Author Topic: Variable results  (Read 5516 times)

Tags:
  • Élite
  • Posts: 431
*
Variable results
« on: 08 Jul 2014, 13:02 »
I used time machine (only using the free markets at present) for the first time recently.

I was very suprised to find that running a fixed trigger (in play betting) against the same 12 markets gives a different result most times.

I expected, as this is using pre-recorded data, that I would get the same result each time but the times that bets fire in and the prices matched and greened up are slightly different on each run.

At first I thought it was the difference between running them in "Normal" time mode and "Fast" time mode but then reaslised that even on one speed it varied.

Am I alone in this and does anyone have an explanation?

It makes the value of testing alternative triggers hugrly difficult to analyse as to do  it properly you wouldhave to average the results of each trigger change over many repeated runs.


  • All members
  • Posts: 829
Re: Variable results
« Reply #1 on: 08 Jul 2014, 14:56 »
Hi,
I ran a set of 38 tennis events twice, both times at fast speed
and got exactly the same result. Same number of bets
and same profit/loss.

I was wondering if there is anything in the pipeline to allow us to examine results ie: total bet amount total collect or P/L ?
R
My grandma taught me not to run downstairs and never back odds on.

  • All members
  • Posts: 905
Re: Variable results
« Reply #2 on: 08 Jul 2014, 15:52 »
TM simulates the real market, so you would expect varying results with TM in-play betting, just as different users would obtain different results in the real market even when using the same triggers at the same refresh rates.    This is simply due to the slightly different times by fractions of a second that bet orders will be requested or matched.

Just stop worrying about it when betting in-play.

  • Élite
  • Posts: 431
*
Re: Variable results
« Reply #3 on: 08 Jul 2014, 16:07 »
Thanks Rubold - I suspected it was that.

How do you mean stop worrying about it - do you mean just acceot the variation?

On that topic, has anyone experimented to find how similar Time Machine is to live betting on in-play horse markets.  I will want to be very cautious if I find a good trigger on Time Machine beofre putting it into practice.  Any thoughts?

  • All members
  • Posts: 905
Re: Variable results
« Reply #4 on: 08 Jul 2014, 17:05 »
Thanks Rubold - I suspected it was that.

How do you mean stop worrying about it - do you mean just acceot the variation?

On that topic, has anyone experimented to find how similar Time Machine is to live betting on in-play horse markets.  I will want to be very cautious if I find a good trigger on Time Machine beofre putting it into practice.  Any thoughts?

Yes just accept the variation as it is normal for in-play betting.  With TM you can see the variation, with live betting you are not aware of it, unless you bought the same market when available and ran it on TM.

You can rely on the TM results for any in-play trigger,  when averaged out over lot of markets.  This is just the same process you would need to follow when testing a trigger with live markets in test mode, except that TM will be more accurate than test mode (and quicker  ;D). 

A good advice is to employ a max refresh rate of 0.5 sec, as this is what is used to record the data.

  • Élite
  • Posts: 431
*
Re: Variable results
« Reply #5 on: 08 Jul 2014, 21:22 »
thanks Rubold

  • Administrator
  • Posts: 8826
  • Gender: Female
*
Re: Variable results
« Reply #6 on: 09 Jul 2014, 11:06 »
I second rubold in everything on this topic.
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!

 

Please note, BetFair is seems to be currently OFFLINE