The value of a model lies less in its predictions than in its capacity for differential revision.
End states are largely accidental. Dynamics are exceedingly difficult to get right.
Box’s Axiom: “All models are wrong, but some are useful.”
Since the release of the first ECDO Hypothesis on 16 February 2020, a number of modelers have stepped forward to offer their own perspectives on how Earth’s geomechanics might fit into an ECDO framework. To date, six individuals or groups have undertaken the task of developing simulations of varying complexity to test the fit between Earth’s geophysical and geohistorical record and an ECDO paradigm. This is a fantastic evolutionary process and I am very appreciative and supportive of all these efforts to date.
The purpose of this article is not to dictate how such efforts should proceed, but rather to offer some guidelines.
My own experience bears the patina of five decades spent developing, applying, and testing models in the real world. I am admittedly less versed in the vast array of modern simulation packages and AI-assisted tools that have become available in recent years. Development of the research is far more important to me at this juncture, than the task of learning a new modeling/simulation package or API.
Nonetheless, a few fundamentals of modeling do not change, no matter how sophisticated the tools become. A scratch golfer remains a scratch golfer whether playing with an old set of Ben Hogan irons or the latest Cobra 3DP X clubs. The inexperienced player, by contrast, often speaks at length about equipment and metrics as a substitute for skill. This can become a distraction.
The same thing occurs in modeling.
Sophisticated tools frequently conceal overly simple assumptions, and may well expertly communicate an entirely invalid or incoherent outcome.
Most ECDO models I have seen fail not because of inadequate software, but because they attempt to force a highly complex geophysical system into a framework which is too simple to reproduce the observational set.
Before continuing to mature ECDO modeling itself, therefore, it is worthwhile to establish a few principles regarding models in general.
On Modeling
My experience running thousands of them in the real world:
TES’s Black Box Paradox
If you don’t understand why a model produces its answer, should you trust it?
And if you do understand why it produces its answer, why did you need the model in the first place?
The above, of course, is not a mathematical paradox, but a philosophical one. Nor does this tension serve to invalidate efforts to create models and simulations. Rather, its purpose is to compel the modeler to confront two questions:
What has actually driven the answer I have obtained?
And,
Am I using a sophisticated tool merely to repackage and conceal a simple a priori assumption?
These are the questions every model developer must keep firmly in mind while applying their tools.
Corollary I — The Convergence Corollary
Models generally produce one of three outcomes.
Multiple Optimal Outcomes
A model yields several competing solutions or equilibria.
This indicates either that the problem remains underdetermined, or that assumptions require further discrimination—or simply that a business or judgment call must be made.
Incoherence
The model fails to reconcile and/or produces contradictory results.
Such failure may indicate ignorance, incompatible assumptions, changing conditions, or an apples-and-oranges comparison.
The unappreciated irony, however, resides in this: The incoherence itself is often informative.
A Single Optimal Outcome
The model converges upon one answer.
Such convergence may indeed represent a true optimum.
Or it may merely reflect unsound assumptions, hidden biases, or sensitivity blindness.
Paradoxically, the most suspect outcome is often the appearance of a single optimal solution.
Corollary II — The Failure Corollary
Because all models are wrong (see Box’s Axiom), failure does not imply lack of utility.
Indeed:
The catastrophic failure of a model may be more informative than the model itself.
Corollary III — The Adaptation Corollary
A model’s utility is measured less by its end-state accuracy than by its ability to survive contact with contradictory observations and constraints.
Or:
The value of a model lies less in its predictions than in its capacity for differential revision.
Conformance to reality is found primarily in a model’s dynamics—not in any particular end state.
End states are largely accidental. Dynamics are exceedingly difficult to get right.
Corollary IV — The Complexity Corollary
If you use too few variables to characterize a phenomenon, you have little hope of capturing its full complexity.
If you use too many, you can forecast virtually any future you desire.
On ECDO Modeling
The purpose of this step in the systems analysis process is to move the discussion away from
“Here is my model.”
and toward
“Here is the language within which competing models may be expressed, leveraged, and compared.”
That is what systems groups do before attempting to develop model fabric.
This is the proper role of dynamic modeling—and at present there are at least six highly motivated groups pursuing that effort. I am impressed with each and every effort.
My own role is not to be the modeler. It is to provide coherence to the effort. As managing partner in a model-development practice, and president of a large systems integration corporation, this has been my job for some time.
After defining the ECDO problem (Cunningham, Inversion, “Define the Problem—Then Seek Falsification,” ch. 3, p. 15.),1 a Wittgensteinian system description should precede any attempt to formulate constraints or differential equations.
Before constraining equations come objects.
Before modes come solutions; before solutions come states.
Before simulations come coupling coefficients and dynamic variables.
These things must first be identified and assigned coherent meanings. Otherwise, one is left with a collection of IITPW models whose variables, assumptions, and state definitions differ in ways that preclude meaningful comparison.
Accordingly, the following two pages constitute a proposed system description for ECDO States 1 and 2 and their transitions.
They are not intended to be equations. Nor are they intended to represent a complete model.
Rather, they are meant to define the principal entities, coupling factors, state variables, and dynamic parameters which should be considered before any mathematical treatment or simulation is attempted.
