Tuesday, January 27, 2015

The time of Sands

[Is there anything more irritating than blog post titles that try too hard to be clever?  I don't really think so.]

Feeble attempts at cleverness notwithstanding, as Sands gets closer to publication, I have been thinking about possible areas in which the game's theming might need to be polished; the terminology and whatnot that one uses in playtesting are mostly arbitrary, because one just wants to know if the systems work, but now that the design is "done", it's appropriate to take a step back and start considering those aspects more carefully. 

One criticism that pretty much all civ games receive, and almost all deserve, pertains to how they handle time.  On the one hand, these games are supposed to cover time periods spanning centuries or perhaps millenia.  On the other hand, you have individual actions like battles or construction projects or piece movements that seem quite out of proportion with the overall time period covered by the game.

Now I think that Sands is guilty of this criticism in spades.  But, I thought it might be interesting to see whether judicious choices of terminology, and thinking about the game in the right way, might be able to produce at least a hand-waving harmonization of the game's action and the time scale it purports to cover.

With its Roman-era map and Greco-Roman structures and advances, Sands is understood to occur primarily within classical antiquity, which (roughly) covers the period from about 800 B.C. to about 400 A.D or so. (This is entirely contained within the Iron Age). So, that's about 1200 years.   It's not crucial that every single thing that happens in the game occur in exactly the right chronological location or sequence, because we're remaking history here, but just roughly, the game should feel like it covers about 1200 years.  Good.

Now, there are two different "clocks" we therefore need to synchronize with this overall duration:  the population growth clock, and the individual player turn clock.

The basic unit of time in the game is the player turn, each of which consists of two actions.  Turns are grouped into multi-turn "clusters" of 2-4 turns (currently called "generations" in the ruleset, but let's just call them "clusters" for now).  Scoring happens after a certain minimum number of turns have elapsed (6, 7, or 8 turns, for the first, second, and third scoring round, respectively), but only after the current "cluster" ends as determined by a die roll.  So, let's refer to these groups of turns (and, simultaneously, groups of clusters) as "scoring rounds"; a scoring round, it can be seen, typically contains 2-3 (or rarely, 4) "clusters".  And the total game lasts 22 turns or 44 actions.

 If we divide 1200 by 22 or 44, we get 54 or 27 respectively.  So, could we simply view each action as taking ~25 years, and call it good?

With respect to population growth, this seems to work.  Each player starts with 9 "citizens" (6 peasants, 3 warriors), and population growth is abstracted:  rather than occuring every turn or cluster of turns, it occurs automatically in three specific turns of the game (4, 9, and 15, currently), and adds a single peasant to each of the player's owned territories. Making an allowance for the addition of warriors (which happens through an independent process), and sort of roughly compensating for war losses, a typical population growth curve might be:
Turn -- Pop
0 ---- 9
4 ---- 13
9 ---- 18
15 ---- 23
(No mention of how many people each citizen represents -- it's "quite a few" -- because the main thing is to get the rate of growth roughly realistic)

If we make the above assumption that the full game spans about 1200 years, then 15 turns should correspond to about 750 years, and this would give a population growth rate of 0.2% annually. The actual growth rate in the Roman Empire from 1100 BC to 200 AD was about 0.1%, according to an extremely reliable source (wikipedia). So, for a crude system, it seems that population growth gives a pretty decent model of the "correct" behavior under the time scale we're assuming.

The individual player action clock may be a bit more problematic.  Let's 
look at a few actions, and ball-park some time estimates that we might associate with those actions:

  • War -- A decade or less (?)
  • Major building project -- Several years to a few decades (?)
  • Establish a regional trade route -- A decade (?)
  • Implement a tech advance -- Several decades to a century (?)
  • Manage unrest through reforms, etc. -- Several years to a decade
This suggests that a turn (two actions) might be about 10-20 years or so.  (Call it 15, split the difference).  That would make a cluster of turns, which, again, can be 2-4 turns, 30-60 years.  I've called that a "generation", i.e., the reign of an individual ruler; and maybe that's about correct.

As for a "scoring round", which typically contains 2-3 "clusters" or "generations", the game currently refers to that as an "epoch" which is clearly wrong.  At 90-120 years, a scoring round seems to coincide not-too-unfavorably with a "dynasty", i.e., the length of rule of a typical ruling house before it falls or is overthrown or dies out or whatever.  So maybe "epoch" should become "dynasty".

There's a problem, of course:  at 15 years a turn, the 22 turns of the game only correspond to 330 years rather than 1200 years; and with only three "scoring rounds", only three dynasties rule over a 1200 year period, which also seems wrong.

I can see two possible reconciliations, both of which require a compromise to accuracy.  The first is to view the turns as lasting about 50 years each, and then the clusters are "dynasties" and the scoring rounds are "eras" or "periods".  The compromise this requires is that it over-estimates the time scale of the individual actions (more egregiously for some than others).

The other option is to view the turns as 10-15 years, the clusters of turns as "generations" or "reigns", and the scoring rounds as "dynasties", BUT, the idea is that we're not tracking EACH AND EVERY ruler or dynasty, but only the MOST CONSEQUENTIAL ones.  There is, to me, great thematic appeal to this reconciliation -- it acknowledges that some rulers didn't do that much or didn't establish a reputation that was preserved for posterity, and this is in keeping with the game's theme.  What I also like is that it keeps the theming of the player-as-individual-ruler intact.  Yes, with each generation, the player "magically" becomes the next ruler in a succession, but at any individual instant, the player's representation is concrete:  he is the current ruler of his civilization, and his current horizon is simply the length of the current "generation" -- his own life span, which is uncertain.

