don't mean to disregard crazy, [ I ] was meant to be an assumption -- we must first assume that TA 'works' before we can discuss if some actors can break its methods.
Mhm. (That's where I was trying to get to, starting from some simple rules, with
"Bitcoin", an n-player, zero sum game.)(Some forms of technical analysis have some predictive power, and can be employed to make money in some markets,
specifically those markets that have some set of properties, which may or may not apply to Bitcoin, and perhaps only under some conditions, and here are the reasons.)
It's complicated... but TA can be a valid and valuable form of analysis. (Otherwise it would be a losing strategy in major markets and it isn't.)
Corollary: Saying outright that all TA is stupid is stupid.
Which is not to say that:
• Elliot Wave theory or any other
particular method of TA isn't stupid.
• Or that any
particular analysis is sound or will make you money,
only that TA can be used to win. (Better than even odds.)
(Mentioned because some people here dispute the validity of TA in general, and I was unsure, so I thought I'd try to answer the question for my self.)
as for [II]b, that is really the pith and point of this thread.
Yep.
Is there any case in which a rational actor with manipulative potential would make a move that breaks these methods?
I can imagine an irrational actor doing so, but only at a loss.
Is this not begging the question?
Sounds like:
If manipulation loses, the manipulation attempt was irrational, and should not occur. Whereas if it succeeds it was rational.
And perhaps, such "manipulation" is just what anyone should do under the same conditions, hence not "manipulation" in the first place?
Is there any case in which a rational actor with manipulative potential would make a move that breaks these methods?
I imagine that there is. (Will hopefully get to this, in detail, with examples.)
One salient albeit indirect example:
1: You unload Bitcoins. DDOS MtGox. You predict price will fall. Wait for price to fall.
2: Outlook and sentiment change as a result of your actions. (VWAP, volume, support/resist, moving averages, etc.)
3: You buy more BTC back cheaper. (At some target price, well below daily VWAP.)
4: Wait for market sentiment to recover. Repeat as desired.
Generally, you try to:
Make a move, (any move, by any means) - such as posting technical analysis on a public forum - that will (you predict) change market sentiment,
preferrably near some critical point, and leverage the
resulting actions of other traders to achieve some better-than-linear effect on the market.
Which could result in:
Making your, or any, technical analysis (way) less reliable, if (the risk of) "manipulation" is not properly accounted for.
With regards to TA, I suspect that manipulation attempts, successful or not, can ONLY be accounted for if
the effects of historical and similar manipulation attempts are reflected (recorded) in the market data of the time window you're analyzing,
AND the frequency, outcome and effect size accurately reflects the actual outcome, frequency and effect of
future manipulation attempts.
This should hold
even if the manipulation attempt is not "Place limit order for X BTC here" or "Sell X BTC here" to affect some set of indicators in a so-and-such manner,
but perhaps instead "Let's DDOS MtGox today, see what happens."