Random as catch-up...or not

A friend invited a bunch of us to help him playtest a game. I like it, because generally everyone can come up with SOME game they'd like to see made. This will probably become a fallback question that I use when meeting people. What game would you make?

The friend, he is heavy on analysis and generally anti-dice. I guessed correctly that there would be limited randomness (a stack of tiles), and would be driven by player brawls. (Brawl over race...it's not that much of a guess.)

Randomness can be a useful catchup mechanic. Look at it this way - if everything remained status quo, the current leader would win. The other players want some dynamic behavior to shake things up (hopefully in their favor).

In my friend's game, the leader is generally attacked directly by other players. This leads to plenty of see-saw action, but most of the mechanics are driven by money. Want to do X? Pay money. Want to do Y? Pay money. If you don't have money, you can't do actions...and it can be pretty debilitating. In some ways, similar to discard hand mechanics in trading card games. There's a reason why discard is feared as a mechanic. Create an imbalanced economy, the economy doesn't right itself.

His fix was to create an official 'dark horse' position. If you're doing extremely poorly, there is a point at which you get crazy discounts on all of your actions. Again, money. Money is everything, except at the end...and then you need the victory points from buildings that you purchased and still control. It's still got all of the makings of a good game - it's just not my ball of wax.

I shared a game that I am working on. (The other designer's final comment was "I was just trying to figure out how to reduce the randomness in your game." Bwhahaha.)


So I like some randomness. I like being able to choose my odds - maybe play risky if I'm behind, or even just to cement a win in the mid-game. Or just to have smidgens of good/bad luck to overcome. I don't want a game that is flip a coin, get a point for heads. I also don't want chess, where it's very static and players SHOULD paralyze themselves, thinking further ahead.

In currently named "Magic Masters", players start with somewhat random spells, based on the Talents that the players draft in the beginning. As apprentice wizards, they also have some dice that generate mana to cast spells (or research new ones). As the player explores the world and takes over locations, they gain more regular sources of magic as well as other trinkets/abilities. The tiles they explore are random (within the limits of "all are one of 5 elements", "all are octagons", "all can be entered from 4 directions", and "all have some location of interest where you can battle a creature to get something").

What's random?

• Tile exploration (you can reduce some of the randomness by taking the Explorer talent)
• Mana dice (you can reduce some of the randomness by conquering elemental nodes, finding familiars, using certain magic items)
• Combat (it'll get its own post, but your spells use dice to determine if the spell hits or not...and each spell has different combat speeds and different combinations of dice to resolve combat)
• Quests come from a constrained deck, as do monsters. The quests are broken up by difficulty - you won't be taking a Tier 3 quest until the end of the game, but the Tier 1 quests are all roughly even.

A fair bit of mitigating randomness, if the player chooses to spend effort on it, instead of charging straightforward towards victory. How much mitigation the player wants - it's one more way the player can choose their method of play.

Puerto Rico and its blue and yellow socks

We had an excellent game day today, before the snow started falling in downtown Seattle. We opened with quick games of Dominion, switched to 5-player Race, and then we hit Puerto Rico with 5. (We wrapped up with 4-player Settlers and a final game of Race.)

They were all games played with good gamers (who are also good friends). There's always a bit about what feels random...it was on my mind after reading the "Statistically Speaking" article on Gamasutra (link in the sidebar). The topic crept into tonight's game, as we started Puerto Rico. How do you determine who goes first? My standard way is take 5 barrels, 4 corn, 1 indigo...whoever gets the indigo starts. Fast, quick, doesn't need dice or prolonged discussions of how to best use rock-paper-scissors in a group environment. (I'm also of the belief that RPS is not random, and would not be useful in determining a necessarily random behavior!)

My friend said the magic words of "Wait - let me do it!" He has a new way. Take 3 indigo tiles, 2 corn tiles, and let players take tiles clockwise until the system is revealed. For example, Player A takes a tile, reveals indigo. Player B takes a tile, reveals corn. This would mean that Player A is third player with the last indigo, Player B is fourth player with the first corn! (In Puerto Rico, the first three players in the order start with indigo, the last 2 players get corn.)

That certainly seems magical, right? Something seemed fishy, but the bottle of wine that we had been sharing in celebration of the snow seemed to lighten our spirits...so we continued as it was decided. (He showed the system for Indigo/Indigo/Corn or Indigo/Indigo/Indigo - you only needed to go as far as three players.)

Of course, being gamers we yearn to have complete knowledge. We argue over what the best Settlers placement was, whether it's better or worse for random distribution of numbers, ways to eliminate/reduce/create 'luck' in favorite games...and we will now circle back to the magical method.

It makes for a great scam.

We'll start at a table where I'm playing with Alice, Bob, Charlie, myself, and my buddy Jack. I'm the person getting everyone to sit down, and I will use the magical method as described above. I'll choose Alice to start, going clockwise around to Bob and then Charlie as needed. 5 face-down tiles, 3 indigo, 2 corn - just like the game.

Yes, it's true that any player has a 60% of starting the game with an indigo, or a 40% chance to start with a corn plantation. But the next step is where it breaks - we're not dealing with replacement...they're related events.

Let's start with that 60% indigo tile draw for Alice. Alice is happy - she is now one of three indigo players, and therefore has a 1 in 3 chance of going first! Not quite. Bob has 4 tiles left - 2 indigo, 2 corn. If Bob gets a corn tile, he is first corn and therefore Alice is third indigo. Bob draws corn half the time - half of that 60% we had, Alice goes third (30%). If Bob gets indigo, Charlie has 1 indigo, 2 corn left. If Charlie gets a corn (a 2 in 3 chance), he becomes first corn, and Alice's indigo draw earlier puts her as 2nd player. (20% chance, as it's 2/3 of 30%.) The remaining 10% chance is Alice's chance to go first.

Alice could have drawn corn, 40% of the time. Alice would be happy again, assuming she hasn't learned her lesson from indigo discussions. There's a 50/50 chance that she is first corn player! Sadly, the magical method will fail her here as well. Bob would be left with 1 corn, and 3 indigo to draw from - 3 times out of 4, he'll draw an indigo, making her last player.

Alice's breakdown:
1st player: 10%
2nd player: 20%
3rd player: 30%
4th player: 10%
5th player: 30%

Now, if I were a devilishly clever person of low morals, and wanting to be first indigo or first corn (first player & fourth player, respectively), I would look to the person to my right and say "Hey, choose a tile and reveal it for yourself. Now do the same for me, as I finish setting up the game!" 60% of the time, I'd start as first corn or first indigo.

And that ends the story of the horribly magical method, that you should never ever use in polite company...or mine.