Introduction
Victims
Suspects
Witnesses
Ripper Letters
Police Officials
Official Documents
Press Reports
Victorian London
Message Boards
Ripper Media
Authors
Dissertations
Timelines
Games & Diversions
Photo Archive
Ripper Wiki
Casebook Examiner
Ripper Podcast
About the Casebook


Most Recent Posts:
General Suspect Discussion: Sir Robert Anderson's sixth victim - by Herlock Sholmes 1 hour ago.
General Suspect Discussion: Sir Robert Anderson's sixth victim - by etenguy 2 hours ago.
Scene of the Crimes: Broad Shoulders, Elizabeth's Killer ? - by The Rookie Detective 2 hours ago.
Scene of the Crimes: Broad Shoulders, Elizabeth's Killer ? - by New Waterloo 3 hours ago.
Non-Fiction: The Five - by Linotte 3 hours ago.
A6 Murders: A6 Rebooted - by Sherlock Houses 4 hours ago.
Motive, Method and Madness: Escalation: What would Jack do after Mary Kelly? - by The Rookie Detective 4 hours ago.
Scene of the Crimes: Broad Shoulders, Elizabeth's Killer ? - by Herlock Sholmes 5 hours ago.

Most Popular Threads:
Pub Talk: Irritations - (16 posts)
Scene of the Crimes: Broad Shoulders, Elizabeth's Killer ? - (13 posts)
Non-Fiction: The Five - (6 posts)
General Suspect Discussion: Sir Robert Anderson's sixth victim - (4 posts)
Audio -- Visual: Release Date for Kosminski Documentary - (3 posts)
Motive, Method and Madness: Escalation: What would Jack do after Mary Kelly? - (3 posts)


 A Ripper Notes Article 
This article originally appeared in Ripper Notes. Ripper Notes is the only American Ripper periodical available on the market, and has quickly grown into one of the more substantial offerings in the genre. For more information, view our Ripper Notes page. Our thanks to the editor of Ripper Notes for permission to reprint this article.
Connecting the Dots: Were the Ripper Crime Scenes Chosen to Form a Pattern?
By Dan Norder

Ripper Notes Editor Dan Norder lives in Madison, Wisconsin. He owns and operates the website at www.mythology.com and is writing a book examining the origins of some popular modern legends.


Human beings are born with the desire to look at the details in the world around them and try to make meaningful patterns out of them. This is fortunate, for without this capacity we would be unable to come to even the most basic conclusions about how to plan for our lives and the future. The drive is so powerful, however, that we often convince ourselves that some meaningful pattern exists when it was actually nothing more than mere chance at work. The same intellectual process is responsible for both our greatest advances in science and also our most arbitrary superstitions.

For example, if it weren't for centuries of accumulated knowledge about motion, chemistry and other fields built up from successfully identified patterns, mankind never would have set foot on the moon. On the other hand, improperly attributing the dark spots on the face of the moon to an intentional design has led people of various cultures to identify the markings as a rabbit, a girl holding onto a tree in the wind, a man carrying a stick followed by a dog, and many other figures. Elaborate stories and myths were created as a result of these ideas, sometimes influencing beliefs in other areas.

The man in the moon, to take the version most of us are more familiar with, was at times believed to be the Biblical Cain or the Wandering Jew, each damned by God for his individual offenses.(Note 1) The character symbolized opposition to the rule of God, thus the full moon, which displayed the figure's entire body, was believed to be a bad influence. This belief still persists, minus the legends that helped shape the idea.

Interestingly enough, there is a small but real possibility that the name of Jack the Ripper himself may have been influenced by these legends. Another folkloric character who was doomed to wander the world forever borrowed some of the attributes of these earlier man in the moon stories. This one got into trouble after repeatedly tricking the devil, typically over money. He was refused entry into hell by the upset Satan but had lived too sinful of a life to be allowed entry into heaven. His name? Jack o' the Lantern, as carved into the faces of pumpkins all across America on Halloween and lit up to shine like the moon at night. You might call him the original lunatic.

We now know those spots on the moon's surface are the result of random collisions with meteors, as well as other astronomical events that happened over millions of years. That didn't prevent different people from weaving elaborate stories around the figures they had convinced themselves they saw. The patterns they came up with were often not the same as what people elsewhere imagined there, even though the moon, thanks to gravitational forces that locked its rotation to match that of earth's, always shows one side of its surface to everyone on Earth. The exact same set of meaningless spots were explained in completely different ways.