The objective is simple:
To establish a common language within which competing ECDO models may be constructed, tested, falsified, and compared.
Thoughts are appreciated.
Modeling Factors for ECDO Exothermic Core-Mantle Decoupling
What permits the ECDO mediated Dzhanibekov rotation event?
Modeling Factors for ECDO States 1 and 2 – Along with Their Transitions
Given permission, what are the ECDO system rotation dynamics?
Simulation Success Touchpoints
End states are largely accidental. Dynamics are exceedingly difficult to get right.
Forty geohistorical touchpoints are offered as benchmarks against which competing ECDO models may be evaluated.
The underlying premise is simple: the greater the number of touchpoints a model successfully reproduces, the more confidence one may place in its validity within an ECDO context. This does not maek the model = truth. Rather it simply identifies those models that produce a complext set of dynamics, that just so happen to emulate the complex dynamics of Earth very well.
No single observation proves the model. Rather, credibility emerges through the coherent reproduction of an increasingly broad set of independent geophysical and geohistorical observations.
Below are my suggested touch points for model evaluation (along with the test latitude and longitude for those packages using GCS-specific geographics):
- Kalahari Desert (Return to State 1 potential) – 26°47’03.63″S 19°43’02.41″E
- Ogaden Desert – 8°21’31.88″N 48°08’42.56″E
- Rub’ al Khali/Arabian Desert (Return to State 1 potential) – 21°07’08.80″N 51°25’21.82″E
- Syrian Desert (Return to State 1 potential) – 32°22’52.39″N 40°42’05.01″E
- Thar Desert – 26°20’59.54″N 71°54’36.67″E
- Taklamakan Desert – 39°05’07.05″N 83°01’16.75″E
- Gobi Desert – 40°51’32.60″N 102°53’15.79″E
- Great Victoria Desert – 25°01’29.68″S 126°55’45.24″E
- Russian Black Soils Displacement Range – 58°13’42.41″N 62°27’59.30″E
- North American Black Soils Displacement Range – 54°24’35.89″N 120°25’00.69″W
- Great Basin Desert (US) – 37°16’40.60″N 110°41’54.45″W
- Vizcaíno Desert (Baja) – 27°29’21.18″N 113°36’50.98″W
- Patagonian Desert – 43°46’33.68″S 68°16’04.43″W
- Salar de Uyuni Salt Flats (Bolivia) – 20°11’20.17″S 67°34’40.43″W
- Bonneville Basin/Salt Lake (Utah) – 40°32’36.81″N 113°33’26.64″W
- Doggerland (Emergence) – 54°41’30.89″N 3°32’10.07″E
- Cuban Underwater Formation (Emergence) – 22°07’54.86″N 85°07’41.13″W
- Chihuahuan Desert – 23°37’24.31″N 102°13’55.00″W
- Eastern Antarctica – 71°55’34.56″S 63°41’21.62″E
- Western Antarctica – 74°59’59.32″S 73°36’13.84″W
- Emi Koussi Pass (Sahara Desert) – 18°16’46.31″N 20°13’02.82″E
- Giza Plateau Egypt – 29°58’38.86″N 31°08’04.68″E
- Sardinia Island – 40°01’21.47″N 9°06’35.35″E
- Grand Erg Oriental (Sahara Desert) – 32°03’41.22″N 6°44’27.64″E
- Mauritanian Slide Formation – 17°54’21.96″N 17°04’07.81″W
- Cape Verde Islands (Emergence/Atlantis) 15°11’44.52″N 23°42’09.49″W
- Bahama Bank (Emergence) – 24°17’54.59″N 78°11’05.62″W
- Galapagos Land Bridge (Emergence) – 1°05’27.42″S 85°55’32.83″W
- Easter Island Land Bridge (Emergence) – 25°56’36.91″S 100°34’19.77″W
- East Euler Axis Survival Point – 0°00’00.00″S 120°55’35.14″E
- West Euler Axis Survival Point – 0°00’00.00″S 59°47’30.39″W
- Caspian Turan Depression – 44°34’04.25″N 59°51’58.29″E
- Zealandia (Emergence) – 49°19’25.31″S 170°05’46.64″E
- Lemuria (Emergence) – 10°44’56.00″S 61°09’10.24″E
- Lake Tuz Salt Flats Anatolia – 38°41’15.33″N 33°26’59.04″E
- Faiyum Basin (Egypt) – 29°22’05.52″N 30°49’13.34″E
- Erg Amatlich Desert Striations Mauritania – 18°55’29.92″N 13°50’00.45″W
- Azores Bank (Emergence) – 38°00’47.99″N 29°09’33.71″W
- Canary Islands (Emergence) – 28°03’37.00″N 14°40’41.55″W
- Mu Civilization (Emergence) – 11°20’14.58″S 148°48’25.87″W
To all our highly-motivated, intelligent, and creative modelers: “Well done, and I look forward to continuing to work with each of you.”
The Ethical Skeptic, “On Modeling ECDO Theory”; The Ethical Skeptic, WordPress, 20 Jun 2026; Web, https://theethicalskeptic.com/?p=117051