Spatial Delivery

Originally posted on Tue 27 of May, 2008 [11:06am]

Spatial Delivery won the 2008 KublaCon game design contest, held on Memorial Day weekend in Burlingame, California.

I'll admit I was surprised. But the host and playtesters were enthusiastic about the game, and several of them told me individually that they hoped to see it published so they could buy copies for themselves. This feels really good, especially since the last game I submitted to them (Brimstone, in 2005) was judged to be "monotonous and quite boring." (A phrase like that sticks in your head.) But apparently my design skills are improving!

The best news is that the host has recommended Spatial Delivery to a major game publisher, and the publisher has asked me to send a prototype (not sure if I should name the publisher, so I won't). And then, of course, I get to sit back, chew my lip, and wait (and wait, and wait, probably) to hear back from the publisher. Fortunately, I've got a new design on the front burner to keep me busy and distracted! If the first solo playtests show promise, I'll start a new blog for the new design.

Originally posted on Mon 26 of Nov, 2007 [09:56pm]

The rules themselves haven't changed (nor have I managed any more playtests) since the last post, but I have at least re-written sections for greater clarity. Clarity (I've often found) comes from good illustrations, so I've added one big one full of examples of legal and illegal builds, and updated a couple of others.

The updated rules are here.

Originally posted on Wed 30 of May, 2007 [02:41pm]

At KublaCon on Friday evening I met with Seth (sedjtroll) and others to test Spatial Delivery and Seth's Terra Prime (which is much improved from last year, by the way, and a game I'd be happy to play again.) Spatial Delivery got its first five-player playtest, with Seth, Tom Jolly, Candy Weber, and Seth's friends Peter and Julie. I just watched and played banker.

The good news: nobody hated the game, everybody got interested and involved, and it all basically worked. The challenge system was well-received, with players saying both that they liked it and that they had never seen anything quite like it before. Five players were not too many, which was a relief; someone even suggested I should try it with six. The new scoring system that I devised the week before was a definite improvement. All players gave me useful feedback and suggestions, especially Peter and Julie (and Seth has promised a shedload of suggestions when he has time to write them down).

There was bad news, too. Two problems surfaced which must be addressed. The first was that at the start of the game, a player who gets no Exploration Cards (or "rocket cards" as the players began to call them) cannot build and therefore may have nowhere to move or deliver. There are a number of fairly obvious ways to solve this problem, but I'm going to leave this one alone for now because the solution to the next problem might also solve this one.

The biggest problem is that it's difficult to tell when the move-and-deliver phase is over. This is because the phase ends when every player announces that he is done with whatever he wants to do. But movement is free, so there's really no reason for a player to declare that he is done unless he is completely out of playable cards. There's always the chance that some other player will pull off a late build and open a new market for deliveries, so players want to wait and see whether that happens. If nobody actually does a new build, they wait for each other forever.

There are a couple of possible solutions. One is to have players declare when they are done building, after which they may still move and deliver but may not build. That would break the deadlock, but it feels a bit like an inelegant "patch" rule.

Seth suggested that movement should not be free, and that players should be required to either spend cards on something or drop out. This makes good sense, but I am afraid that it will unbalance the whole economy and greatly change the feel of the game. That isn't necessarily bad, but it's daunting! Yet if done correctly, it could make the game more interesting by giving the players more significant decisions to make.

I'd like to fix these problems with minimal violence to the current ruleset. I've gotten enthusiastic responses from playtesters who would like to play again, and one who has said she would buy the game on the spot if I had a copy to sell! So I am reluctant to risk big changes that would alter the feel and flow of the game. At the same time, I'm willing to consider almost anything that will add interest.

I'll post again after I've had time to playtest a number of possible fixes. In the meantime, my thanks go to all the playtesters, and especially Peter and Julie who took time on Monday morning to give me more feedback, discussion, and suggestions.

Originally posted on Wed 09 of May, 2007 [02:58pm]

There are two problems I've been worried about. One is ensuring that the players are differentiated: that is, that they're not always after the same things all the time. Different goals make for a more interesting game. The second is that I want to give my players something to build during the game. By "build" I mean a large construct built up over time, not just (for example) a scattering of unrelated stations. This is a personal preference — I like building things — but it also contributes to story arc, and it can lend tension to a game if your final score depends on some construct or accomplishment whose completion is uncertain.

In early versions of Spatial Delivery, the construct was your private network of routes and stations; but the fact that it was "private" was hurting interactivity. In current versions the network is shared, but that took away much of the feeling of building something of your own. To restore that feeling, I instituted a pyramid scheme of bonus rewards: you were building a towering score, so to speak. And that has helped, but not as much as I'd like.

