Prediction Registry

Introduction
Scheduled Regular Events

Scheduled Unique Events
Unpredictable Events
Other Events Introduction

Scheduled Regular Events

  • Eckhard Etzold, pred. 1998-12-22: Christmas eve: All around the world Christians celebrate Christmas with family centered evening celebrations, mostly in Europe and in America. Christmas eve is a wide spread event celebrated around the Christian world. This makes it difficult to give an exact prediction. There will be probably no peak deviation at a single time but two, or perhaps three modest waves of effect, one for Europe, one for east coast US, one for the west coast. In Europe the time zones of interest include Moscow, Middle-European and Greenwich. Specifically, I expect statistical deviations in a two hour period centered on UTC 17:00 for Europe, 22:00 and 01:00 (25th) for American east and west coast, respectively.

    Hand calculation: No significant effects: 98-12-24, 16:00-17:45 UTC: Chi-square 52.497, 64 df, p=0.847 98-12-24, 21:00 - 17:45 UTC: Chi-square 62.411, 64 df, p=0.533 98-12-25, 00:00 - 01:45 UTC: Chi-square 55.598, 64 df, p=0.764

  • RDN, pred. late 1998-08: New Year Celebrations: Expect peak deviation at midnight 5 minutes. Expect correlation structure to proceed through the 24 hour period. An approximate analysis using 15 minutes around the midnight transition shows a consistent negative trend. The 5 minute pattern will be examined later. In August, 1999, adequate tools for this analysis were finally available. The second-by-second outcome is highly significant, with a monotonic trend across the 24 periods, and a Chisquare of 136430 with 135000 degrees of freedom and p-value of 0.0031. Both Bierman and Broughton used second-by-second data, and this was chosen as the formal analysis for the Nelson prediction as well. Because the block-size had not otherwise been specified, this analysis was supplemented by explorations of 1-minute, 2-minute, and 5-minute block sizes, and a split into maxi- and mini- celebrating zones was made to complement the "evoked potential" analysis by Richard Broughton (the present analysis is a sequential accumulation, Broughton's does signal averaging). All of these explorations are displayed in a separate page.

  • RDN, pred. late 1998-08: Times Square New Year: Expect concentration of effect around midnight 5 minutes in New York City, with its strong tradition. Here, we will also look for, and predict, a progressive increase in the deviations or strength of effect in the data as the midnight hour approaches. The second-by-second data for midnight 5 show a pronounced "spike" centered about 1 minute before the stroke of midnight, but although positive, the deviation for the predicted period is not significant.

  • Dick Bierman: Prediction on 1998-12-31 for the upcoming New Year event. Deviations from expectation will be tested on the basis of the trial level variance, with a time-period of one second.

    a) The European RNGs 1000, 1003, 1022, & 37 will show larger chi-square at local New Year (00:00 - 00:01 MEZ = 23:00 - 23:01 UTC on 31 Dec) than the US - RNG's. This prediction is for a one minute period and relies on an intuition that this excess will be present for less than a minute.

    b) The same prediction for the RNGs on the US eastcoast compared with the Europeans 6 hours later: larger chi-square at local US New Year time (05:00 - 05:01 UTC) in the US compared with Europe.

  • Prediction by Kishor, Bangalore, India, made on 16 Sept 1999 concerning Indian elections. Kishor believes, "Most probably the results start on 6th afternoon. The better time is 5.00 PM to 10.PM IST (Indian Standard Time) on 6th and 7th of Oct '99." This will be treated as an informal prediction, because the Indian elections will not garner the worldwide attention of a GCP global event. Exploratory analysis will, however, be done on local and global basis, with 15 min blocks. The predicted times are 11:30 to 16:30 UTC on the 6th and 7th of October.

