*****New Football Thread*****
-
you are 2 for 2 today
nice -
i also went for this trade.
2-1 and the profit is for us. i greened up
lets find a other trade if there is a oportunity -
hedged
-
having a lay on gangwon
-
SHG in the Jeju United v Gangwon match. 0-1 at HT
-
SHG in the Suwon Bluewings v Suwon match. 0-1 at HT
-
I’ve went for a SHG in the Jeonbuk v Ulsan match. 0-0 at HT
-
For the weekend. Lay Under 1.5 in blue/Lay over 1.5 in purple:
-
I’ll be looking a SHG in these two matches tonight:
19:45 - Shamrock Rovers v Dundalk
19:45 - Derry City v Shelbourne
-
Utrecht 0 2 down to Rotterdam, Utrecht have won the last 5 matches between these teams. Have layed Sparta Rotterdam and a small back of Utrecht.
-
Anyone done LTD on cup games, is it a no go or work well?
-
SHG in the Landskrona v Trelleborg match. 0-1 at HT
-
So after 8 months of testing, Winning, Learning, failing, losing and starting again, like with Racing tomorrow is D-day.
I have decided to follow Tomacz ONLY......2k Bank and see how it goes.
My reasons are every strategy I have made up and followed has dips and troughs but the Tomacz 1.5 has been relentless in Profit making over this period, so if you cant beat it......Join it and don't look a gift horse in the mouth!
-
@Martin It seems to work for me, when I start a new strategy, I paper trade the first 80 selections, then calculate the Strike Rate and min/max odds required, then go live with small stakes, revising the odds every 50 selections if necessary, until I have hit the 385 mark and then if the strategy is still working I go fully live with a bank and % of bank staking dependent on the ROI and Drawdown..
As I say, I'm no statistician and it may not be technically correct but it hasn't done me any harm (yet!)
-
I'm already guaranteed another red month after the events of Sunday but hopefully this will stop the rot and begin a decent if quiet June. Lay Under 1.5:
-
@Greg-Mitchell interesting that! I like a number of 2000 for certainty on total strategy effectiveness (or as close to as can get) but 385 makes sense for odds filter!
-
Basically, to be 95% confident that your margin of error is 5% or less, you need a sample of 385.
If you want greater confidence and a lower margin of error, you need larger samples.
The above is largely used for survey data.
A better way to test betting/trading results is to test the p value (probability that your results are due to chance alone).
-
@Greg-Mitchell said in *****New Football Thread*****:
@Joseph-Henderson said in *****New Football Thread*****:
How many selections would you suggest collating before applying a min odds filter? Thanks
Statistically, if you have a strategy whose number of selections will be unlimited, then 385 is the most you should have to track to be 95% confident that they are representative of an unlimited number of selections.
Obviously this is quite a lot and could take some time (strategy dependent). When starting a new strategy, I tend to use 80 as the first stop, as this represents an 95% confidence for a sample number of 100 selections. Then I review every further 50 selections until I have a total of 385 and then stop changing things.
Whether this is right or wrong, I dunno, I'm sure some statistics whizz will disagree, but it works for me.
Why 385? what's the significance of this number? Curious
-
@Joseph-Henderson said in *****New Football Thread*****:
How many selections would you suggest collating before applying a min odds filter? Thanks
Statistically, if you have a strategy whose number of selections will be unlimited, then 385 is the most you should have to track to be 95% confident that they are representative of an unlimited number of selections.
Obviously this is quite a lot and could take some time (strategy dependent). When starting a new strategy, I tend to use 80 as the first stop, as this represents an 95% confidence for a sample number of 100 selections. Then I review every further 50 selections until I have a total of 385 and then stop changing things.
Whether this is right or wrong, I dunno, I'm sure some statistics whizz will disagree, but it works for me.
