Monday, June 17, 2013

Goals of balancing and tricks behind the scene

I like to think that each game has 2 sides - 1 side being the theme, story, flavour and the other being it's rules, mechanics, the mathematics of the game, it's like the brain having 2 sides - right and left. Now when talking about balancing we usually refer to the mathematics, the rules, and usually we are talking about offering all players equal starting position and having lots of viable strategies. Now I believe there are more things to consider when tweaking numbers and rules for balance. And moreover I believe balance is, like all rules connected to the story as well. In the end each game tells a story, a metaphorical, literal or abstract, story is told.

So what are those other aspects to consider? Making sure to have equal starting positions is important for fairness. Having multiple of viable strategies makes the game more interesting and replayable. But what about intenseness, immersion?

Consider a traveling merchant game situation where a player has 2 road choices to a single destination, but either way player would have to role a dice and in the case of 1 or 2 bandits would attack the player and kill him. Suppose the game would be otherwise equal for all players, now what's wrong here (we assume that there were no mechanics to reduce the risk in any way)?

Well, firstly the inevitable death of the player, unless he simply stays in some safe city for the whole game. That however is quite a boring option.

Secondly if you transfer this situation into a story it would sound something like: "Merchant Toby set out with his caravan on a journey to the eastern cities were riches awaited for him. During the journey he came across a crossing and could choose if he'd take the hilly road or the road by the river. They both lead to his destination so he chose to take the road by the river. But oh god he got attacked by bandits and they killed him. In his last thoughts he figured that the same would have happened if he had taken the hilly road. The end."

I think we can all agree it is a pretty horrible story. Throughout play-testing my last game Pactio I often went the path of seeing how some game interaction would be transferred to a story, helping me decide if the game was immersive enough, and that it didn't contain some awkward interactions.

Now suppose we improve upon that. That instead of always having the same 1/3 chance of dying you would instead draw and reveal a bandit card, where it shows the odds for each roads. Moreover you could pay scouts some money to see the next card. Then you could also invest a bigger amount of money to hire caravan guards, so that when you do get attacked by bandits you have 4/6 chance to drive them off for example.

This leads to a much interesting story, where a merchant would die because he was too greedy to pay the scouts some coin, or live because he chose to spend some money to protect himself or earn riches cos he took a great risk and he got lucky.

So we fixed immersion quite a bit, but a game where you have a chance to die in the first turns is not really enjoyable. A game where you struggle and make all the tactics in the world to still have a relatively high chance of simply losing and going out of the game might have some sort experimental value, but won't be something people would want to play, at least not for a second time.

Now the easiest solution would of course be that you take out dying and replace it with losing some gold. If you lost all the gold then for the game's purposes it's same as dying. However I think something more resolute would be more interesting from a game design perspective.

However it is important to give players a solid chance of survival. Suppose the game lasted for 10 turns, each turn involving going from 1 city to another meaning risking your life 10 times. If the survival for each turn  was 2/3, then the chance to survive the game is 1.7 %. In the case of 5/6 (only rolling 1 is deadly) the survival chance is 16 %. If you hired guards starting from the third turn and bandits attacked you with a 1/6 chance your chance of survival would be 27%, still way too low.

So when designing that aspect of the game we should start by setting the goal of what the odds of total survival should be. I'd suggest 97-99 % for a solid safe strategy, 75-95% for a occasional risk taker, and only most reckless daredevils chance of survival would drop below that to below 50% and so on to a almost certain death.

First of all we would have to lose the 6 sided dice and replace it with a 20 sided dice, as that would give us more accuracy when giving odds. Secondly, suppose the story went that players are the pioneers of making those trade routes, and bandits don't start to show up before players have totally earned a certain amount of gold. Then you could have multiple of bandit decks representing different densities of bandits, each one used once the players' total power level has achieved a certain amount. Then we'd like to set the amount so that in an average game the first deck would come into play on third turn, second on sixth turn and third one on ninth turn.