  • For the turn of the Millennium, we will be making several formal predictions. Some of these have been in place since the beginning of the project, and some are based on experience with other events, especially New Years 1998. A brief list of the intentions as of September, 1999 includes three specific, formal predictions (4 is a duplicate):

    1. Duplicates RDN prediction for New Years, 1998, saying that 00:00 +/- 5 min in every timezone will deviate from expectation (1-sec blocks)
    2. Duplicates Richard Broughton's maxi vs mini celebrating zones, using the same parameters as in 1998.
    3. Predicts an effect of the "Just A Minute" event, 12:00 to 12:01, 1 Jan 2000, in all timezones, (1-sec blocks)
    4. Predicts an effect correlated with the NYC countdown, which I hope to coordinate with the folks who run that (see below). Parameters not yet set. As of early October, the prediction for the NYC celebration appears to be the same as the RDN prediction above, since the Times Square BID celebrations will begin at 6 am, corresponding to the first midnight in Fiji, and will continue with a special program each hour thereafter for the 24 timezones. See The Big Idea for a complete description.

  • RDN (Duplicate, retained for explanatory context and timing information -- see above, points 1, 4.) Prediction in late 1998-08: Millennium: We may be able to coordinate with Gretchen Dykstra, Times Square Business District director(BID). In an interview in the United Airlines magazine, Hemispheres, July 1998, Dykstra says of the celebration for the year 2000, "... we're going to have 24 special minutes [for each time zone]. And as the millennium travels around the world we're going to celebrate the world here." The GPC will predict a progressive increase in the size of deviations near the transition time.

  • RDN, pred. late 1998-08: Millennium: Larger temporal spread of the effects of the unification toward which we gravitate on New Year's eve around the world. We expect there to be increasing attention to events on a world stage that will see the transition to the 3rd millennium of this period. In specific predictive terms: we will coordinate and assess the small bursts of information in the Unpredictable events category, to see whether there is an objective acceleration in rate of occurrence, or an increase in information density, and correlate historical trends with structure in the data taken in the time leading up to and following the actual shift into the year 2000.

  • September 2000, and subsequent months. A regular meditation event every other Sunday is being organized by the World Mind Society. The theme is a Global Group Mind Meditation, and the time is regularly set to begin at 3:00pm PST. The first of these was on Sept 24 2000. Matthew Webb, who is the primary organizer, is willing to provide a specific time period which he feels may be most appropriate for a GCP prediction of deviation, usually a period of 20 or 30 minutes beginning around 3:15 or so. We have been predicting these formally with medium expectation, and using raw trial data for the analysis. This description will serve as a general prediction for a series of these meditations over the next few months. A note will be added when we decide we have enough samples to judge the effect on the GCP.

  • The US presidential election, Nov 7 2000. Paul Bethke suggests that even though this is a US local event, there is understandable world-wide interest and attention. The election is predicted to be close, with polls showing a 2 or 3 percentage point advantage for Bush over Gore. After some discussion of a reasonable approach, Paul suggests that

    ... we set a data collection/observation time period of one hour before to 3 hours after the first definite "projection" of a winner. This [should identify the time of greatest engagement and attention. We also should note] the time of the losing candidate's "concession" call/speech. We can use CNN as the official news media driving the clock.

    ... The thrust of the prediction would be to observe the data at and around the time when most people would be impacted by the election results. A side experiment may be to see if there are times of coherency throughout the voting day, or maybe a buildup as time approaches result announcement.

    The formal prediction will be for a departure from expectation in the raw, trial-level data during the four hour period beginning one hour before CNN's first definite projection of a winner, and continuing for three hours after this time. Medium expectation (based on the US-centric nature of the event). Some explorations will be undertaken to see, for example, whether the US eggs will differentially respond. We will also examine the data during the losing candidate's concession call or speech, and In addition, we will take an exploratory look at the whole of election day, noting the time polls open on the east coast and finally close on the west coast.

    Further formal predictions were made for the crucial decisions in the long month of waiting for resolution. These are documented on the elections web page.

  • For the New Year transition 2000 to 2001, we have two major predictions based on experience with the prior occasions, 1998 to 1999 and the Y2K transition.

