Empire: A Data-Driven Thought Experiment

This didn’t begin as a project . Just idle curiosity and a vague question— the kind that gets you going – if you’re both a history nerd and newly curious about the capabilities of Generative AI.The question was simply this:  How do you compare the achievements of history’s great empires, fairly? Not just by size or fame, but by what they accomplished given their time, tools, and constraints.
So I teamed up with ChatGPT — as assistant, not author — and began building a framework.
This post outlines that framework, the metrics we used, the rankings that emerged, and the questions it raises about both past power and present perspectives. It’s  structured curiosity, not a sweeping claim. A little nerdy, hopefully thoughtful, and definitely open to debate.Also , this is no way romanticizes or condones imperialism and the undesirable consequences that imperialism has had on various peoples over the centuries. I do think it belongs to a dark chapter of our collective history.

1. Choosing Empires

I used two simple criteria to shortlist empires . Choose empires post 1 BCE and choose empires which spanned at least two continents.

Here is the list:

EmpireContinental SpreadTime Period

Roman

British

Ottoman

Mongol

Russian

Umayyad

Alexander

Sources: Britannica, World History Atlas. Criterion: Continental spread at peak, post-1 BC

The first question sounded deceptively simplistic :

Who expanded the fastest? However ,  consider two empires A and B . If A conquered a million sq. kms in 10 years but only had horses at their disposal vs B which also conquered a million sq. kms in the same period of time but had railways at their disposal – A ranks higher. I think that is only fair.

Now you could add on further nuance to it. What if both A and B conquer x sq.kms and A does it faster with horses and B does it slower with railways. A needs to get a further bump.

Means of transport also dictate speed of communication. And communication becomes important not only during war and for conquering territories but also to maintain conquered territories. You can come up with several such combinations. And clearly we need a metric that can normalize for this. And the below is what we came up with:

Key MetricRationale

Achievement per Unit of Speed

Sources: Comparative history frameworks, e.g. Ferguson, Diamond, Toynbee.

We also examined each empire’s fastest available transport and communication systems — from Roman roads and Mongol Yam relays to British steamships and then ranked them on a relative , simple 1-10 scale.

Refer to the table below. The Romans had great roads , yet they were amongst the slowest. Alexander and the Russians rank slightly above the Romans. The British had the best speeds.

Empire	Primary Transport	Relative Speed (1–10)
Roman	Road Network	5
British	Steamships/Rail	9
Ottoman	Caravan + Sea	6
Mongol	Horse Relay (Yam)	8
Russian	Sledges, Rail (late)	6
Umayyad	Camel Caravan	5
Alexander	Horse, Foot	6

Sources: UNESCO Silk Road Programme, Transport History encyclopedias

2.. Expansion Efficiency

Now it was important to evaluate expansion efficiencies. Like I said above , one cannot take the area conquered as the sole reliable indicator. One needs to look at this in the context of the both the fastest means of transport available as also the time it took to reach the peak conquered area. We calculated a metric based on land area × continental spread ÷ years to peak. This revealed which empires expanded quickly and widely given their constraints.Refer to the table below to see this in action. Alexander scores the best. At his peak he had conquered almost a similar area like the Ottomans in 10^th the time it took the Ottomans . And in relative speed , the ottomans and Alexander rank the same.

The Mongols are second . The Russians and the Ottomans fare the worst here.

Empire	Area at Peak (sq km)	Years to Peak	Expansion Index
Ottoman	5.2M	130	0.40
British	35.5M	180	0.54
Roman	5.0M	200	0.25
Mongol	24.0M	54	1.00
Umayyad	11.1M	89	0.69
Alexander	5.2M	13	1.15
Russian	22.8M	200	0.34

Sources: Historical atlases, peer-reviewed geography datasets.

3. Geographic & Cultural Adaptability

Then we ranked empires on their ‘adaptability’. How many different climates did they have to operate in? What levels of cultural and religious variability did they to contend with? And how well did they integrate ? Empires were scored on how well they adapted to unfamiliar climates, languages, and cultures. Homogeneity wasn’t rewarded; integration was.

Interestingly , the Russians again score the lowest – they were a vast empire – but perhaps not much diversity. The Ottomans score the best , followed by the Romans.

Empire	Adaptability Score (1–10)
Ottoman	8
British	6
Roman	7
Mongol	5
Umayyad	6
Russian	4
Alexander	6

Sources: Scored based on geographic variance + cultural integration ability.

4. Diversity in Governance

The next natural question was , how did they integrate the new peoples into their governing mechanisms.? How tolerant were they? Did they suppress or did they integrate? Measured. Ottoman millet system scored well — but devshirme was penalized.

The Ottomans and the Romans come out on top here too.

Empire	Diversity Governance Score (1–10)
Ottoman	7
British	6
Roman	7
Mongol	5
Umayyad	6
Russian	4
Alexander	5

Sources: Cambridge History of Islam, Roman legal systems, colonial records.

5. Benevolence Index

Next we looked at ‘Benevolence’. How exploitative were each of the empires? Did they persecute? Did they give local autonomy? Were their taxations systems draconian?

Please note that there is a lot of subjectivity involved in this question. Interestingly , the Ottomans and the Romans come out on top again.

This heavily weighted metric penalized extractive, brutal rule.

Empire	Benevolence Score (1–10)
Ottoman	7
British	5
Roman	6
Mongol	3
Umayyad	5
Russian	3
Alexander	4

Sources: Evaluated based on taxation, persecution, local autonomy.

6. Legacy Index

Then we evaluated ‘legacy’- what lasting achievements did each of these empires leave? In law , education and government? These results should not be very suprising.

Empire	Legacy Score (1–10)
Roman	9
British	8
Ottoman	7
Mongol	4
Umayyad	6
Russian	5
Alexander	5

Sources: Measured legal, educational, and cultural contributions.

7. Decline Dynamics

Then we assessed decline dynamics. If an empire declined over a longer period of time – we considered it to be more robust. We also took into account the preventability of the decline factors. For example , freedom movements are not preventable , where as corruption and lack of reform are.

Interestingly enough , the Ottomans score the highest in preventability score. Both them and the Romans had a very long , very painful decline.

Empire	Years from Peak to Fall	Preventability Score (1–10)
Ottoman	250	7
British	60	6
Roman	300	5
Mongol	100	3
Umayyad	89	3
Russian	75	4
Alexander	10	1

Sources: Reform readiness and internal dynamics literature.

8. Composite Ranking

All individual metrics were normalized (0–1 scale), with higher weights on benevolence and legacy.

Finally each of these metrics were normalized on a 0-1 scale and then weighted with benevolence and legacy together accounting for 60% of the weightages.

Here is the final table. I have always been fascinated with the Ottomans -not really sure why. I was surprised though to see them topping the chart.

Empire	Composite Score	Rank	Confidence Band
Ottoman	0.89	🥇 1st	High
British	0.84	🥈 2nd	Medium-High
Roman	0.80	🥉 3rd	Medium
Umayyad	0.75	4th	Medium
Alexander	0.72	5th	Medium-Low
Mongol	0.68	6th	Medium-Low
Russian	0.65	7th	Low

Confidence band reflects documentation consistency and metric stability.

Final thoughts: This is no way an acedamecially rigorus work . There are error margins and the confidence levels are not uniformly high.

However , I do think it is an interesting exercise – not only in terms of taking a fresh lens to empires , but also how do we use powerful tools to ask the right questions.

Would love your thoughts.