Some people have identified what they consider to be intentional patterns formed by the locations where the Whitechapel murders attributed to Jack the Ripper were committed. Unlike the spots on the moon, we know that they aren't the result of a natural process at work. An individual -- one with murderous intent -- is, in fact, responsible for them. Of course that by itself doesn't answer the question of whether the killings were specifically planned to form a pattern, as someone can set out to kill without taking the physical location of each site compared to each other into much consideration. But one thing the Ripper locations do have in common with the figures imagined on the moon's surface is that the unifying patterns that are described by various people often don't match each other.

A surprisingly large variety of symbols have been advanced as the supposed goal of the placement of the bodies. The ones that have probably gotten the most discussion are a cross, an arrow, a pentagram of sorts, and a figure shaped like the intersection of two circles.

Let's spend a little time examining each of these alleged patterns.

The Cross

The first of the symbols mentioned in a theory that the Ripper murder sites were planned was probably a cross.

Some people were (and others currently are) under the belief that the Whitechapel murderer was on a religious crusade to kill off sinners and had specifically targeted prostitutes for that purpose. In such a case, a cross would be an unsurprising choice if this kind of a personality were to pick an identifying mark. People have in the past perverted the meaning of the symbols of various religions by using them in ways that the vast majority of its followers would find highly offensive. As one example, white supremacists have burned crosses on the lawns where minorities lived while following their own unique brand of Christianity.

Another related theory is that the killer may have wanted to form the symbol of a cross not to advance that religious belief (or a murderous offshoot of it, as the case may be) but to denigrate it. A proven example of this concept at work is how the Nazis used the Star of David to identify and persecute Jews. Here, the theory goes, the killer considered himself an enemy of the Christian religion, perhaps all the way to practicing Satanic rituals.

Not too long after the end of the murders of the victims now known as the canonical five, the following theory was advanced in a London newspaper :

"Further, in the practice of evocation the sacrifice of human victims was a necessary part of the process, and the profanation of the cross and other emblems usually considered sacred was also enjoined. In this connection it will be well to remember one most extraordinary and unparalleled circumstance in the commission of the Whitechapel murders, and a thing which could not by any possibility have been brought about fortuitously. Leaving out the last murder, committed indoors, which was most probably not committed by the fiend of whom we speak, we find that the sites of the murders, six in number, form a perfect cross. That is to say, a line ruled from No. 3 to No. 6, on a map having the murder sites marked and numbered, passes exactly through Nos. 1 and 2, while the cross arms are accurately formed by a line from No. 4 to 5. The seventh, or Dorset-street murder, does not fall within either of these lines, and there is nothing to connect it with the others except the mutilations. But the mutilations in this latter case were evidently not made by any one having the practical knowledge of the knife and the position of the respective organs which was exhibited in the other six cases, and also in the mutilated trunk found in the new police-buildings, which was probably the first of the series of murders, and was committed somewhere on the lines of the cross, the body being removed at the time. Did the murderer, then, designing to offer the mystic number of seven human sacrifices in the form of cross - a form which he intended to profane - deliberately pick out beforehand on a map the places in which he would offer them to his infernal deity of murder? If not, surely these six coincidences(?) are the most marvellous event of our time."

-The Pall Mall Gazette,
Dec. 1, 1888
(Note 2)

This particular piece is noteworthy not only because it advanced the cross theory but also because it contributed to getting its author, Robert Donston Stephenson (a.k.a Roslyn D'Onston),(Note 3) named by others as a Ripper suspect.

There are a few aspects of this excerpt which deserve further explanation for those readers who might not have studied the case in-depth yet. (The rest of you can bear with me for the length of this paragraph.) The first is that, although these days we are quite used to hearing that Jack the Ripper had five victims, the murders were originally thought to have started with the deaths of Emma Smith and Martha Tabram earlier in the same year as the "canonical five" victims. So when Stephenson mentions drawing a line from victim three to victim six, he is actually referring to the women generally accepted these days as the first (Mary Ann "Polly" Nichols) and the fourth (Catherine Eddowes) of the series. Similarly, his numbers four and five would be our second (Annie Chapman) and third (Elizabeth Stride) victims of the string, while number seven on his list is the last (Mary Jane Kelly) of the canonical five. The "mutilated trunk" he mentioned refers to the human torso that was found in early October and called the "Whitehall Mystery."(Note 4) The police never linked that to the Whitechapel murders, and only a small minority of modern researchers think there is any connection.(Note 5)

Other than the inclusion of Smith and Tabram, an idea which has fallen out of favor, this description contains many features that we will see over and over.