    1. The first duplicates RDN prediction for New Years, 1998 and 1999, specifying that 00:00 +/- 5 min in every timezone will deviate from expectation (to be analysed in 1-sec blocks).
    2. The second formal prediction is for a repeat of the variance effects predicted by Dean Radin for the Y2K event. The analysis will use the smoothed, normalized variance procedure, but will also explore other approaches.
    3. As an exploration (not formal prediction) we will again assess Richard Broughton's maxi vs mini celebrating zones, using the same definitions as in 1998.
    4. A subsidiary, but formal prediction is for the "010101" event, 11:45 to 12:15, 1 Jan 2001, in the GMT timezone, (1-sec blocks), as detailed in the 010101 webpage.
  • The Presidential Inauguration on Jan 20 2001. This is a US event of course, but of interest to the wider world. The main ceremony began about half an hour before the inauguration proper, which was at 12:02, EST (17:02 UTC), and continued with a speech by President Bush lasting until about 12:18. The formal prediction was set for the hour surrounding the inauguration, 16:32 to 17:32. Seconds resolution and low or medium expectation, in line with other political events. An exploration of US vs Other eggs will be done for context.
  • Prediction, Lettieri and Nelson, 31 Oct, 2001 about noon. Arnold and Roger were discussing the concern and security precautions attending the World Series baseball game in Yankee Stadium scheduled for 8:00 pm today. As a sporting event it draws the attention of lots of Americans, and because of the US government's announcement of a warning that some terrorist activity might occur around the end of October, though with no indications of where or exactly when, this evening seems likely to be a focus of attention on a broader scale. Our prediction is for an increase in the Chisquare (standard analysis) from 7:00 to midnight EST, thus including some time prior to and following the actual game. Though it is not intended as a testable hypothesis, this formal prediction will serve as a framework in which we may consider any terrorist attack, should there be one, God forbid. The formal prediction is for 00.00 to 05:00 on 1 November, GMT, with medium confidence, seconds resolution.
  • For the New Year transition 2001 to 2002, we have two major predictions based on experience with the prior occasions, 1998 to 1999, the Y2K transition, and the 2000 to 2001 transition.

    1. The first duplicates RDN prediction for New Years, 1998, 1999, and 2000, specifying that 00:00 +/- 5 min in every timezone will deviate from expectation (to be analysed in 1-sec blocks).
    2. The second formal prediction is for a repeat of the variance effects predicted by Dean Radin for the Y2K event. The analysis will use the smoothed, normalized variance procedure, but will also explore other approaches.
  • For the New Year transition 2002 to 2003, we will again have two major predictions based on experience with the prior occasions, 1998 to 1999, the Y2K transition, the 2000 to 2001 transition, and 2001-2002.

    1. The first duplicates RDN prediction for New Years, 1998, 1999, 2000 and 2001, specifying that 00:00 +/- 5 min in every timezone will show a positive deviation from expectation. This is the "standard analysis" using the Stouffer Z composite across eggs, squared to give a Chisquare distributed quantity. For New Years, there is an extra step to form a composite across time zones using signal averaging (superimposition). The resolution for analysis is 1 sec, i.e., full detail. Medium confidence.
    2. The second formal prediction is for a repeat of the variance effects originally predicted by Dean Radin for the Y2K event. The analysis will use the smoothed, normalized variance procedure, and signal averaging to combine the results across time zones. As before, the probability will be estimated using permutation analysis to determine the combined probability of the magnitude of the peak deviation and its distance from midnight. (This was done in +/-30 min range in previous years but should be done in the +/- 5-min range set in the original predictions). Because past experience has shown that deviation of the variance is sometimes positive and sometimes negative, we will assess both the unsquared and squared value and choose the stronger result, paying a factor of two Bonferonni correction to obtain the appropriate probability calculation. Both versions will be presented in the detailed analysis page. Essentially, this amounts to doing a two-tailed test, or making a prediction about magnitude of deviation irrespective of direction. A four-minute smoothing window will be specified, conforming to the previous years' analysis protocol.

      Note: prior to actual analysis the decision was made that the prediction is about deviations of the variance -- not variance squared. This will be the standard variance analysis now and in future similar applications.


Introduction
Scheduled Regular Events

Scheduled Unique Events
Unpredictable Events
Other Events Introduction


[next page]