Conference hydra::dejavu

    There has been some confusion over the address I posted (CRA::Cooper)
    and the address which NOTES says I post from (ERLTC::Cooper).  Both
    refer to the same cluster and therefore, as far as mail is concerned,
    refer to exactly the same machine.  It makes no difference which
    you post to.  If you post to my old address (PBSVAX::Cooper) it
    will be forwarded to me, and this will work fine also -- except
    that you are more likely to have network problems on forwarded mail
    (two machines instead of one must be accessible).
    
    If it reaches me, it counts.
    
    					Topher

    Hi folks.  Despite DEJAVU's inaccesibility the response so far has
    been good.  So far there have been sixteen contributions from thirteen
    different contributers.  The more the merrier folks, and the deadline
    is noon on Monday, so get your contributions in.
    
    A lot of people have found the required format for the contribution
    confusing.  All except one have resubmitted their contribution.
    Let me restate the format required:
    
    The *subject* (not the body) of your mail message must look like
    
    		Cooperative ESP: xxxx
    
    where xxxx is your four-digit contribution.  Their must be four
    digits, even if the leading one, two, three or four digits are 0.
    Case is ignored, the ":" is optional, extra spaces where there
    are already spaces (or at the beginning) are ignored, and anything
    following the contribution, or in the body of the message is
    paid attention to only by me.
    
    Remember you can contribute as many times as the spirit (;-)?) moves
    you -- it makes neither more nor less sense than contributing once.
    
    Thanks.
    
    					Topher

    Can you reveal to us how you are going to map the individual
    contributions into the final result?
    
    Or would that be telling (at this stage of the experiment)?
    
    As I see it, we're not just testing for the precognitive lottery
    event, we're also testing for the clairvoyant "What's Topher going
    to do with the individual contribution?" back-compensation.
    
    Then again, Heinlein claims that "anybody who can do precognition
    has a high-enough voltage brain that clairvoyance or psychokinesis
    is child's play".
    
    	-Bill 

RE: .3 (Bill)
    
    Sorry I didn't get a chance to answer this before the deadline, Bill.
    There is no need for secrecy about the algorithm, but either it
    is irrelevant or this can't work anyway.
    
    The test is based on the idea of ESP being goal directed.  This
    means that it doesn't matter how complex the relationship between
    the correctness of the call and the physical target (if there is
    a physical target) or how many "pieces" of information have to be
    obtained and integrated via ESP.  A lot of experimenation has shown
    that this really does seem to be the case -- its only limititation
    being the confidence of the percipients.
    
    I have a theory, which I call the BT (Branching Time) theory of
    ESP, that explains how this rather odd behavior can occur.  Simplifying
    grossly, this theory says that ESP operates by "looking" down the
    branches in time which are created by a choice and comparing the
    sense of satisfaction of the percipient in each of those possible
    futures.  The choice encouraged by ESP is the one in which there
    is the greatest sense of satisfaction for the percipient.  What
    varies between the branches is the choice made (in this case, which
    four digit number was sent to me and when) and how that turns into
    personal satisfaction (e.g., how that is put together into a final
    number whose accuracy affects the satisfaction of the participants)
    is irrelevant.  In a sense, according to this theory, all ESP is
    explained in terms of a very limited, precognitive auto-empathy.
    
    This theory is a number of similar theories (called collectively
    observation theories) of ESP which similarly explain the goal oriented
    nature of ESP.  They all put the critical essence on ESP at the
    moment of observation by the percipient (or others) of the results
    of the percipients choice.
    
    Anyway: the algorithm is -- a high quality pseudo-random number
    generator is used to shuffle a vector of the numbers between 0000
    and 9999.  When a contribution is processed, it is used to index
    that array and the value found there is used as the "decoded"
    value.  Then the value at the selected location is swapped with another
    selected by the pseudo-random number generator.  Then the next
    contribution is processed.  The decoded values are added together
    and the remainder when that sum is divided by 10000 is the "answer".
    
    The reason for the decoding process is to "isolate" the results
    from the purely non-ESP cognitive patterns and tricks that the
    subconscious can produce.  For example if someone contributes
    7777 then someone else (or the same person) on learning of that
    contribution could subconsciously cancel that result by contributing
    2223.  The result would be as if neither contribution had ever
    been made, not because ESP had guided such a "decision" but through
    some non-ESP psychological reasons.  By filtering the contributions
    any such patterns which get through would *have* to be by ESP.
    