Secondly we could give certain characteristics to each of those decks. For example the easy deck would always have 2-3 roads with 0 risk, and max risk would be no bigger than 3/20. The medium deck would always have a single safe road and max risk would be 5/20. The high risk deck would have an average risk of 5/20, so you would have to hire guards to decrease your risk. So that if you wanted to play safer in the late game you could invest into guards. Also the more you invest into them the higher your chances for survival are.

This leads to safer merchants being able to go through the first 5 turns with 0 risk, and then with the implementation of guards and scouts could play through with risking as little as possible.

So now the rules have transformed the game into something which tells a meaningful story. Some merchants take knowingly risks for a chance to win the game, while others prefer to play a steady but safe game.

So to sum things up, when balancing a game, one needs to keep their eye not only player vs. player fairness, but also player vs. game and some parts of game vs. other parts of game balance to keep the game interesting and intense. Also the story balance is important, as this is the immersion part of the game. When tweaking numbers try to understand the meaning behind the numbers and feel the story it creates at the end. Good understanding of combinatorics is really important for that.

Saturday, April 13, 2013

Balancing - let players do it?

The purpose of balancing is giving a fair chance for all players no matter the side they pick, and no matter what kind of starting customization they choose or what kind of starting random setting they get.

One of the simplest balancing forms is giving all sides identical starting resources, and abilities. For example chess. Even then however, it matters weather you start or go second. However for non professional chess players like me the starting advantage is quite irrelevant.

Once you start adding different abilities, or different starting resources, or randomness, then the balance gets disrupted. If you take civilization series, then the starting area is random, it really plays a huge role, how you get your game going. Of course when I start to play the game, I know what I'm getting into, and the randomness is also good as it creates for different situations, making me seek out different strategies from game to game. In games like these you only care about having the fighting chance. So how is this achieved?

Suppose we were playing three player chess and one player starts with 1 rook instead of 2. Does he have a disadvantage over the other 2 players? What if he had no rooks? As the goal of the game is to defeat all other players, meaning for the both other players point of view will be that there is a weakened opponent, and a normal opponent. So when they start to build up their strategy, they will create a certain perception of both opponents, and if they perceive the situation right at all times then eventually the 2 rook advantage that 2 players have will be taken away from them. In a 3 player game, where all players depend on each other, the perception and keeping the game in balance is one of the things players have to keep their eye out if they want to succeed.

Let's take a more trastic example. Suppose that in 3 player chess 1 player starts with 1 king and 7 queens in the back lane. Now he has a stronger army compared to other players, but since all 3 players will want to win, then the other 2 players will most likely start cooperating from the very beginning until they perhaps come to a situation where the armies of all 3 are somewhat more equal. However, what if the overpowered player picked 1 of the 2 other players and started focusing him, saying that if he doesn't betray his "temporary ally", he will lose anyways. This is a situation, where picked player has no chance to win, but  he gets to decide who of the other players wins. This happens because the alliance is bound to be temporary anyways, and the breaking is inevitable, since all players want to win. Propably there are some ways to play around it, but a situation like that, where you can't win but can decide who else wins is something that can happen in a 3 player game, where everyone influence other players very strongly. Like in chess, through taking pieces.

Similar is the board game "diplomacy", where you can make deals but don't have to keep them, leading to inevitable betrayal, and leading to similar situations as prisoner's dilemma. The emotions play a very strong part of the game, and by creating negative kind of emotions, many players don't like it, even though it has "balance" in it. My personal idea on the matter is that a 3 player war game - it represents 3 tyrants fighting for the overrule of the land, no honor, only deception and millitary is used... and to look it like that you can see the non-appeal of such games. Especially for the people in the democratic states, who propably don't relate to the idea of going to an all out brute force war just to be the overruling nation in the end.

So, as you can see having more than 2 players can be a balancing experience, but also a ridiculous experience, as there are lot's of ways to lose just because other players choose less rational and more emotional decisions.

