, BetFair is seems to be currently OFFLINE