Author Topic: Designing Difficulty  (Read 18773 times)


  • Rogueliker
  • ***
  • Posts: 355
  • Karma: +0/-0
    • View Profile
Re: Designing Difficulty
« Reply #30 on: March 05, 2012, 10:15:17 PM »
Chess is always the best comparison in my mind.  You know the immediate result of your move, and you can guess how your opponent will react, and so you can plan ahead a few turns.  But it's not fully predictable and your tactics have to change and evolve as you play.  The game rewards thought, and because it keeps changing you have to think every turn.  It is never ever boring.

I'm in the camp that doesn't think of Chess as a 'game,' but more as a competitive puzzle.

The problem with chess is that it has no fog of war and the initial board state is always the same. It is not predictable what an opponent may or may not do, but both players can be aware of ALL possible outcomes of any given move.

Chess is then a battle between who has more of the game state tree of chess memorized. If that doesn't determine a winner, it is a battle of objectifiable (word choice? >_>) heuristics. If that doesn't determine the winner, then it's who is more familiar with whatever sub-tree of board states of chess we are in. IFF that doesn't determine the winner, we will finally have a game that is interesting.

Why do I dislike all of that? Because it has the potential to become-->

Too much predictability and it becomes a set puzzle, which is boring since you solve it at the start and then just go through the steps.

The other thing we're getting at here is accessibility. A player's skill rating is more dependent upon memorization than on one's ability to utilize the rules. Because Chess is a solvable game, it would eventually become a set puzzle. A low-ranked player has virtually no chance against a high-ranked player due to a difference in the number of early moves memorized. If those early moves didn't matter, then it would become a contest of who utilizes the rules of chess in the most interesting way.

A simple exercise to change things up is to allow players to place pieces on the board within the first two ranks in turns. If white places a piece first, then black has a slight advantage in initial board state, while white has an advantage in tempo. The change in starting conditions mean that memorization won't favor one player or the other. Intelligent building of the opening board state and clever utilization of the rules would play a larger role. After all the pieces are placed, the game is still completely calculable- we just cut out memorization as a factor in skill. Of course- there are obviously 'good' and 'bad' initial states and these may not matter much what the other player is doing, so memorization can still play a huge role if this style of chess is something that a player practices regularly.

An interesting game to look at, in terms of initial state and fog of war, is this little Italian card game called Scopone (Specifically the variation Scientific Scopone).

The gameplay is simple-
--40 card italian deck, each player is dealt 10 cards.
--Right of dealer plays first.
--Each turn a player MUST play a card (10 rounds per hand).
--If there exists a card or group of cards on the table that add up to the exact value of the card played, those cards and the one played are captured by the player's team. If multiple sets of cards exist, the fewest number of cards must be taken (player's choice on a tie).
--If a player plays a card and captures all the cards on the table, it's a 'scopa' (pronounce it loudly and proudly!) or 'sweep'- unless it is the last round of cards played.
--When there are no longer any cards in hand, the last person to capture is awarded all of the remaining cards on the table.

Points are calculated at the end of each hand. One point is awarded for each of the following:
--Most Cards (unless there is a tie)
--Most Golds (unless there is a tie- 'gold coins' is a suit)
--Highest Primes (generally who has most sevens)
--Gold Seven
--For each Scopa

The first team to reach 11 points wins. Points are only calculated at the end of a game and at the same time- therefore, it's possible to tie. If there is a tie, continue playing until the difference in score is 2.

Table talking is absolutely allowed- There are even specific italian words that players are encouraged to use to tell their partners what they want them to do. The only restriction is that a player is not allowed to announce a card in their hand.

What makes this game awesome? All of the cards are accounted for AND there are a variety of objectives for each hand. The rules are very simple and strategy gets very interesting.

If there is a 4 and a 5 on the board and I don't have a rider (9), I might play a 3 or a 2. My opponent, who has a 7, will want to win a prime, taking the 4+3 or 5+2 (pending on what I played) with his 7, leaving behind a 4 or a 5 for my partner which could result in a scopa for my team. The opponent knows that this is risky, but if he does something else there may still be a combination of 7 on the board passed to my partner- who might also have a 7. Say the board is 4,3,5- he might play a 2 instead of a seven. If my partner plays a 7, he leaves a 7 exposed to the other opponent who could take it as a scopa. Lots of things have to be considered.