In my board game Pactio, players get to pick roles, and some of them are with really unique play-styles, strengths and weaknesses. On 1 vs 1 level many of the roles are far from balanced, the strengths of some roles actually assume having multiple of players. However since the game is played with at least 3 players, the players can start balancing the game them-selves. However compared to 3 player chess, Pactio is an economic game. So the whole game revolves around growth. There are a few destructive aspects, but mostly it is about growth. Also players can do sealed deals between each other, meaning all their interaction have much more strategical meaning, since you can make deals, which you think will be fair and balanced and having a guarantee that they will be kept, therefore having the calculated effect on the game. So what happens is that once players learn to play the game, they understand that it is all about perception, if you manage to perceive players right you are able to understand what kind of deals are actually profitable, fair or bad for you. Of course the perception part can still involve emotional play, but taking into account all the play-testing I believe it has done way more good than bad to the game.

Also since there is little to no destructive element there is much more solid ground to base your actions, and players, when actually having a situation where they can decide the winner other than themselves tend to simply choose the path that gives them the higher score, creating for a better emotional feeling after the game. As this game is about becoming economically best. You might argue that from a rational perspective it doesn't matter, but from an emotional perspective you simply take it as one player traded the best and others came close being second, third etc. while in a war game 1 player crushes all others and the others lose - there is no second or third place.

To sum things up, one way of balancing a game, is to make it a over 2 player game, and give players tools to effect or interact with other players. That leads to a much more robust system, where players can balance the game by keeping stronger players in check. However note, that it can also lead to wierd situations, and my personal opinion is that, games like 3 player chess, or diplomacy can have rational situations, but the goal itself leads to lot's of emotional dilemmas rather than rational game play. Furthermore 3+ player games may be as balanced that a perfect game-play would lead to an inevitable tie for all parties, ruining the point for a war game, but being a completely acceptable outcome for an economy game. Never the less, playing such games where, players themselves balance the game teaches a lot about balance in real world, and I strongly advise it. At least for a few times.

And again, don't hesitate to comment and share your ideas on the matter.

Wednesday, April 10, 2013

Strategy RPG - a tragic lovestory?

I've always been fascinated by the ideas how different game mechanics can create the feel of things you could also express in poetry or literature. Sadly however I feel like the math behind the games is way too often not related to the story or the theme. Is it because attempting to do that would make the games way more complex? Or is there something we could do to change that, without sacrificing the game experience, and instead make it ways more enjoyable and immersive. Or is there a simple explanation to why it is not being chased? Let's explore these questions by delving into the deep mysteries of combining story telling and strategy, poetry with math, analytical mind with emotions. Or to put simply modern strategy games with RPGs. You think the last one was inappropriate, don't worry, I'll prove you wrong.

Strong math underlays strategy games. RPGs on the other hand are basically interactive story telling. When strategy games origin from chess, go and all the other mind games symbolising the strategies of war, then RPGs origin from stories, told by artisans, travelers, parents to child, the guy in the town market, preachers and so on.

One of them is a game where the rules are most important, and the outcome depends on how well you understand those rules, how well you play, and the outcome is a ruthless defeat if your not careful, or if your simply not good enough.
Stories however, first have an ending, or a moral, and path to the ending is merely for explanation purposes, to make it plausible. That means victory or defeat is decided already before you start experiencing the story. Usually the path is unlikely if you were the one who would take it. Or the character who is taking it simply unlikely powerful to exist in the first place, making it plausible to ignore the rules set otherwise for us mere mortals. Of course there is always some truth in stories, but let's face it, a good story is filled with dreams, unreality and imagination.

Nowadays, strategy games and rpgs, the children of two so contradicting schools of art are seeking to take pieces from one another, seeking to combine and form something new. And I too have caught in the fascination of this ideal, what seems to be an awkward beginning of Romeo and Juliet, and we all know what happened there. But luckily for us reality is not storytelling, for us the ending is open, and until proven impossible everything is possible (then again there I was back in high school wondering if I could come up with a positive solution to NP=P problem, obviously without any success).

So what is happening? If you look at civilization series or Europa Universalis games, they are no doubt very good strategy games, but they include quite a lot of story telling elements. They give players the opportunity to personalize, to define what kind of ruler they are, or what kind of kingdom or civilization they are leading, they still leave the outcome open, will you win or lose, stand the test of time in civilization series, or admit defeat during the path. But you get the option to choose really neat paths, and even if the game is balanced in it's core, then by finding the right niche, you simply steamroll over all your opponents, something a pure strategy game wouldn't allow (I believe if you take away the campaigns from Starcraft, it is an excellent example of a modern pure strategy game, so was age of empires 2, and of course the many board games out there).