So I tentatively prefer the second, but ultimately, it's mostly a cosmetic consideration, and decisions of this sort ultimately fall to Spielworxx to implement.  Happy to hear any other comments, suggestions, or considerations that others have on this subject.

Sunday, January 4, 2015

Solo play

Back into the weeds for a bit.

A while back, I wrote about how, originally the game had an idea of "trade routes" as adjacencies between players' cities, but my good friend John Velonis (designer of, among other things, a really fun beer and pretzels game, Venus Needs Men) observed that these were a pain to visualize, and additionally weren't entirely historical -- trade routes were about access to resources that were far flung over a vast geography.  This prompted an addition to the board of "trade good" spaces (the star spaces that show up in the image in this post) and of "caravans" (think Settlers road pieces) that players use to connect their cities to those trade goods.  In the present ruleset, connections to trade goods are a scoring category, but connecting caravans to other players' cities or to trade goods let you increase your heritage and other players' caravans that touch your cities authorize you to hit them up for tribute.

Now the nice thing about this is that, somewhat like Knizia's Lord of the Rings, different player counts give a different feel.  The more players in the game, the more potential cities you can connect to (more access to heritage) but the more exposed you are to tribute demands from other players (and of course the more players you can potentially collect from).  With fewer players, you can still get heritage from your caravans, but you'll have to build longer routes to reach all the stars (at the same time incurring less tribute risk), or you can rely more on the other means of acquiring heritage.

But what took me a while to realize was that, for the first time, none of the core scoring mechanisms depend on the presence of other players, and in fact, it's possible to play the game solo.  I've tested a solo variant a number of times now, and it seems to play pretty well.  I'm hoping it will end up included in the ruleset inside the game's box.  It's a variant because it does include two ghost civs, which exist merely to give you someone to attack (so military development is still a plausible path) and to give you some risk that you have to factor into your plans.  The AI for the ghost civs is trivially simple -- when population grows, they add a warrior in each of their territories, and each time you roll a 6 when testing for the end of a generation, one of them attacks one of your territories. (Since the game lasts 21-24 turns, statistically you'll face about 3-4 attacks per game, and this feels just about right).  They don't build or advance or anything like that, so you have to do the legwork of caravaning mentioned above by yourself, and there's no one to discover new techs on the tech tree for you, so scores are probably a bit lower on average, but it's still a fun way to get to play the game and a good way either to learn the game or to play around in the sandbox and try some new strategies.

I don't know if players who only plan to play solo would seek to acquire the game just for solo play, but I do think it provides a pretty similar experience with only a very few simple additional rules, so it has some value.  And it would be neat to be able to say that the game seats 1-6 players; I don't think too many civ games can say that.  Because the lion's share of the game time happens in the simultaneous action planning, the game length shouldn't really change much with player count in principle, but of course in practice 5 or 6 players will generally take longer to play than 1-2 players.

Thursday, January 1, 2015

Sands in a slightly different nutshell

The news is public, so I guess I can officially announce that Sands will be published by Speilworxx, as announced here.  This is great news for the game; Spielworxx has set a high standard with a string of excellent gamer's games, and it will be great to be a part of an excellent brand.

It's been almost two years since the last "nutshell" post, and while it's likely spielworxx might change some aspects of the game before release, I can at least update the previous "nutshell" post with a few of the changes that have taken place since that post.

The overall scope of the game is the same; it's a civ building game set loosely in classical antiquity, with the action taking place on a Roman era map, and the goal (which I believe to be unique to Sands, and its most distinctive feature) is to record "Chronicle cards" boasting about your exploits, so that your posterity will remember your greatness long after your civ has succumbed to the ravages of time.

A few things have been streamlined.  First, the game now DOES have tech trees; the game has three civilization categories (cultural, political, and civil), and there are two tech trees in each, but it's really more of a " tech line" than a tech tree, as they don't "branch".  The trees are card- based, and the first player to reach a given tech level on a given tree gets to choose what card at the next level on that tree will be.  So, there's some variability from game to game, but everyone has the same trees available and it's easy to see what everyone's capabilities are.

The "achievement tokens" have been replaced with "heritage tracks", which reflect your reputation in the three civ categories.  Your heritage in a category increases when you score in a category (so, heritage, literally), when you build trade routes that contact other players' cities (reflecting the idea that your reputation is spreading), or when you complete an Emphasis card.

Emphasis cards are one of my favorite new additions; on each turn you select two Action cards, but once per "generation" you can instead play an Emphasis card, and this does two things; as mentioned, if you take at least three actions during that generation that match the civ category of the emphasis card, you get a boost to your heritage -- you've established a reputation in that area has increased by your having specialized in that action.  The second is that it unlocks some capability on other action cards.  So emphasis in building lets you build a city or demand tribute, emphasis in culture lets you put a marker on an additional tech tree or reduce your unrest for free, and so on.  

I'll post updates as additional changes emerge during development, but I hope this post, and the previous "nutshell" post give a good sense for what the game is mostly about and maybe what might be interesting about it.