    					Topher

    In that case, what about the "quenchers", people whose presence causes
    oujii boards fail to slide, crystal balls to cloud, tea leaves to clump
    and sink, and tarot decks refuse to shuffle? 
    
    This isn't to say the BT theory is wrong, but it doesn't explain
    _all_ of the known phenomena, hence it isn't the truth.  BT may
    be _part_ of the truth but we still need an interference mechanism.
                                                               
    -----
    
    The second problem is that BT assumes a lack of self-destructive
    urges on the part of the percipient.  A person who (for some reason
    either concious or subconcious) wanted to spoil your experiment
    could, if they were strong BT-sensitive.  
    
    I admit that I like the experiment, but it seems that there should
    be a redesign of it such that a single BT-sensitive spoiler couldn't
    bring it down. 
                  
    I'll think on how to do that and blortz back if I find a solution.
    
    	-Bill
    
                   
        

RE: .5 (Bill)
    
    As I said the description I gave of BT was grossly simplified and
    incomplete.  I described in a few sentences what required three
    rather long technical papers to present in whole.
    
    In any case the correctness or lack of it is rather irrelevant to
    this test.  Its design was not based on my BT model of ESP but
    rather on the experimental finding which BT was meant to explain
    (among other things) -- the apparent goal directedness of ESP and
    the independence of the complexity of the process linking the
    target system with the success condition.  I presented my brief
    description of BT in hopes that it would make goal-directedness
    more intuitive.
    
    The theory which I described is actually BT1, which is a deliberately
    simplified model of ESP which asssumes that there is only one psychic
    individual involved with the process (and even BT1 is grossly
    simplified in the description I made).  BT2 is a second model
    which tries to account for the way that psychics interact -- including
    the effect of psi-jinxes.
    
    Essentially BT2 says that there is only one channel for information
    about satisfaction levels to be passed back through time on.  It
    is impossible to distinguish whose satisfaction (and when) is being
    detected and compared.  Psi ability then becomes logically separable
    into two distinguishable characteristics.  First is the ability
    to signal back ones state of relative satisfaction (influencers)
    and second the ability to read and react to that information at
    choice points (sensitives).  A psi-jinx or quencher is then simply
    a strong psi-missing influencer.
    
    BT in no way assumes "a lack of self-destructive urges" -- it indeed
    assumes them as one major source of psi-missing (this is pretty
    much a traditional view.  By the way, self-destructiveness is not
    really the most useful way to look at this.  The percipient(s)
    may act contrary to the experimenters wishes, or even to their
    own *conscious* desires out of self-preservation (perhaps in
    minor ways) rather than out of self-destructiveness.  For example,
    success in an ESP test may result in emotional conflicts, changes
    in self-image or undesired demands from others.
    
    It has been hypothesized (by a parapsychologist named Batchelder,
    who just recently passed away) that one source of psi-missing
    subconscious motivations might be a fear of the responsibility
    of having psi.  He collected a lot of evidence which suggests that
    ESP works most effectively under conditions in which the responsibility
    for ESP could be displaced -- where no one can be "fingered" as
    the cause of the positive results.  This test was in part inspired
    by this speculation.
    
    There is an interesting symetry in this design.  It is true that
    a single psi-misser could bring down the efforts of a large group
    of psi-hitters.  It is just as true (perhaps more so) that a single
    psi-hitter could bring down the efforts of a large group of psi-
    missers to avoid the target.
    
    A commonly used technique for "protecting" an experiment from the
    individual psi-misser and in general to eliminate "noise" is to
    use majority counting techniques.  One makes a large number of
    separate guesses for, say, the first digit either with a lot of
    seperate people, or with a few or a single person guessing repeatedly,
    eliminates patterns by a decoding technique such as I used and then
    choose that digit which most often comes up.  Unfortunately, while
    there have been some successes with this technique which indicates
    that it might provide some improvement in reliability, there have
    been enough weak results and outright failures to indicate that
    any improvement is marginal.
    
    For some time I have had in mind a fairly elaborate experiment in
    group-ESP which deals explicitly with interactions and makes a
    stab at compensating for psi-missing.  I have not been in position
    to set up this experiment and run it.  It occured to me after the
    number guessing note that I could run a sort of quickie, simplified
    experiment in a similar vein.  The result is this note.  If it
    works it will mean a lot, but there are enough things which can
    go wrong with it so that a failure -- as discouraging as it would
    be -- would not mean much.  I certainly do not see this test as
    the prototype for a perfectly reliable ESP procedure but simply
    as a "what-the-h**l let's see what happens" type of thing which
    might help in the design of future, more complex designs.  Succeed
    or fail, I have already learned some things about the mechanics
    of experiments of this type which may prove useful in the future.
    Most important -- I thought it would be fun to try.
    