Similarly- If one player of team A knows that the next player of team B has a 7, but that the other player of team A does, he can literally pass a 7 to his partner. The only way to prevent this from happening is for the other player of team B to ensure that an ace, two or three is on the board so that their partner can catch that 7 with a royal (face card)- that, however, could then expose a scopa with the remaining card. If card passing does succeed, it can oftentimes result in a scopa for the other team.

Say there is a 2 and 3 on the board and something else. You play an ace and your opponent has a 6. If two to three of these four cards are gold, a player will consider taking the risk of exposing a scopa to your partner. If they don't, however, they're going to pass a bunch of gold to your partner. Same sort of situation as the above two examples.

Being aware of how players pass on opportunities gives us hints as to what they have and don't have- but a player is just as likely to intentionally not play something to prevent the chance of a Scopa happening or to dupe you. With tabletalk as a component, both the rules and social component create a very fun and interesting game. You end up playing and thinking about the game in some really complex ways.

When a team is pressed into a position where they can't make a strategic action it's called a 'whirlwind' (because one team is taking cards when the other can't). A 'whirlwind' of scopas is the ideal progress of a game for one team (typically lots of taunting will result: "And it's become a whirlwiiiiiinnnd"- "is it breezy in here? Oh oh... look what just happened again"), but the game is very interesting because it's not difficult to come back from a losing position because there are so few points awarded. It's possible for the game to end in 2 hands, therefore dealer dis/advantage (though there isn't really any) is balanced by the way scoring is handled and by how many points you need to win. After all, each game both teams can get the same number of points in different ways. You have to play both to pursue objects and deny your opponents.

You can play it as a heads-up game also- just deal the second set of 20 cards after the first has been used up. The first hand is a lot of probing strategies, while the second one ends up dealing with card-counting and tricking your opponent.

The scientific variation causes the game to start as a whirlwind, which many players find frustrating (though IMO a better game)-- traditional rules have 9 cards to each player and 4 starting on the table.

If you have a friend or group of friends interested in alternative games, I highly highly recommend it-

Games are fast and a lot of fun. The regular version of two-player Scopa is in itself a great game. I find it to be much more strategic than Bridge or any other trick taking game.
« Last Edit: March 05, 2012, 10:33:26 PM by requerent »


  • Guest
Re: Designing Difficulty
« Reply #31 on: March 17, 2012, 11:21:12 AM »
  I'll have to check out that game next time I'm up for a card game night. Those happen very rarely nowadays. My group mostly gets together nowadays to play FPS frag fests.

  As for your project; I'm a huge fan of the methodology of Kinsey. His brute force information gathering so as to create rock solid statistics has influenced my academic pursuits in many ways.

  That said I think this year's 7DRL competition will probably produce a TON of games for you to study. I'm expecting there to be at least a dozen top games this year. All short and with easy to access source code (I for one would release my source to anyone who asks). It could take a lifetime to become an expert in more than 1 or 2 major roguelike games. With the smaller ones it is possible to really know the mechanics inside and out and use that knowledge to further your study of design and difficulty.

  Just a thought. 


  • Rogueliker
  • ***
  • Posts: 1112
  • Karma: +0/-0
    • View Profile
Re: Designing Difficulty
« Reply #32 on: March 17, 2012, 09:38:59 PM »
Because Chess is a solvable game, it would eventually become a set puzzle.

That's technically true, but in practice, no human player is ever going to be able to do that.


  • Guest
Re: Designing Difficulty
« Reply #33 on: March 18, 2012, 08:37:58 AM »
  On the contrary. Top Chess players have complained that the game has become a memorization exercise during the several opening moves. Meaning the more creative chess player will lose if he is not also the more 'memorized' chess player.

  You do not need to memorize the entire game for it to be solved. Just enough to beat everyone else.