Doubts over long-term profitability might be resolved soon

about bloody time 2! have u tried ameoba for back testing

I've used Amoeba, but this is limited to files that you have saved by running the bot or downloaded from another user, so is limited for backtesting. If x-feeder can pull it off, then it will be the ultimate tool - no more live testing to see if something works.

Sounds good but I think I know the answer already.

All my xmas's would come at once if I could backtest a trig against thousands upon thousands of games. I never appreciated how important a feature this would be until I started using amoeba. Trigs that had promise would fail and others that failed short term despite being good on paper proved profitable 'long term'. If xfeeder does pull it off, it would be a massive coup for them. I would expect it to attract a lot of new players to the x games.

Or alternatively, as Jazza suggests above, put the nail in their coffin once and forever.
We shall see.

If this does prove successful for x-feeder, I'm hopeful that it will be extended to Welldone's sports product Marketfeeder Pro. Now that would be a coup.

Let me pose the following.
If one finds a strategy that backtests profitably over a very large number of games then, if the strategy is not explainable in terms of there being a clear, understandable mathematical edge ( that is the odds are clearly wrong), will this test simply mean that if and when one goes live that the future games will probably result in a loss to compensate for the extraordinary successful backtest results ?.
That is , a reversion to the mean will occur and the more successful the backtest the more vicious the future loss is likely to be in live mode ?.
A very,successful backtest could in fact be the ultimate undoing of many ?

Although possible, I don't believe that mean reversion cycles will be longer than all the games played. But who knows, I certainly have a few strats I would like to test against at least 100K games. It would be nice to have an answer, although I would agree that the conclusion can never be absolute.

well good on x feeder if they do it! also they can improve getting 100% of my bets! instead of 90%!

I suppose Grunweld that one should cross bridges as one comes to them.
That is find the successful long term backtested strategy first and then worry about the rest then.
Btw I only proposed my thought in connection with those strategies that have no clear, mathematical reasoning underlying them.
Also I have only proposed it, because it is something that I worry about all the time.
That is am I only setting myself up for a big fall one day by backtesting numerous strategies endlessly ?.

FAFH
"...future games will probably result in a loss to compensate for the extraordinary successful backtest results..."
Is this a shift towards the Mean reversion theory and away from the 'Gambler's Fallacy'? Does the RNG really have a memory? Can this be explained mathematically?

Not at all.
Everything reverts to the mean over time.
You just can't make money out of that simple, unquestionable fact, because of the odds spread and commission.

and you might have to wait 200 years?

Patience is a virtue Jazza ?
But not in the case of appyling a mean reversion strategy to Xgames.
The sometimes snail pace of a reversion means that the commission kills you.
A classic case of so close yet so far.

So, let's get this clear, if a series of results runs away from the expected odds, we can rely on future results to bring it back?
So if in blackjack, the dealer wins 8 out of 10 times, can we be sure that it won't keep doing that?
Does it mean that we can make some predictions because of statistical maths?

You can be sure that the dealer won't keep winning at that frequency. But we cannot predict how many games it will take the for the frequency to normalise. All we know is that over a large enough number of games the frequency of wins will be as expected from the true odds of 3.76. The more games observed, the closer it will get to these odds.

Exactly.
As Grunweld says essentially it can revert virtually to the norm, for example, simply by occurring in the future at the correct probability for a long time.
No way to make money out of that sort of prediction unfortunately.

as I said, it may take 200 years. Too long for me.

Isn't Jazza just a really happy fricking bunny.

Time he developed a brain.

Can all of you tell me what channel is your priority at the moment, if not Poker?

Well my main interest is in Omaha mainbets, but I suspect that will be at the bottom of anyone's list.

Next would be Blackjack, Baccarat (main and side), HiLo and Card Derby - in that order.

BJ, hilo and bcct, in that order please.

ludyburrows 04 Mar 07:27

Isn't Jazza just a really happy fricking bunny.

Time he developed a brain.

This may surprise you, but a developed brain can evolve to a point where it can deal with all sorts of emotions. Happiness is but one.

Keep smiling

Your posts keep me smiling Jazza.
Keep em coming.