    					Topher

    I'm not saying it's not fun to try...
    
    I'd even venture that it is _worthwhile_ to try...
    
    I just like to make trouble, you know... :-)
                                                               
    
    When is the drawing again?
    
    	-Bill

    OK, I've made the tally.  The file CRA::[COOPER]CoopESP.ENC contains
    the encrypted form of our collective guess.  I would appreciate
    everyone with ENCRYPT on their systems copying it over, and as many
    other people as possible as well (in case noone with ENCRYPT copies
    it, I can still prove I didn't tamper with the file).  When you
    get it say:
    
    		SET PROT=(W) CoopESP.ENC
    
    which will block me from monkeying with it (if I desired to).
    
    Tomorrow evening is the drawing, and as soon after that as I can
    manage (sometime Wednesday morning), I'll post the results and
    make a VMS command file for decrypting the file available.
    
    By the way, in the final count -- there were 33 contributions from
    27 people (two people contributed three times apiece, two people
    contributed two times apiece, and 23 people contributed once apiece).
    The person who previously had not resubmitted has done so, but one
    person who contributed twice on sunday and used the wrong format
    did not get their correction in on time (since I was in meetings most
    of the morning, I was unable to tell them about the problem
    until they didn't have much time to try again).  Also one person
    submitted after the deadline, and so wasn't counted.  (The preceding
    tally of 33 contributions did not include these last three which
    I couldn't use).
    
    					Topher

I like that discussion about psi-spoilers; I even start considering buying one
for my car. ;-) 

This BT-theory appeals to me. Especially the BT2 one, where a group of people
are considered as one unit, potentially having ESP power. Somehow, it makes it
easier to think less of the source of power, and more of the attributes of the
power itself.

It took some time for me to let the penny drop, but now I think I know why I had
some uneasy feelings reading the discussion about psi-spoilers: you discuss the
part, when analyzing the whole.

To me, it doesn't seem useful to discuss how the parts of this whole will react
as a part; it is the whole that is analyzed now. Let me give an analogon: a
glass of water. The water may have no in- or outflux, yet looking at the
individual (H2O molecule) level, you will see a lot of movement. It is, of
course, possible to calculate the overall movement of the water from the
movements of all individual molecules, but I personally would find it very
unsatisfactory to find out in the end that all movements together end up in a
zero movement. 

Try and imagine this incredibly long list of numbers, a line under it, a "plus"
sign next to the line, and a simple zero under the line. Wouldn't you need a lot
of agression paper ;-)? 

The same applies to observing ESP phenomena where a group of people are supposed
to have ESP power *as a group*. You can try and "calculate" the result on the
basis of individual behavior, but can we? And if we can: do we *really* learn
something from it about the group power? I don't think so.

Hey Topher: if this test succeeds, I know what test to do next. Have the group
linked to EASYNET as a node. Add an extra command to DCL with null meaning. The
group must be able to assign meaning to this command via PK, and use it as the
link command for itself. You want a suggestion for the command?




SET GHOST DEJAVU


(I would be excited as a network partner to log in)

:-)
Arie

RE: .8 (me)
    
    The correct file spec is
    
    		CRA::ADM3$:[Cooper]CoopESP.ENC
    
    Sorry about that.
    
    					Topher

    Well, the lottery number drawn was 9988, and our collective guess
    was 6482 a miss.  There was one digit matching but out of the 10000
    possible numbers in the lottery 2916 of them match in a single digit.
    That is, if nothing at all were going on and only chance were
    operating, we would expect to do this well about three times out
    of ten.
    
    I am not discouraged.  The test was a long shot done for fun in
    the first place -- a positive result would have been considered
    spectacular.  In addition things were confused by the overly strict
    format required for peoples contribution and by DEJAVU being
    unavailable for a large part of the test period.  Perhaps sometime
    soon we can try again, if people are interested.
    
    					Topher

    Sure, let's do it again.  After all, one data point does not a theory
    disprove.
                            
    May I suggest we repeat the current experiment for a few weeks,
    just to get a few more data points?  My back-of-the-wrist calculations
    show that if we can repeat this level of performance four more times,
    we are functioning at the 1-in-100 chance level.  Not merely
    spectacular, but publishable!
    
    -----
    
    I'm still thinking about the upgraded experiment, the one that 
    somehow doesn't let psi-spoilers smash the results.  The big problem
    I'm having is how to deal with the time factor.
    
