> The single most powerful proof in science is
> replication. And the number of replications that
> defy chance is five, according to the binomial
> expansion. Example: I meet old classmate Sally
> at the airport, and she introduces me to her five
> daughters. Then, she pulls me side, and indicates
> that she and her husband would go for another
> child if there was a chance of having a son. By
> chance she has a 50/50 chance of one girl, 1
> chance in 4 (1/4) having two in a row, 1 chance in
> 8 (1/8) of having three in a row, 1 chance in 16
> (1/16) of having four in a row, and 1 chance in
> 32 (1/32) of having 5. If the probability of any
> sequence is less than 5 per cent, the usual
> scientific decision rule would be that her five
> daughters are rarer than would be expected by
> chance. If we adopted a more conservative
> decision rule of 1 per cent, we would need six
> consecutive daughters, having 1 in 64 (1/64)
Understood. In my own work I try to establish such thresholds when I can. Sometimes it seems quite easy, other times not so straightforward, as I will explain below.
What grabbed my
> > was that this turnaround began in Leicester
> > first match after King Richard III was buried
> > there on March 26.... this was a significant point in the
> > City story.
Ray... The statistic that works here is called a
> regression analysis: "wins X time" One line
> would be chance, and the comparison line (the
> regression line) would rise. The decision rule
> would be the same as the earlier example I gave.
I think I see where you are going, but wonder if you can clarify on the greater example. This game features what I call 'convergence', where number of meaningful associations concern on the same element, King, which is relevant to the Richard III narrative.
So, the 'event' is the winning goal in this key game. But here's what I see that gives it a synchronistic signature.
1. The winning goal was scored by a player named Andy King.
2. This King was an unusually 'pure' Leicester City player, had played for the club his entire career.
3. LCity is owned by a corporation named King Power, which superbly complements that King Richard-Leicester backstory.
4. On the day the winning goal was scored, it was King Power CEO's birthday.
Clearly, we're dealing with a scenario that's more complex than a binary one. How might you begin to frame an analysis in a regression or Chance analysis here. My layman's guess is to seek to apply suitable ranges.
For example, since Andy King was the only Premier League player by that last name, perhaps it is more relevant to turn to society and ask, what percentage of people in the UK have last name, King? If one percent I would think there's a 1-in-100 chance that our scorer would have such a name. That seems straightforward. Is it Ray?
note, according to this page, King is not in the top 25 of Britain's most common names. The 25th name, Clarke, is the last name of 108,000 out of 65,640,000 Brits, for a frequency of 0.16 percent or about 1-in-600. King will be less common than this.
Just to complicate matters on the sub-point point alone, these days the Premier League has highly significant foreign content, which would drive that figure down. These days about 70 percent of players come from outside the UK. In 2015 they came from 60 other countries. One can be sure that most of these players come from countries where English is not the main language, so in the percentage of Kings in those 60 nations will be far, far, far less than our sketch 1-in-600 UK rate. So, that a player named King would score this particular goal, in this day and age must surely be much much less than 1-in-1000.
This is where we hit diminishing returns on our probability analysis. The point has been made that this particular coincidence is highly improbable and maybe extremely so, depending on one's use of language. But can we provide a minimum number to work with in an aggregate probability analysis? Maybe at this point it is best to stay with English demographics, and say, "When we only consider UK players, the chance of a King scoring this goal is (clearly?) x percent, based on known demographics. This approach, of addressing what can be known while affirming that the greater picture is more improbable, is what I call the 'minimal inference' approach. Would you and your colleagues find that kind of approach reasonable?
More generally, I would be interested in how you might begin to frame these four attributes.
> > So, according the 'template' I mentioned to
> > Daniel, this was a logical place to look for
> > of intervention, or hints of the non-random,
> > the April 4 game was THE turning point. As I
> > in the essay I posted online, a number of
> > King-related themes "converge" on that match.
> > This is one of the say I have found that syncs
> > unfold: in ways that are clearly uniquely
> > relevant to the main event and, or seemingly,
> > improbable.
> > In that sub-example, I would say that the
> > turnaround features a 'convergence level of 4'
> > based on the King cluster. This is what I call
> > thematic redundancy, and I have found that
> > redundancy comes up with surprising frequency.
> > Staying with my template that, when the
> > Footballing Story was finally over, I decided
> > find out when exactly Leicester had actually
> > the Premier League. Everybody celebrated the
> > victory in early May, but in fact the Foxes had
> > clinched a few weeks earlier. That made an
> > earlier game, and likely its winning goal,
> > 'logical' place to look for non-random
> > since it was clearly 'salient' or stood out in
> > this storyline.
> > The results are posted in the link. Not only
> > this goal take place at a very English time,
> > winning goal proved to be highly significant
> > respect to three related settings - LC's
> > the European League and the Premier League.
> > is another kind of redundancy, based on
> > 'significance' rather than kings...
> > These results were not surprising Ray. On the
> > contrary, they appear to be oddly normal. The
> > best-synced events display a lot of unique
> > redundancy, and here I will remind you of that
> > essay I sent where the number 87 came up a
> > times, in or in clear relation to another very
> > well known even.
> > I am beginning to think that I may be asking
> > right questions. While I can't say what causes
> > syncs, I often treat them as if they were
> > For example, So, Leicester City's amazing
> > turnaround began in the match just after
> > was buried, and everyone's talking about that
> > in the UK? Let's take a closer look at that,
> > the 'King' theme in mind...
> > To come full circle, I suspect that I may be
> > rewarded for following breadcrumbs that have
> > left there to pique the curiousity of those
> > willing to pause and reflect.
> > I would appreciate it if you would take a look
> > the Leicester essay with attention to these key
> > events. As a stats man, can you comment on how
> > regression might come into play with those
> > examples? I can't think of any yet. Really
> > know my limitations. I did stats in university
> > and went as far as calculus, but I'm well aware
> > that I'm a layman. My hope has always been to
> > present data that may be of interested to
> > professionals like you, who may be able to sort
> > out where I've hit and missed, and build that
> > something that I seem to be on to. Looking back
> > over the years, I would say that I'm batting
> > 90 percent when it comes to turning to events
> > seem to be synced for whatever reason: that is,
> > the events usually reveal multiple other
> > meaningful data - and in all cases we are
> > about events that loom just as large, in their
> > settings, as the world's most prominent
> > do in their respective settings.
> > A final thought, if syncs are the handiwork of
> > designer who chooses to conduct himself as our
> > aforementioned card reader might - installing
> > highly improbable outcomes on its terms, not
> > - then how far can we run with a statistical
> > analysis? Can that be done, Ray? Or, - my
> > present hope - can we establish reasonable
> > where the likelihood of non-randomness is
> > reasonably established? Is it not better - at
> > least until we have a suitable search engine,
> > perhaps - to seek only to assess the
> > non-credibility of Chance, rather than try to
> > that AND determine the root cause of
> > synchronicity.
> > Please take your time on this, Ray. And thanks.
Edited 2 time(s). Last edit at 04-Oct-18 04:09 by Poster Boy.