First, the theorist ends up picking and choosing which people to count as Ripper victims based upon the geometry of where they were found and not upon other criteria. Stephenson ignores Kelly and tries to explain away her death by coming up with an alternate scenario -- in this case, that her death supposedly doesn't show the same features as the others so must have been caused by some other hand. While many modern Ripperologists argue over whether Stride should be considered as one of the Ripper's victims, Stephenson and others would not even entertain such an idea for the simple fact that all of their various alleged patterns don't work if the location of her body were removed.

Second, they start mentally making connections that are rather arbitrary. Connect-the-dot puzzles that children play are actually numbered so there's an objective process for linking them together. The only rules the crime scene pattern theorists use is whatever they have to do to get some generally recognizable figure. There's no concrete reason to see a cross formed from four locations instead of, say, the outline of a diamond or a skewed letter Z. (A letter Z? Quick, someone check to see if Zorro has an alibi!)

Third, there is apparently some amount of wishful thinking when looking at the arrangement of the locations. Here Stephenson claims that bodies fall "exactly" on a straight line when they actually don't, and the claims other use that the four main spots are at perfectly equal distances don't appear to hold up on the maps that I've seen. On the other hand, I do have to concede that any murderer trying to form a pattern may have been using a map that didn't line up with the ones I'm looking at, and the four major locations are at the very least spread out so that they'd probably be close enough for a mad killer's purposes.

And lastly, once a pattern is identified, it's typical to declare that it had to be intentionally planned because things like that just don't happen by coincidence. Here we are talking about visual shapes, but the same sort of argument is advanced over and over in various other theories about the case. We'll discuss this whole question at the end of the article.

The cross theory was picked up and modified by various other people over the years. The major change everyone who has used it has made is to drop Smith and Tabram so the figure has four points instead of six. While there has been a movement by some to reintroduce Tabram to the list of Ripper victims, her location doesn't fit the crime scene pattern theories very well. Smith is almost never included as a Ripper victim, based upon her testimony (she survived for a few days after her attack) that she was assaulted by two or more men and the differences in the wounds. It's possible that Stephenson said those two lined up with Nichols and Eddowes based upon a map of the crime scenes published in the Nov. 10, 1888, edition of the Daily Telegraph.(Note 6) On this map, locations are marked with little daggers and numbers so it's hard at glance to tell which is which. Between that and the little inaccuracies that creeped into the illustration, it would be easy to believe that Smith, Tabram, Nichols and Eddowes formed a straight line. Of course it's likely that similar maps were put together by other papers, so it's uncertain what map Stephenson was using.

The Arrow

"But the biggest clue of all was only just becoming evident: the placement of the bodies. [...] If you draw lines from Mitre Square to the other three sites, a perfect arrow becomes apparent. If you have an intentional arrow, it must point somewhere. It goes southwest, directly to the heart of government, the Houses of Parliament at Westminster."

-Sue and Andy Parlour
(Note 7)

The arrow theory is essentially the same as the cross hypothesis, other than the fact that it connects the same four locations in a different way. Instead of connecting dots on opposite sides, it connects Eddowes' location to the three others.

Once again, one has to wonder just where the choice of lines connecting the dots comes from. It's not like Eddowes' apron was found under graffiti explaining that her body was the head of an arrow (not that the whole Juwes thing, if he wrote that, was much clearer). With four spots to choose from, you can make an arrow pointing in four completely different directions. One way isn't any more obvious or natural than another. It seems to me that someone explicitly trying to make an arrow would put two of the bodies closer to the spot they intended to be the head. Of course that could have also been interpreted as another type of cross. Without more points to work with or some other message to get the same point across, it's all rather ambiguous.

This theory again ignores Kelly's death, but with a different rationalization for doing so. Like many versions of the Royal Conspiracy, the Parlours believe that Eddowes was killed by accident in an attack that was meant for Kelly. This unlikely idea is based upon Eddowes' use of an alias similar to Kelly's name on the night of her death. But then if a group of killers had actually thought that this plan (killing prostitutes in bizarre ways over the course of a month to respond to a secret allegedly known by one of them) made sense, it's not a stretch to believe that they would accidentally kill someone who not only looks nothing like their main target but is more than a decade older.