    It's a very nasty problem that even game theory doesn't help in,
    because game theory allows hidden variables inside a player's
    decision-making process, and the nature of the experiment is to
    show that there _are_ _no_ hidden variables anywhere in the universe.  
                                                               
    	-Bill

       Perhaps I should just keep quiet, but I'm not...
    
       I think that if all of you wish to play at this, it's fine.
    
       HOWEVER,...
    
            All of this is *testing*...it IS NOT *trusting*.  It is
    a wonderful opportunity for the negative ego to show it is "right"
    and you are "wrong" (stupid, undeserving, bad, evil, rotten, and
    on and on...)  Can you play these games and expect success?  Very
    unlikely.  Why?  Because, again, the negative ego has too much
    at stake here and isn't going to let you get past it.  Can you
    get past this negative ego?  Sure.  But if you do, you won't bother
    with any testing.  So, from my perspective all of this proves 
    nothing, except the flexing of negative ego...a condition which
    is apparent to me in virtually every aspect of my reality.
    
    
    Frederick
    (see if I ever ask any of you for my lotto numbers!  ;-)  )
    

    WELL,  it may not have worked, BUT, last nite I had the "urge"
    to play the number 8888, (not my guess by the way).... I won on
    the last two digits... only $25.00, but.  
    
    Carla
    

    re .12
     
    	There are those of us who look, simply because we rejoice in
    not knowing what we'll find when we open our eyes.
    
    The universe is a wonderful and fascinating and bizarre place.
    
    	-Bill

    I'm teaching a course this week so I only have time for some quick
    replies.
    
RE: .12 (Bill)
    
    <<Repeating right away>> Your enthusiasm is appreciated.  But first
    I'm rather busy until Christmas.  Second, guessing numbers can go
    stale very quickly and if this isn't fun its not worth doing (for
    one thing its unlikely to work, for another nobody is likely to
    contribute, and for still another fun is the main motivation in
    the first place).  Third, I need to do some work on the design
    and the programs which implement it.
    
    <<Publishable>> Its the policy in parapsychology that the results
    of an experiment not determine whether or not it is published, so
    I could probably publish it right now, but I would rather refine
    the design and publish something a little more meaningful than
    what we would have going after a marginally small, though significant
    result.
    
    <<Four more times>>  The usual standard for "significance" these
    days is 1-in-20 so 3 (more) tests with this result would give us
    a "positive" experiment.  But this assumes that our one digit hit
    represented a real, albeit weaker than looked for, psi effect. 
    I think that it is much more likely that we were just lucky, and
    that we should only expect one out of those three to "hit" with
    one digit.
    
    <<Assumptions of game theory>>  Yes, psi turns all our assumptions
    on its head.  This test was "demonstration oriented" -- just trying
    to show that something was happening.  The complete meaninglessness
    of a "closed" experimental situation and the assumed leakage of
    any "blind" conditions makes "process oriented" experiments --
    which try to determine the characteristics of psi -- very difficult
    to interpret.  The scientific method *as practiced in the past*
    is on shaky grounds when psi may be present (whether or not it
    is a psi experiment -- i.e., always).
    
RE: .13 (Frederick)
    
    I think that we can all agree that your note clarifies something
    about the operation of negative ego.
    
    Frederick, you believe that the only thing worth learning is your
    own desires, while I believe that there is much much more to learn.
    You feel that finding out what you *can* do is rejecting trust,
    while I feel that it is learning about yourself as you are.  You
    feel that this is pushing forward the negative ego, while I think
    that it is asking the Universe -- of which I am a small but essential
    part -- about itself.  You prefer your way, I prefer mine.
    
    As to whether we can expect results, I can give a definite answer
    to that.  The task I set us was very ambitious -- a positive result
    would have been spectacular.  But similar tests -- mostly less
    ambitious ones -- have been successful in the past, not once but
    hundreds, even thousands of times.  If I can be faulted here, it
    is not for my lack of trust, but in my naive belief that "what the
    hell, it might work."
    
RE: .14 (Carla)
    
    Although it didn't count that much, in an abstract sense your guess
    caried alot of the "feeling" of the actual number -- repetition
    of a "high" digit.  If this were a judged matching procedure, I
    would rate your guess much higher than say "9158" even though
    the number of digits guessed correctly were the same.  Without being
    able to say for sure you weren't just lucky, I would say that
    was an encouragingly good match -- keep working at it (but don't
    force it -- that's a good way to get poor fast).
    
    						Topher