In Civilization V, you have a choice of two social policy trees early in the game, Tradition and Liberty. Tradition is supposed to be the "tall" choice to better enhance a smaller number of cities, while Liberty is supposed to be the "wide" choice to rapidly expand to many cities. However, with the Gods & Kings expansion, I keep feeling across many games that Tradition is always better even when building wide. This article is a detailed discussion.

Just to recap the mechanics, as of the Gods & Kings hotfix patch in March 2013:

• Tradition opener: +3 culture in the capital city. Faster border expansion in all cities.
• Aristocracy: +15% production for wonders. +1 happy per 10 citizens in a city.
• Legalism: Provides a free culture building in your first 4 cities.
• Landed Elite: (Requires Legalism) +2 food and +10% food surplus in the capital.
• Monarchy: (Requires Legalism) -1 unhappy and +1 gold per two citizens in the capital.
• Oligarchy: Garrisoned units cost no maintenance. Cities with a garrison gain +50% ranged combat strength.
• Tradition finisher: +15% food surplus in all cities. Free aqueduct in your first 4 cities (even without Engineering tech.)

• Liberty opener: +1 culture in each city.
• Republic: +1 production in each city. +5% production when constructing buildings.
• Collective Rule: (Requires Republic) Free settler. +50% production on settlers in the capital.
• Citizenship: Free worker. +25% tile improvement rate by workers.
• Republic: (Requires Citizenship) Starts a Golden Age. -33% to additional social policy cost per city.
• Meritocracy: (Requires Citizenship) +1 happiness per city connected by trade route. -5% anger in all cities.
• Liberty finisher: Receive a Great Person of any type.

The changes from vanilla pre-expansion Civ 5 are the addition of aqueducts to the Tradition finisher, and reversing the order of Republic and Collective Rule. Also, a patch shortly before Gods & Kings changed the Liberty finisher to increase the Great Person cost counter as if it were a naturally spawned GP.

These points are all in Tradition's favor and against Liberty. The Liberty finisher change is a serious nerf, since the GP isn't really free if it increases the cost of all later GP. Really the Liberty finisher functionally awards just 100 GPP. By the industrial era, the costs will catch up and you will not actually be ahead by any great people, so Liberty really just means you get the same Great People a bit sooner.

Reordering Collective Rule to after Republic was also a deceptively serious nerf. Now the free settler requires either skipping the worker or waiting until the fourth policy. This wait is killer. Liberty was a lot better when it could go opener - worker - settler. Three social policies come quickly, but the fourth takes as much culture as the first three combined (25/30/60/115). This delays the settler and settler bonus from roughly turn 40 to turn 60, which is too late. Turns 40-60 are exactly when you become able and willing to produce more settlers, either by building or buying. By the time Collective Rule arrives, I've already put out three settlers (escalating policy costs and delaying CR further), can't support happy for another, and thus also miss out on the production bonus.

So Tradition openings to me just feel infinitely smoother. Liberty feels like fighting against the game, dangling that settler just beyond when I need it, getting taunted by Meritocracy promising happy then oh crap I need forty turns of worker labor for the roads, and finally the "free" great person that is anything but. Tradition feels like a warm practiced cooperative teammate, handing over exactly what I need at each step like a surgeon calling for the scalpel. Culture first, then food, Monarchy just as the happy starts crunching, Aristocracy just as you start building wonders, Aqueducts just as Civil Service arrives and food blasts off.

Legalism in particular is highly underrated early. I love starting with two scouts, but then fitting in the monument is troublesome alongside demands for the shrine and granary. Legalism allows happily skipping the monument. And those monuments greatly help new cities. There is a significant difference between a border pop in 8 turns via monument (add a tile just as you grow to size 2) and 15 with only the Liberty opener (tap tap tap will you acquire that stupid deer already so I can stop working the bare forest!)

And Tradition accelerates into itself much better than Liberty, thanks to the stronger opener. With Liberty, I always feel that the worker policy is lagging if I don't get a second ruins culture pop. With Tradition, one pop is sufficient because the opener accelerates into the second policy, and even no culture pops can be playable since I don't really care that much about the timing of Legalism since it's not on the critical pathway to selling my luxuries.

Finally, Tradition has lasting power for the entire game that Liberty doesn't. The opener doubles culture expansions for the entire game, Aristocracy always helps, Monarchy is steady, and the +15% food finisher matters too. While nothing in Liberty really has much impact beyond the first hundred turns. So Tradition wins.

Of course, my readers don't come here just for opinionating, you can find that in any thread on Realms Beyond or CFC. I'm here to give this a full-blown mathematical treatment. I conjecture that Tradition beats Liberty even with a wide city build where Liberty is supposed to excel.

Let's measure the competitors over the first 100 turns. For the playing field, we will consider a simplified but reasonable scenario that our civilization builds cities at ten turn intervals starting at turn 30, and that each city grows at ten turn intervals. So by turn 100, we have reached nine cities, at sizes 11, 8, 7, 6, 5, 4, 3, 2, and 1. This is reasonably achievable for a typical wide build with sharp play, a few ruins pops, ample luxuries, and some religious happy.

The opening policies come out about equal in productivity. Assume they come on turn 15, splitting the difference about getting a culture ruin or not. Tradition produces 255 culture over the next 85 turns, Liberty a total of 365. But Tradition acquires just as many tiles with its expansion discount.

Conclusion: Tradition and Liberty are extremely close overall through those first 100 turns, including nearly identical gold and happy outputs. Liberty produces more hammers, Tradition slightly less food plus hammers. Liberty gives a bit more overall, but only by fully counting the finisher as a Great Engineer wonder rush and neglecting the fact that it burned the 100-GPP great person slot.

But watch what happens next.

Consider the next 50 turns. I am going to simplify the civilization model. Assume a capital that averages size 20 running +30 food surplus, and nine other cities averaging size 10 and +10. I will neglect the cultural effects of both the Tradition opener and Representation, acquiring tiles and policies faster respectively. I think they're broadly similar in strength.

Tradition now outshines Liberty. Most of Liberty is over, while Tradition's wonder and food and capital bonuses keep trucking on. Liberty's hammers are ahead on paper thanks to Republic (actually I didn't realize until writing all this that Republic is actually pretty strong.) But by this stage of a maturing civilization, food can generally be traded for hammers 1:1 by working lumbermills and mines instead of farms, so Tradition can just use its food to outhammer Liberty.

And Tradition actually produces a lot of gold, from Monarchy and the various cost savings. I definitely didn't realize that. But there's my answer on wondering when my games sometimes have a lot of money or not. Sure enough, my rich games (Dutch culture, Rome science) went Tradition and my poor games (France culture, Korea science) went Liberty.

Finally, it's notable how Tradition gets ahead for happy. Meritocracy is supposed to be the happiness solution to a wide civ, but Tradition beats it. Once each city hits size 10, Aristocracy is as good as the trade route portion of Meritocracy. And the 50% capital anger reduction of Monarchy always beats the 5% civwide of Meritocracy, as long as your capital is at least 10% of your empire's population, which is true in all but the ICSiest games. Meritocracy only wins for happy if your cities never reach size 10 but you build so many that the cumulative population comes to 10x the capital. In Civ 5's early days, that actually was the best way to build, but not so much now that religion can lift so much extra happy.

Conclusion: Tradition wins. Only if you go pure max ICS might Liberty come out ahead, but for any reasonable wide build, it's Tradition.