So in the other side, there are role playing games which instead of simply telling a story, also give you an illusion that you might actually lose, or your choices change the world. Why am I saying illusion? Well, you can die, sure, but you always have a save game/autosave so on. You may do bad choices, but they effect the story either not at all or in a very controlled manner. Only way to lose, is to quit the game, and as long as you don't do that, the choices you make and how you play the game effect only how long it takes to finish the game. But they are using more and more strategic elements to make the illusion better, buying items for money (resource management), advanced combat (tactics) etc. They can even have multiple outcomes, involving the exploration of morality, cause and reaction, but like said before, in a controlled environment, meaning, if you do this, then that happens and it's not going too much more complex than that.

So what should a true strategy RPG game be like? What elements of each genre would it take into itself? And how to solve the conflicts of them both?
First the world definitely has to have rules, by which it works, your actions, and possibly actions of other characters will definitely effect the world in a fair, rule based manner, otherwise it is by all means not a strategy game. What do I mean by that?
Well, suppose you were a hero defending a fort from orcs. Suppose you can position the fort soldiers however you like and give some orders. So, in a fair ruled RPG strategy game, the number of orcs and number of men protecting the fort is finite, meaning that suppose there were 100 orcs and 50 of them died without making progress then the orcs, being rational entities would retreat, and not spawn additional ones and wait until you kill their leader or complete any other objective. Also, suppose you succeed defending the fort but only 4 men remain on your side, out of initial 100 for example, it was a bloody battle. So what would a village with 4 men, and a few hundred of elders, women and children do? Well, they could migrate to another village, increasing their population, but abandoning this village, making it a ghost village. Or maybe the land is in turmoil and they cannot migrate without taking a great risk. Maybe women take on hunting and become soldiers and it becomes a matriarch governed village, whatever happens, it has to be deduced from the rules of the world, they would need to guard themselves, or they risk being overrun by raiders, they need to hunt, or they starve. And village as an entity should be able to deduce the course of action from those rules, and that would become a story.
So here you can see the first obstacle of making a strategy RPG, if you want to have story, you need to have entities filling all the roles in the story, and all the entities would need be able to deduce their course of action from the rules and current situation, without conflicting their personality. And that's a hell of a AI challenge.
If I would wild guess an algorithm for the AI, I'd go for Monte Carlo modification, where either instead of having random set of options entities only attempt from a set of options plausible for their character, or instead their choice evaluation function is modified by their characteristics. In that way you can still have them look ahead for a couple of moves and make strategically sound choices, while still making for a plausible story. Of course using those 2 strategies makes for a hell of a difference. For example in the first choice a kind village elder will never consider a ruthless decision, if it might be strategically very strong move, saving maybe many in his village, while in the second scenario he considers all possibilities and chooses the ruthless one when it significantly out weights all other options. Hell, if I'd do the game I'd give some characters one algorithm and others the other one based on weather their personality is idealistic or pragmatic, and a mix for those in between. Why, cos the first algorithm is basically saying that I will only consider choices morally acceptable for me, while the other one will say that I will take morality into account while deciding what option I choose. Here's some combining math with story.

Second main aspect would be the story, the immersion it creates. The feeling that the story is more important than the gold you possess or the size of the army at your command. That everything else is a mere tool to craft your story. So when those two genres are to be combined, I would avoid all sorts of winning by score, or winning at all. There can be an end, end date, end condition, like solving a conflict, or defeating your foe, or losing yourself, or a end goal, defined by yourself (for example at the beginning of your game you say that I want to play until my capital becomes the biggest center of trade in the game). 
So there has to be story, and there has to be rules for the world, meaning that rules and story must intervene, unless of course you want to occasionally ignore either of them, but that's not the grail we seek. No we'll only be considering options consistent with the dream. However since rules are mathematical definitions of legal moves, which when combining with story, rules become the mathematical definitions of plausibility. To experience deep stories in this game we would need to go as deep as starting to mathematically define emotions, feelings, theme, like the mad going J. Nash did in Beautiful Mind.
Of course it is already being done all the time: Vassal levies and taxation system in Crusader kings to explain the partial independence of the vassals from their superiors, Casus Belli used again in crusader king and also Europa Universalis to define the reason for war, which by itself has the potential to create a lots of story, governments and civics in civilization series, that give character to your civilization. If you've played some of the grand civ IV modpacks, then you have most likely witnessed the efforts people have done to try make it more realistic, to explain reality through mathematical rules.