Of course, even if that were the case you'd think that maybe one of these conspirators would, call me crazy here, read the newspapers or something and realize they had the wrong woman. Once they knew that Kelly was still alive they would naturally want to finish the job right away. Are we supposed to believe they'd just let her run around for more than a month knowing this terrible secret and potentially telling it to everyone she knew? It just doesn't make much sense.

The idea of a giant arrow over Whitechapel is probably one of the least outrageous parts of the whole theory. On the other hand, there's not really anything to support it either.


The Pentacle

"You can't outrun it, Netley. It surrounds us... This pentacle of Sun Gods, obelisks and rational male fire, wherein unconsciousness, the Moon and Womanhood are chained. Its lines of power and meaning must be reinforced, according to the ancient ways... What better sacrifice than 'Heiros Gamos'? Than Diana's priestesses?"

-William Gull,
as imagined by Alan Moore
(Note 8)

This one is another variant of the Royal Conspiracy theory but is probably even less logical.

A pentacle is a figure shaped like a five pointed star formed by connecting alternating points with line segments. It is also variously known as a pentagram, pentangle or pentalpha. As probably the most well recognized magical symbol by the general public,(Note 9) and considering that most people believe Jack killed five victims, it's not too surprising that it has been suggested as a symbol that the Whitechapel murderer was trying to form.

Without even taking a look at the locations in question, however, it seems difficult to reconcile the idea of a figure with five equal sides when others are proposing ones with four equal sides. For each of those two concepts to work, the points would have to be spread out in very different formations. In fact, the five crime scenes (now including Mary Jane Kelly) are not distributed in a way that one could make a figure with five equal sides.

Trying to draw a pentacle from these locations leaves us with a very lopsided figure. Alan Moore suggests (at least for the purpose of his story) in his graphic novel From Hell that it was made that way so that Christchurch in Spitalfields would serve as its mystical center. The problem with that idea is that the church isn't in the center of these five killings, as it is located only a short distance away from Kelly's room in Miller's Court. The map that artist Eddie Campbell drew to illustrate Moore's concept places a number of the killings in incorrect spots, most notably with Eddowes erroneously located most of the way to Goulston Street (where the apron taken from her body was found) instead of in Mitre Square where she was actually found.

This alleged pentacle is described by another writer offering enthusiastic support for a wide range of conspiracy theories centered around black magic(Note 10) as an asymmetrical pentagram. What this person fails to acknowledge is that any five points, random or not, will form an asymmetrical pentacle, as long as no three of those points line up on a straight line. As a matter of fact, Kelly's body is not all that far from being on a straight line between the places Chapman and Eddowes were found. The lack of symmetry in this supposed pentagram is quite extreme.

To sum up, if a group of killers actually had been trying to make a pentacle with the bodies of dead prostitutes in 1888, they sure did a lousy job of it.


Vesica Piscis

The most recent of the more popular theories about an alleged pattern is championed by author Ivor Edwards and fans of his theory.(Note 11) I won't go into much detail about it for the simple reason that another article in this issue is devoted to it. But, briefly, the first four canonical killings form the cross as described in Stephenson's earlier theory and then Kelly's location is used to further define an area. This creates what Edwards sees as a vesica piscis figure, an almond shaped object formed by the overlap of two circles that is symbolic of a womb, and thus fertility, rebirth and so on.

The biggest difference between this theory and the others is that it suggests a pattern composed of curves instead of straight lines. It's rather difficult to visualize a shape like this from a small number of spots without outside reference. And, similar to what happens with all the previous theories as well, once you are told what to look for there's always the possibility that you see it because you expect to and not because it was created that way intentionally.

And talk of possibilities and trying to determine whether something was random or intentional leads us into the next section...


