Fig 1.
Relative agricultural productivity.
The modeled Afro-Eurasian geography used in the simulations. Relative agricultural productivity due to improved technology available to agrarian states in different eras are highlighted by different colors. Map and biome data republished from [16] under a CC BY license, with permission from Proceedings of the National Academy of Sciences, original copyright 2013.
Fig 2.
Agricultural regions within nominal strike zones highlighted for nomadic tribes (red, 200 km) and confederations (black, 550 km) in different eras. Agricultural regions and steppe colored as in Fig 1. Map and biome data republished from [16] under a CC BY license, with permission from Proceedings of the National Academy of Sciences, original copyright 2013.
Fig 3.
Mean large-scale polity density for the Old World at the end of each 1000-year interval (A, B, and C) and over 3000 years (D) beginning in 1500 BCE. Left panels reflect historical data from [16] for large-scale states (≥ 10 regions); right panels reflect model predictions. Darker grey indicates available agricultural regions in each period; light blue indicates steppe. Map, biome and historical data republished from [16] under a CC BY license, with permission from Proceedings of the National Academy of Sciences, original copyright 2013.
Fig 4.
Predictions for the Old World based on 32 trials (polities ≥ 10 regions). A: Predicted mean and standard deviation of total area under all large-scale polities (red) and agrarian states (magenta) compared with historical values per century from [16] (black and blue, respectively). B: Predicted mean and standard deviation of total population (red) compared with historical values every 20 years from KK10 (black). C: Predicted mean and standard deviation of cumulative (red) number of large-scale polities surviving 50 years compared with historical values (black) from [16]. D: Mean and standard deviation of spatial R2 metrics for 500 (blue), 1000 (red) and 3000 (black) year intervals.
Fig 5.
Polity size and duration distribution.
Historical (A) and mean predicted (B) distributions of peak (agrarian and nomadic) duration (in centuries) and peak size (in millions (M) of km2) between 1500 BCE and 1500 CE.
Fig 6.
Mean and standard deviation of predicted versus historical Old World polity sizes for the largest 3 and largest 20 states (agrarian and nomadic) between 1500 BCE and 1500 CE.
Fig 7.
Effective military distance to agricultural regions (dark grey) from four ‘capitols’ (Rome, Babylon, Mumbai, and Beijing) under different discount factors for long-distance sea and desert travel. Colors indicate how distant ‘200 km’ (green) ‘600 km’ (red) and ‘1000 km’ (blue) appear for given a set of efficiency factors. Darker grey indicates agricultural regions; light blue indicates steppe. Map data republished from [16] under a CC BY license, with permission from Proceedings of the National Academy of Sciences, original copyright 2013. As Beijing enjoys no sea or desert travel advantage its distances remain unchanged in each panel. A: No travel advantage; the distances from the western capitals resemble Beijing’s. B: Mediterranean sea travel discounted to 30% of normal overland travel; Rome is able to reach territories along the lip and deep into Europe; Babylon has some access. C: Near East desert travel discounted to 30% of normal overland travel allows Babylon and Mumbai to interact with one another if powerful enough. D: Combined discounted Mediterranean and Near East travel, used in the nominal simulation.
Fig 8.
Predicted polity densities given different effective military distances.
Corresponding impact over 32 trials on 3000-year predicted mean polity density of different discount factors applied to the western Eurasian sea and desert transport shown in Fig 7. Darker grey indicates agricultural regions; light blue indicates steppe. Panel D and polity density scale is identical to Fig 3D, right panel. Map data republished from [16] under a CC BY license, with permission from Proceedings of the National Academy of Sciences, original copyright 2013.