Finally there is the question of control? While story maybe doesn't mind rules, and strategy games might not mind some story or theme, then control is something that really rips those two genres apart. In strategy, you have no control, until you are winning which is when you relax and just finish the game. In story telling or RPGs, the whole thing is controlled. That is one of the most conflicting things there is in Strategy RPGs. For example in King Arthur II, which was a solid take on Strategy RPG genre, propably the best so far, I got conflicted, cos I wanted to finish a quest while at the same time my land was being raided from the north. I had the expectation of letting me finish my story while having to make strategical choices at the same time. And because killing that stupid evil general (I think he was Roman) in the north didn't seem like much of a story, hence, I felt conflicted.
The question goes even deeper. Would we, the gamers be able to accept, that instead of our ideal of the story, an AI story happens instead, I mean not always, but since it is a strategy game, we shouldn't have all the control, we should have a chance to lose, when being careless, or by doing mistakes. How would we feel like, when instead of us freeing the land from slavery we fail and a tyrant takes complete control over the land, our allies and friends would be butchered, the odds of ever succeeding our goal dropping to near 0, would we be able to accept that story? I mean imagine, if that story would be told to you similarly to Dragon Age series end game, maybe even from that AI perspective, how would you feel?

So to sum things up, as you can see the reason for why so many people are fascinated and driven by the idea of combining RPG and strategy has a deeper goal, a hidden agenda that has always existed. Cos by wanting to do that, you want to create rules of plausible story telling, you want to be able to define both real and imaginary worlds. If one would ever want to make a strategy RPG as deep as the best hand-written stories, you would need to understand both emotions and mathematics in the same time at a very deep level, and not go insane in the meanwhile.
And what if we succeed? What if the game feels like life, being ruthless, bias and conflicted? I suppose it depends on what kind of world we define. Maybe the first breakthrough in RPG strategy world will be of a world nothing like ours.

Whatever happens, if you have the fire inside you that urges you to attempt it, it propably won't go away, I know mine won't. All we can do is open our eyes look at the dragon we want to slay.

Thanks for reading, I hope you enjoyed a slight less rational sidekick. Don't hesitate to let me know your thoughts on this.

Monday, April 8, 2013

Simple card game using draft mechanic - Draft

Hi all,
Today I thought some more about pre-game customization, and somehow my mind wandered to an old card game named Bartok, which I learnt from a punch of Australians during one of the IOI-s I participated during high school and came up with a new simple card game based on drafting mechanics. Bartok was based on a simpler game, which English name I sadly do not know.

Basically the original game is that each player starts with 5 cards in hand, and there is one card in the middle. Then players take turns and put the same suit or same number/picture card on the top as the topmost card. If they cannot do that they take a card from the pile and skip their turn. Goal of the game is to get rid of your cards. In Bartok you have the additional 3 rules:
1) If you make a mistake, you are punished with 1 extra card.
2) Asking questions is a mistake
3) At the end of each game, winner makes 1 new rule.

Back in the day I enjoyed the game mostly because of the rule making, and the punishing aspect, it lead to ton of fun, especially if good rules were suggested, and people got more and more tired.

So today I had this idea for a similar card game, but using drafting mechanics, for now lets call it simply Draft, haven't thought of a better name for it so far. You only need a normal deck of 52 cards. Take jokers out. I play-tested it with a friend of mine and here's the final setting we ended up playing.