The Probabilities


Most of the people promoting these theories have claimed that the pattern they identify is extremely unlikely to have happened purely by chance and then try to use that belief as support for their particular suspect or scenario. This general strategy was followed by assorted writers for years without attempting to actually quantify the probabilities involved. This all changed when Melvin Harris claimed to have had the statistics for his favored theory calculated. Here's how he explained it:

"Jay Clarke and John Banks, the two Canadian criminal lawyers that I mentioned in the interview, took a close look at yet another aspect of D'Onston's 1 December 1888 article in the Pall Mall Gazette. They considered his idea that there was a deliberate choice of sites leading to a giant cross straddling Whitechapel. This involved the placing of the four outdoor bodies at the essential four tips of the cross. They took this question to a university professor who specializes in statistics: How do you calculate the probability of finding four bodies randomly distributed in a city so that they form the points of a cross? They were advised to draw a map of the area and mark it with a grid: Eight squares down and eight across will do. All calculations could then be based on these squares. The results were startling. The odds against D'Onston's Ripper Cross scenario being wrong were one in fifteen million, two hundred and forty-nine thousand and twenty four. Repeat: 15,249,024! This was reached using a markedly course grid. If the squares were made smaller and increased in number, and they could quite legitimately, then the odds would soar even higher. The lawyers concluded: With odds like that, logic says the distribution was planned." (Note 12)

Others followed and used this argument themselves without really understanding where that number originated from. They assumed that if the odds for a cross randomly appearing were that bad, then the chances that, say, the points of the cross would happen to form in a specific order would be even worse. That would be a logical conclusion, provided the numbers given by Harris were correct. But they aren't.

We are lead to believe that a math professor came up with the conclusions, but what Harris says about it is quite vague on the details. The most important point the idea that the math shows the sequence had to have been planned is expressly shown as the opinion of the lawyers instead of the professor. That's important not only on its own but also because it implies that the professor may not have had much input into the other conclusions they presented either.

It turns out that Harris' statement has two fundamental errors in understanding of the statistics involved, either one of which completely invalidates this particular argument. On top of those two mathematical mistakes, he also made some other errors that would make the statistics a lot less relevant, even if they were done right in the first place, which they weren't. I'll explain all those in just a minute, but we need a little bit of mathematical background first.

The professor suggested using an arbitrary grid that is 8 squares across and 8 squares down to start with. This is a pretty reasonable basis upon which to try to calculate odds for these locations appearing randomly. What that gives us is 64 different possible squares that a body could appear in. If a murder were to happen randomly in an unplanned location, the odds that it will take place in any one square on that grid is 1 in 64. So far that's easy. The number they end up with here, 15,249,024, comes out to be 64 x 63 x 62 x 61. That means that, for the sake of calculating the statistics, they decided that once someone was killed in one square that the killer wouldn't kill a subsequent victim in the same square. So first there are 64 possible end results for the initial crime location, then there are 63 left because one square is used up, then 62, then 61. That's where this long number comes from. It's the odds of that exact sequence of specific locations in that specific order.

Unfortunately for Harris' argument, that's not only not the odds of forming a cross on a grid by chance (which is much, much more likely to happen on a purely random basis), it's also completely meaningless as a way of determining whether the distribution was planned or not. Let's take a look at the second of these fundamental flaws first.

Coming up with the statistics for specific sequences like this proves absolutely nothing. What you end up getting with this method is the mathematical equivalent of circular reasoning. When doing calculations of this sort, the size of the number alone in no way proves whether something was random or not. It just proves that there are lots of different possible outcomes. There is no valid threshold using this method at which you can legitimately claim that something must have been planned.

This is a difficult concept for the average person, who typically doesn't have a background in statistics, to understand. It's probably easiest to explain with the use of an example.

Let's go back to this imaginary grid. If it has 64 squares on it, it's basically a typical game board on which you could play chess or checkers. If you took a board like this out and set it up on the floor, you could randomly drop checkers, coins, or other small objects on it. Assuming that you ignored any that fell off the board or landed on a square that wasn't empty, you could quickly end up with a sequence of four specific squares in a specific order, representing where and when the checkers or coins landed. If you wrote your results down and then figured out the odds of that happening, you'd find out that there is only a 1 in 15,249,024 chance each time you do this experiment that you'd end up with that specific sequence.

Wait a minute, wasn't that the same number Harris was using? And wasn't his argument that, because the odds of that happening were so low, it must have been planned ahead of time? What, were you trying to aim for specific squares? If you aren't sure, wear a blindfold, number the coins, and drop them again. Whatever your results are, the odds of that specific sequence of four spots happening in that order was again only 1 in 15,249,024!

Does that number prove that the results couldn't be random? Absolutely not. On the contrary, it's mathematically exactly what you'd expect from a random result in that situation.

