Possibly... I genuinely do not know. In stats each technique is good for some things and not others. It may be that we can use language but are not good at knowing what it can and cannot be used for. In stats it can be clear when it has been used wrongly, maybe not with language
Does this mean that I will have to understand language before I use it as a tool?
Well, *you* would say that, you are always far too positive! ;-)
After a premptive raid on their assumptions? ;-)
New paper in which I played a very small part. academic.oup.com/bioscience/a...
#abm Edmonds, B. Carpentras, D. and Chattoe-Brown, E. (2025) 'Strongly Empirical Modelling' in Czupryna et al. (eds.) Advances in Social Simulation (Springer), pp. 99–111. doi.org/10.1007/978-...
New Contribution: A Reminder – computability limits on “vibe coding” ABMs by Bruce Edmonds. rofasss.org/2026/02/12/v... #rofasss #ABM #agentbased #programming #LLMs #AI #vibecoding #computability
A pre-print in which I played a very small part:
Rosello, C., Hamilton, S./ H. and Grimm, V. et al. (2026) FAIR principles for good modeling practice and standards. Available at SSRN. ssrn.com/abstract=621... or dx.doi.org/10.2139/ssrn...
Book just published (Free to download). This collects together work done by the groups at the Research Incubator held in Groningen in the summer of 2023 which I participated in. link.springer.com/book/10.1007... #abm #polarization #socialscience #cohesion
The other function of collecting data for any researchers to analyse/use is very much downplayed. Indeed it is hard to get access to much good qual data.
Many qualitative presentations I have heard stress the "give a voice" story and then cherry-pick "interesting" examples from their interviews, ignoring the majority of more mundane comments.
Yes, a major function of qualitative methodology (and particularly its language) is to give more importance to its findings. [this complements the major function of quantitative methodology (and its language) which is to give the impression of more rigour & reliability of results than is warrented]
New Contribution: "Why Object-Oriented Programming is not the best method to implement Agent-Based models" by Martin Hinsch. rofasss.org/2026/02/03/oop #oop #abm #objectoriented #agent-based #agentbased #rofasss #simulation #modelling
Sorry! This is a real pain, but it follows "Edmonds law of simulation" - that results become more boring the more bugs you eliminate ;-)
"10 simple rules for good model sharing practices" journals.plos.org/ploscompbiol...
New joint paper that only took us 3+years to get published! This inferrs the cognitive model of agents from data.
Batzke, et al. 2025. From cognitive coherence to political polarization: A data-driven agent-based model of belief change. doi.org/10.1111/pops...
"Consipracy theory" is a bit of a misnomer. What seems to characterise anti-vaxx, JFK shooting etc. is the asymettry to evaluation as noted above - nit picking contrary evidence accenting evidence for unquestioningly. Of course the ultimate denial of contrary evidence might be "its a conspiracy"
Though the gap between prediction rhetoric and achievment is widespread in science - serious and otherwise.
promiscuous associativity? This has the advantage of sounding clumsy and so might be taken up by philosophers ;-)
Maybe not a "pangloss journal"... but "Pangloss Reports" probably has a future 😉
Mind you, simplistic discussion of the economy is endemic among politicians - rather like arguing over how to fix a complicated engine by pressing the brake or the accelerator :-/
He got there, finally with health care... ‘Nobody knew health care could be so complicated’. He just needs some time to learn via trial and error.
new one developing...
- avoiding the spend/tax trade-off by growing the economy
With an increasingly turbulent society, the demand for social scientists who are capable of understanding and analysing what is happening using computational methods is rising. New free online course on polarisation! Good tonic for the times. Available to all. www.youtube.com/watch?v=6vmV...
OK, but the categorisation by YouGov might be who they last voted for and their opinion based on the current policies/rhetoric. That is reasonable
Post by Thorid Wagenblast, Nick Roxburgh and Alessandro Taberna on "Modelling Deep Structural Change in ABSS" #abm #socialsimulation #rofasss
Sorry, what has considering competence have to do with registering failed predictions? Are you thinking about a particular incident here?