Rules for Draft: 

Right now it's a 2 player game, it uses a standard 52 card deck. Game starts with deciding the beginner.
Both players are given 7 cards and 1 card is put in the center of the table, visible. The rest of the cards are not used during current game. In this game one of the main differences is that you don't get new cards, if you can't make a move, you lose.
From your 7 cards both players pick one card and give the rest to your opponent. They repeat the process until both players have picked 5 cards, the remaining 2 for each player are put away.
That ends the drafting part. Then you take turns and you have to put the same suit or the same card as the topmost card to the top. If you cannot play, you lose. If you play and that was your final card, you win the game. Of course if your opponent loses (cannot play) then you win the game.
At the end of each game the loser chooses who starts the next game.

Basically in this game, most of the game takes place in drafting, I'd actually say that drafting part is the game itself, and then there is an after-game, deciding the result. Of course the after-game is also important as there were lots of cases where there were paths of play for either player's victory, taking into account that there was a 1 card both players were unaware of. 
For us drafting lasted at least 4-5 times longer than playing out the hand itself. Playing out sometimes lasted only 1 or 2 moves, but as we got more experienced, we managed to get usually to 4-5 or more moves.

At first we tried with drafting from 5 cards, so all cards were used up, it lead to too many games which ended first turn dead-end. Having the 2 extra cards really helped to make plans and gave each player more opportunities to outsmart your opponent.

So, what happens in the game, from a strategical point of view. Well, basically, while drafting you will want to be able to achieve a situation, where your opponent cannot play anything anymore, and you want to make sure that you are able to get to that situation. Notice that for the entire game, both players are unaware of only 1 card, guessing the card and noticing avoiding possible bad surprises is one key part of the play, the other one is, how do you make your hidden card count the most. 

The game got even more interesting when we added the 3-rd clause of Bartok to the game. We configured it that if a player wins 2 consequent games, then he may do a new rule. Reason is, unlike in Bartok, in this game each new rule can make the game much more complex and deeper. 
First rule we tried was that on top of diamonds you may also play higher clubs. Simply this rule made the higher clubs much more valuable, and lead to lots of new interesting situations. 
Second rule was, that all 2-s could be put on top of anything, unlike in other games, this leads to a really overpowered 4 cards :D . They alone can be game-winners, but we didn't know that back then. 
Finally we introduced a rule that you could play multiple of same cards (2 sixes or 3 kings for example) if one of them could be played, the legal one would have to be put on bottom and the rest could be put on top in any order. That was a game-changer as it made it much more possible to achieve the "I have no more cards" victory, even when you were not starting the game. 
We didn't make more rules cos unlike Bartok, each rule made the game more complex from a strategic point of view (not just remembering the rules).
Also during first play-test, it seams that starter has an advantage, but not always, it is very situational, but weather you are the starter or you play second, you have to take that into account while drafting.

All in all, I think the game is a very good entry game for memory training and drafting games in general. Also it doesn't take long, so I think I will be using it as a filler for quite some time now. So far I can only advise to play it with 2 players, but I'll try to play it through at some point with 3 or 4 players as well and see what can be done there. Though I would start the play-testing with the rule that drafting goes counter-clockwise while playing starts clockwise. Also the player from the winners right will choose the starter for the next game.

Hope you enjoy it, and don't hesitate to comment or share your play-experience. Oh, and if you know the English name of the simple version, please let me know :)

Also, if you know of any card games that already uses this idea, I'd also be interested :)

Sunday, April 7, 2013

Pre-game customization

 Hello, For those of you who do not know me, my name is Indrek Klanberg, and I'm a young game-designer.  Currently I am finishing my first economy board game - Pactio. 

 Recently I had a discussion in Google+ Board game design community with another member about pre-game set-up. Both of us had different ideas in mind, but pre-game customization would be the thing to generalize both of our ideas. 

 My idea basically was that, at the start of game, players draft the political and power situation of a city. Basically the idea was for a sequence game for Pactio (www.pactioboardgame.com). The new game would not involve city building, but instead all players draft their initial buildings (for income and passive abilities), active abilities (you could use 1 ability each turn so they are as useful as much time you spend on them) and plot events (those are basically events that will take place later in the game, effecting all players, so you will want to draft events that synergise with your strategy). Basically all the 3 types of cards would be in the same deck, and you would have to decide, how much you need to focus on property, abilities and plotting. Then the scenery would be set and the struggle for power will start. 