Harris also claimed that using more and smaller squares on the imaginary grid would increase the numbers dramatically. Well, that's true, but it's also meaningless. Say you set up four chess boards together so they formed a square and pretended that the area represents the East End of London. The odds of one killing (or dropped coin) ending up on any specific square is now only 1 in 256. Plot any four killings (canonical or non-canonical ones, it doesn't matter) or drop coins onto it and the odds for that specific sequence showing up are only 1 in 4,195,023,360. The number changes as a result of the way you model it, but that doesn't suddenly make it any less random.

If that weren't bad enough, there's also the other fundamental flaw: There's more than one way to make a cross on a grid. You can't talk about the statistics of one specific cross pattern as if they were statistics for forming a cross in general. The larger the number of different sequences that can form a pattern, the easier it is for it to happen completely by accident. To prove this, let's start by actually showing the locations of the four victims in question on an arbitrary grid. In order to visualize the patterns easier, I'll rotate the orientation so the cross shape lines up with the rows and columns of the grid.

The murders happened in order with Nichols first in the square with her circle in it, then Chapman second with her square's circle, then Stride and lastly Eddowes. This sequence does form a kind of cross on this grid, but that's not the only cross that could show up at random in that grid not by a long shot.

First up there are crosses that are like the current one but moved horizontally or vertically. Imagine that the murders had happened in unplanned locations, but one square to the left of the current ones on this grid. That's shaped and formed exactly the same way, but it's another one out of those 15,249,024 possible end results, and it should also count. OK, so what about one space to the right? Two spaces to the right? Those both work too. How about one space up? And then there are all the cases where you go one space up and also either to the right or left. Already the odds are 8 times more likely than the lawyers told Harris, and we're just getting started.

There are a lot more ways to make a cross fit in the grid. You can change the size of the cross. You can change the location of the crossbar so it is farther up or down the length of the other side. You can rotate the cross. You can also make the crossbar longer or shorter in comparison to the other side. And you can do all of those in combination, so you might have the cross smaller, rotated, moved over and shaped more like a typical Christian cross with the smaller bar father up the length. (Although if you were doing all of these for real, you'd have to be careful, because some combinations would end up repeating themselves. Rotating the current cross 90 degrees, for example, is the same as lengthening the horizontal bar and shortening the vertical one.)

If you add these all up, you'll find that the really large number we had to start with is getting whittled down very quickly. And there's another thing to consider that also makes it easier to make the cross.

So far we've been assuming that the killings start with Nichols and then progress the way they did historically. That's all fine and good, but if we're trying to find the odds of a cross happening randomly, the order of each of the four points makes no difference at all. If the killer had started with Eddowes, gone to Nichols, then Chapman and then Stride, we'd still have the exact same pattern on the map. It's the odds of that pattern happening by chance that we're concerned with, so we need to take that into account. In this case, there are 24 different sequences that can be made by switching the order around on four events. This number comes from four different points you could start at, times three that are left to choose from for the second location, times two that are left after that, and then whatever one remains at the end. That's 4 x 3 x 2 x 1. That means for each of the different crosses we can come up with on this grid, there are 24 possible ways they could be put together. (But, again, for some shapes there would be an overlap between rotating the figure and changing the order of the points of the cross. You'd need to keep that in mind if you tried to count all the variations by hand.)

Taking into account all the many and varied ways a cross could show up purely by chance, the final odds that I calculated for the probability of four randomly distributed bodies forming the points of a cross would be a not too terribly bad 1 in 100, roughly. This is a much more accurate measure of the odds of the pattern forming than what we started out with. One way we can tell that it's more mathematically sound is that changing the grid size doesn't make such extreme changes in the end results. Increasing the grid to 9 squares across by 9 squares down gives about 1 in 110 odds.(Note 13)

Those odds would be for any cross with a perpendicular crossbar that isn't at the edge. So tau crosses (T-shaped) and crosses with lines that are not at 90 degree angles to each other aren't included in these numbers. If they were included, the possibility of a cross forming by pure chance would be even more likely.

Of course even there those numbers have to be taken with a grain of salt, because they are just the odds of a specific pattern we are looking for. If the killings were random and they had happened differently, then we might be looking for the odds of an F appearing, or a Y, or a T, or a straight line, or any other number of patterns that could have happened. Odds are good that no matter how the bodies would have fallen, someone out there would have convinced themselves that they saw a pattern of some sort.

