To give the impression he wasn’t from Vallejo?

I can see this question cutting in both directions, to be fair.

And just the one letter to the Vallejo newspaper.

This is the thing right, in the Fincher DVD extra, the geographic profiler Dr. Kim Rossmo says that before analysing the case, he thought that the Zodiac crimes were SF crimes. Upon inspection he finds that they are Vallejo-based crimes (obviously you’d have to control for the Stine murder as somewhat of an outlier).

By sending letters to the large circulation newspapers in SF (and one to the LA Times) killer is distorting his profile. In effect, it’s like he’s puffing himself up, inflating himself to make himself scarier and more dangerous than he actually was. The bomb letters do this further, by making the killer’s area of operation anywhere and everywhere within an 7-mile radius of Mt. Diablo.

Rossmo’s heat-map is available here: https://geographicprofiler.com/crimes/zodiac-killer

Also of note, the first part of the first cipher was sent to the Vallejo Times-Herald hinting that Vallejo was more familiar to him.

Second, I’ve seen Rossmo’s map before, but I’m having trouble reading it. The yellow, oranges, and reds are areas he most likely lived? Or vice versa?

Also of note, the first part of the first cipher was sent to the Vallejo Times-Herald hinting that Vallejo was more familiar to him.

I ought to re-read this letter.

Second, I’ve seen Rossmo’s map before, but I’m having trouble reading it. The yellow, oranges, and reds are areas he most likely lived? Or vice versa?

If you click on the link above there’s various models by which the data can be represented (Centrography, Guassian Rossmooth, Logarithmic, Negative Exponential, Quadratic, Rossmo, Truncated Negative Exponential). Rossmo also mentions in the video that the variables can be adjusted to tweak the output, depending on how the data is to be interpreted. He says that geographic profiling, although an impressive tool itself, is only secondary to forensics, witness testimony, etc.

I’m guessing data points have to be significant events, based around principal evidence (sending letters for example, would be secondary). Obvs the map strongly points to the activity of a Vallejo-based resident (I think we know who), yet since that suspect couldn’t be matched by forensics, the profile has to take a backseat. Fresno St is definitely in the warmer sections of the map, depending on which model of evaluation you use.

Interestingly, because GSK was a forensically trained cop, he messed with his own geographic profile. He would up set up offending patterns in one area with a particular MO, yet would hit another area with a different MO to make the signature look different, when he wanted to commit a different type of crime. He also knew which areas to avoid, in terms of police patrols and also skipped the university area, because he knew they had own separate security.

Some people have suggested that Zodiac liked the area between Vallejo and Benecia because it wasn’t fully incorporated, jurisdictionally. I don’t know about that however.

Why Graysmith goes into ALA into Santa Rosa so much, is because LE’s geographic profiling was telling them to do look at him again.

Why Graysmith goes into ALA into Santa Rosa so much, is because LE’s geographic profiling was telling them to do look at him again.

Not at all.

Does Rossmo’s map take into account the other aspects of the Zodiac case apart from the attacks I.e. letters, ciphers, wording etc.

Does Rossmo’s map take into account the other aspects of the Zodiac case apart from the attacks I.e. letters, ciphers, wording etc.

I was just wondering the same thing. I might try to look for a paper and see if Rossmo has written on it. I’m a sociologist, so have an MA in quantitive analysis, but it isn’t my favourite subject by a long shot.

The variables you mention could easily be added to the model; or alternately, you can weight the variables by scoring their importance (e.g. 10 for a murder, 5 for a phone call, 3 for a suspect sighting, 1 for a mailing). Also you can periodise events, to give different profiles. There would likely be accepted standards and methodologies for profiling.

Usually by slicing the data in different ways, you make discoveries and then evaluate that using theory or comparing to other cases. Obviously you want to keep it fairly simple, since adding complexity can muddle your data. I’d be wary of including non-canonical crimes into any model, for instance.

I don’t think its a substitute for other types of traditional detective work, but can help narrow the field of suspects.

EDIT: btw the full text of Rossmo’s study is here: https://www.scribd.com/document/1812896 … ssmo-Ph-D#

Here is access to the whole analysis, which is cut off in the Scribd link
https://www.zodiacciphers.com/photographs-14.html

I still don’t know if the brighter colors indicate probability of residence or the opposite lol

Why Graysmith goes into ALA into Santa Rosa so much, is because LE’s geographic profiling was telling them to do look at him again.

I’ve read your post four times now and it gets more confusing each time I read it.

I still don’t know if the brighter colors indicate probability of residence or the opposite lol

Here is an explanation Chaucer
https://www.zodiacciphers.com/zodiac-ne … -profiling

I think Rossmo might be wrong in this case. Obviously he’s an expert and I’m not. Still …..

Rossmo’s analysis is conventional. I don’t mean that in a bad sense, since conventional is right more often than not. But his framework is that Zodiac lived near Vallejo, went hunting in the surrounding areas, until he was ready to strike, etc. Then he spreads his activity outward from there.

There’s another way to look at it. It’s what I call the Paul Holes Theory. Holes hasn’t actually pontificated on where the Zodiac lived specifically. But during the Golden State Killer investigation, he felt sure that GSK lived in Sacramento and never moved to Southern California, despite the crimes moving in that direction. He looked at the locations and noticed that the Sacramento crimes were deep in the neighborhoods. Whereas the SoCal crimes were typically not far off major roads, in turn not far from major highways.

If you use the Holes logic in this case …. if you drive on 101 across the Golden Gate Bridge, eventually it forks into Highway 37 that goes through Vallejo, Then 37 stops being a highway and turns into Columbus Parkway just a little ways up the road from Blue Rock Springs Park. Beyond that, it’s a short drive with only 1 turn to the Lake Herman Road site. Meanwhile, in the Stine murder, the killer goes deep into neighborhood. The Holes logic makes a lot of sense here, and points to LHR and BRS being commuter killings, while Stine is deeper in a familiar neighborhood.

Also, Rossmo says that he doesn’t think Zodiac hid in the playground. He went in a different direction to a likely parking place. But then that raises the question …. if he didn’t go to the playground, and he doesn’t live in that area, then how does he know the playground exists?

if he didn’t go to the playground, and he doesn’t live in that area, then how does he know the playground exists?

Maybe he read Gaikowski’s newspaper.