 I started thinking a bit more about it and at first glance, it seemed to me that it might actually be something unique and I couldn't remember where it had been used before. But then I thought further.

 I don't think anything quite like that has been attempted before, but let's see what has done from a more general perspective (pre-game customization). Here's some examples: 

 1) Character creation - I think this is the most used pre-game customization. All RPG-s use it, starting from Dungeons and Dragons. Often it is the only RPG element in a strategy game. Usually however it is not very deep, as games tend to spread out the customization through the entire early game. Examples include choosing your class at level 10, choosing another focus class at level 20 etc. Often the pre-game customization boils down to choosing a name, race and appearance, for example Skyrim. The reason is simple, if you made the customization deeper, players would have hard time relating their choices to the game, might be confused and the flow of information might be just too overwhelming for some players at least. 

 However, I've missed from many games an opportunity for advanced start, meaning that once you've achieved a certain level/beat the game, you would unlock advanced start for your account, meaning you would have deeper customization, starting out as a more powerful character than first time etc. The purpose would be that while playing the game, if you have already played the game at a certain level, it is not fun any more to play it again from the beginner levels, so if you could advance customize your second character, it would be a time saver, and you could start off from a closer to your skill level, not having to waste time to try out a different tactic, or to simply find out if another class is more interesting to play (In guildwars 2, while being an otherwise good game, I just couldn't bother to level up another character, as the second time seemed to be just too... dull, especially considering how little skill was involved during the first 30 levels). 

 2) Magic the gathering drafting - I'm not sure how many of you have tried drafting in MTG, since it is pretty competitive, but basically it's an example where players spend around 30-50 minutes to make decks on the spot from freshly used boosters. Drafting is a separate game itself, with it's own strategy, meta, required skill-sets etc. So it's a pretty much a mini-game, that influences the outcome of MTG matches played after the decks are finished. In Swiss format, you play 3 matches with the freshly drafted decks.

 For some mtg players it is the most fun format, as it creates for much more freshness each time you play it, compared to constructed play (using pre-made decks from your entire card collection). Here, the pre-game customization goes very deep, and it is enjoyed by players, that are already deeply familiar with the game. They (and occasionally I) go through all that trouble to simply have a fresh experience from mechanic, that has lots of chance and unpredictability, but also a ton of skill. 

 3) Civilization 4 advanced setup - It was there, I tried it once, I personally didn't like it. It just seemed to take away lot's of story from the early game - the exploration, struggling with barbarians, and it didn't add anything to the table. I guess the reason why it was attempted was if someone got bored of the early game. Unlike your usual MMORPG, then in civ series, the early game is often the most fun aspect of the game. That's where the most important things happen, that's where you decide which way your civilization will go. That's where you hope you will get something from a hut that will be exponentially beneficial for you for the whole game. So an advanced setup to skip even a little of the sweet old early game was just something I personally didn't touch twice. 

 That's basically all pre-game customization examples I could think of during a small amount of time, but basically the reasons to do it, is to skip some part of the game, save time or create for more variety. 

 In my new idea, I wanted to use pre-game customization for creating an unique scenery every time you play it and to create really this immersive feeling as the initial draft is sort of like telling the pre-story for the game/plot to come. However from drawing the parallel to MTG drafting which is enjoyed by already long-time MTG players, and being intimidating for new players, I might reconsider. 

 One solution would to have a first game mode - all players pick a pre-made starting buildings and abilities and events are drawn randomly from all the events, instead of players putting them in. And the drafting mode could be explored after 1 or 2 initial play-throughs.

Thanks for reading, and don't hesitate to share your comments and ideas on the matter!

Edit: In +Board Games discussion, some extra very good examples were pointed, such as Warhammer Miniatures game army building and Twilight Imperium Galaxy building. From that I get the feeling that adding pre-game is mechanic for usually hardcore games. I guess it's cos what it actually does, it gives players the opportunity to get a fresh experience with some or lots of extra effort. So it really doesn't make sense to use it in a light game, which players would prefer playing as a filler anyways. Only exception that I would see, is if the pre-game, for example drafting, would be as exciting or perhaps even more than the game itself.