I also had mentioned that there were some mistakes that would make these numbers less relevant than they appear to be. One of the most immediately obvious ones is that the people who see patterns are picking and choosing which victims to include and which ones not to. Mary Jane Kelly is typically excluded from the cross making odds for the simple fact that she doesn't fit. If you are doing best four out of five to try to make a pattern, the odds of randomly making a cross, or any other symbol for that matter, go up quite a bit.

Another thing that affects how the statistics play out is that it's likely the killer wasn't being completely random in choosing locations even if he didn't try to make a pattern. The method we've been using assumes that the Ripper wouldn't kill another victim in the same square (or general area, since these squares are only hypothetical) that a previous one had been killed in. If that's a reasonable exception to the concept of pure randomness, which it probably is, one might also suggest that a murderer would want to pick a site at least two squares from the previous one. Even a rather straightforward modification like that vastly increases the odds that some sort of pattern can be identified.

Looking at the psychology of it, the Ripper's real rules were probably more complicated than that. He might have decided, for example, that having a new murder site close to another was less risky if it had been a while since he last killed someone. Or, if he were experiencing hallucinations, it could have been largely based upon which of the gas lamps seemed like they were possessed by demons, or some other strange concept that we couldn't know about unless he had been caught. You never can tell what could influence the mind of a killer.

One important consideration when choosing sites would be where the main roads are located relative to where you intend to kill people. One might point out that Jack obviously wouldn't attack someone in the middle of Whitechapel Road where anyone could be watching. That might lead one to try to find the squares on the grid that contain a major road and remove them from consideration when calculating our odds. Of course looking at the grid with this in mind makes it clear that we can't exclude those squares without ignoring some of the very areas the crimes happened in!

No, Jack didn't kill anyone in the middle of a major street, but the locations were quite close to them. This could be so he could make an easy escape, or it might just be that the prostitutes were looking for customers where foot traffic came by on a regular basis. Taking this one step further, looking at the map shows that the main roads, although somewhat irregular, mostly form the shape of a cross all by themselves.

In other words, most of the features of the patterns people see in the crime scene locations were already present in the layout of the East End before the Ripper killed his first victim.


Odds and Ends

So we've looked at the statistics and the layout of the streets and concluded that apparent patterns behind the sites could have been completely random. Does that necessarily mean that these theories are nonsense?

No, actually. Showing that something could have been accidental by itself doesn't prove that it was accidental. For example, if a person (or group, if you wanted to go the conspiracy route) did want to create a symbol by killing people in specific spots, it would make sense to do so working with the street design in mind. There really isn't a reliable way to answer the question of whether they were planned that way or not. If a suspect could be traced to a prediction of a specific symbol or one or more exact locations before the killings took place, that would at least be a pointer in favor, while specific pointers against would have to be something along the lines of one of the killings happening in a location that would have been impossible to plan for. No examples of such a thing come to mind. Proof, as all too often happens in this field, is frustratingly elusive.

The main argument against these theories is that the process of selecting sites beforehand would add an extra layer of complexity in trying to successfully pull off these already rather audacious murders. It's one thing to simply place a dead body in a specific location, but it's a whole other thing to not only kill them there but also to gut them out in the open where someone could wander along at any moment. Many authors have commented on the amount of sheer luck or cleverness the killer had to have possessed, but to do it all with a master geographic plan in mind would be just that much more difficult. Of course that doesn't make it impossible.

The best hope to trying to sort it all out is to look at the rest of the details of a particular theory and see how well they hold up on their own. Other than the parts already discussed briefly as we went along, that kind of analysis is beyond the scope of this article. If you are interested in the rest of the details of these theories, I suggest picking up the books that support them.
Keep in mind that, although I've been specifically talking about whether the crime scene locations were random or planned, the same debate of whether some detail in the Ripper case is significant or just a coincidence is played out all the time over different aspects.

Some people think they see letters in the middle of bloodstains on the wall in Mary Jane Kelly's room in the famous photo of her on her bed. The police at the time, as well as many others later, believe that the graffito found near where the piece of Eddowes apron was dropped must hold some key significance. Patricia Cornwell claims that Walter Sickert must have been Jack the Ripper because of certain aspects of his paintings she considers significant, or based upon a mathematical possibility that partial DNA on a letter from someone claiming to be the killer may have been Sickert's.

Each of these arguments wants you to believe that some perceived significance is real instead of merely coincidental. Out of all the different potential conclusions that are offered up in this case, some are probably right and most would have to be wrong, simply based upon the sheer number of them which are contradictory. The trick is to try to figure out which is which.

When coming to your own conclusions about what is or is not a significant detail, don't underestimate the power of sheer coincidence.

 

NOTES:

1. Anderson, George K., The Legend of the Wandering Jew, Brown University, 1965. (Back)

2. As quoted on the JTR Forums website, www.JTRForums.co.uk (Back)

3. The article, titled "The Whitechapel Demon's Nationality: And Why He Committed the Murders," was written under the byline of "One Who Thinks He Knows." The writer has been identified as Stephenson based upon comments that the paper's editor, W.T. Stead, made at a later date, as well as similarities between other parts of the article and a letter he wrote to police in October. (Back)

4. For a copy of the inquest and period newspaper articles about this incident, see: www.casebook.org/victims/whitehal.html ­ take note that there's only one L in Whitehall in that address. (Back)

5. For one who does propose a link, see R. Michael Gordon's The Thames Torso Murders (McFarland & Co., 2003), which attempts to link a series of similar crimes to his Ripper suspect, Severin Klosowski. (Back)

6. See the second page in the last photo section of Evans, Stewart P. and Skinner, Keith. The Ultimate Jack the Ripper Companion, Carroll & Graf, 2000. (Back)

7. Parlour, Sue and Andy, "The Jack the Ripper Whitechapel Murders," The Mammoth Book of Jack the Ripper, Carroll & Graf, 1999, edited by Maxim Jakubowski and Nathan Braund. (Back)

8. Moore, Alan & Campbell, Eddie, From Hell, Top Shelf Productions, 2004, collected reprint of a comic book series started in 1989. (Back)

9. In fact, it's so iconic of the concept of black magic that one is featured on the cover of Ivor Edwards' Jack the Ripper's Black Magic Rituals (John Blake, 2003) even though that book offers the theory that Jack was drawing a vesica piscis symbol with the bodies and not a pentacle. (Back)

10. See: econcrisis.homestead.com/JTR2.html (Back)

11. See the previously referenced JTR Forums website or Edwards' book. (Back)

12. Harris, Melvin, "Roslyn D'Onston: An Exclusive Update," Ripper Notes, March 2000. (Back)

13. This number was arrived at in the following way: The first step in making a cross is to draw a line. Any two dots, random or not, can be used to draw a line, so the first two don't matter statistically. Then we can make a grid around two dots. I'll use 9 squares across by 9 squares down since locations on an odd numbered grid are slightly easier to explain. It's a hypothetical construct, so we can place it however we like. The two dots are put in column 5 (middle column, with four others on either side), in the top square (row 1) and bottom square (row 9). Any square that lands in rows 2 through 8 that doesn't end up in column 5 can be the next point in the cross, so the odds of one of those happening are 56 out of 79 (a 9x9 grid has 81 squares, but two are already used). Now, on the last dot, it has to specifically match the third one, but on the opposite side of the middle column. At this point only one specific square will work, so the odds of that are 1 in 78 (81 minus the three squares already used). Multiply 56/79 by 1/78 and you get 56/6,162, or about 1 in 110. (Back)

SOURCES:

Anderson, George K., The Legend of the Wandering Jew, Brown University, 1965.

Begg, Paul, Jack the Ripper: The Facts, Robson Books, 2004.

Cooper, Geoff and Punter, Gordon, Jack the Ripper Whitechapel Map Booklet 1888, ripperArt, 2003.

Edwards, Ivor; Hebblewhite, Tyler and Brown, Howard, JTRForums messageboards, www.JTRForums.co.uk

Evans, Stewart P. and Skinner, Keith. The Ultimate Jack the Ripper Companion, Carroll & Graf, 2000.

Harris, Melvin, "Roslyn D'Onston: An Exclusive Update," Ripper Notes, March 2000.

Jakubowski, Maxim and Braund, Nathan, The Mammoth Book of Jack the Ripper, Carroll & Graf, 1999.

Moore, Alan & Campbell, Eddie, From Hell, Top Shelf Productions, 2004, collected reprint of a comic book series started in 1989.

Ryder, Stephen P., Casebook: Jack the Ripper website, www.casebook.org


Related pages:
  Dan Norder
       Dissertations: Dr. Anderson at New Scotland Yard: An 1892 Interview Redi... 
       Dissertations: Heartless - The Evidence for a Copycat Killer