I am now considering (and haven't yet tested) another tweak, intended to aid both problems. The plan is that the first delivery to any world would gain the player a token marked with the kind of good that was delivered. At the end of the game, there would be pyramid bonuses for each player, for each kind of good (instead of one big pyramid for all first-deliveries together).

There are five worlds of each kind. If the pyramid schedule is 1, 2, 3, 4, 5, then you can get 1+2+3+4+5 = 15 extra points by making all five first deliveries of (say) Droids. The fifth and final such delivery is therefore worth six to eight points (5 + ring number of the world), which is substantial.

I'm hoping that this will add focus to the card draft, as players with three or four deliveries of a particular kind will want more deliverable goods of that kind in preference to other kinds. Success in claiming a high-value first delivery will give that warm "big bomb" feeling to the players, and the ability to beat someone out to such a delivery will tempt players to more interaction and screwage.

To further differentiate players, I'm also considering two changes to movement. One would be to change "you must stop at every station" to "you must stop at your own stations, and you may stop at other players' stations." This will create some speed differences to spice up the delivery races, and will give the players more to think about when planning their routes.

The second movement change would be to give each player a limited number of "Hyperspace Jumps." Using a HyperSpace? Jump lets you instantly teleport to any station. I'll have to playtest to get this right: you might get only one or two jumps per game, or one per round. You might only be able to jump to your own stations, instead of to any station. The idea, of course, is to give the players another decision to make: when and how can I best spend my jumps?

Now all I need is some spare time to get some playtesting in!

Originally posted on Mon 16 of Apr, 2007 [12:05pm]

Apologies in advance for a long, rambling post.

Lots of solo playtests in the last week or two have shown me a number of flaws in my design. I've made a number of experimental changes to try to fix them, and often I find that fixing an old flaw makes a new one pop up. In some desperation I've wandered away from small tweaks into fairly major changes.

Building the board

Case in point: originally, the board was built up by adding a few new hex tiles every round: "new worlds being settled." The problem was that there wasn't room in the early rounds for all players to go into the richest areas of the layout, yet going in a different direction was usually suicide: there were rich areas yet to be revealed but you didn't know where they'd be and you couldn't plan ahead.

So I tried laying out the entire board at the start, and all of those problems went away, to be replaced by a new one: the middle of the board got ignored. It robbed the game of its story arc. Instead of starting small in the inner worlds and working their way outward, players would just save Exploration Cards (XC) for a couple of turns and then go for the more-valuable rim in a single big build.

Shared infrastructure

Another example: The explore-and-deliver phases of the game lacked interest because everyone's best moves were too obvious; and (I decided) they were too obvious because players had too few choices.

So I switched to a "shared infrastructure" system, where you could use other players' routes and stations for deliveries and for starting points for your own builds. This does seem to open up the play; but it takes away the feeling that you're building something of your own. Originally each player built and used his own routes and stations, so there was a strong feeling of ownership. It bothers me to take that away: I like games where you get to build something, and your success depends on how well you have built.

The latest test

In my latest solo test, I tried a couple more innovations. With only one test so far I can't be sure how well they work or how much I like them, but I think they're at least worth another playtest or two.

First, the board: I compromised by laying out the two inner rings at the start of the game, then adding the entire third ring after the second round of play. This seems to have the desired effect: players develop the inner rings because the outer doesn't exist yet; there is room for all players to act; and the shared infrastructure means nobody is penalized for guessing wrong about where rich worlds will appear later on. So far so good.

Second, the shared infrastructure: I instituted a "pyramid" bounty system. Each world has two tokens placed on it when its hex is first placed: one goes to the player who first builds an adjacent station, and the other to the player who first delivers to that world. You always get VPs as usual for building stations and making deliveries; but at the end of the game, you get extra VPs depending on the number of first-build and first-delivery tokens you have claimed. The reward schedule is the usual pyramid: for 1,2,3,4,... tokens of each kind you get 1,3,6,10,... VPs.

The pyramid rewards motivate the players to build often and to build into new areas of the board, so that by the end of the game the entire board is accessible. They also motivate players to be first to reach new worlds with the right goods; this puts some extra bite into the resource-gathering phase. And they provide a "big bomb": getting a seventh first-build or first-delivery is hard to do and is worth 7 points over and above the usual rewards.

Some of the anonymity of the shared infrastructure is eliminated by giving a player 1 VP every time an opponent uses his station for a delivery. So you not only want to build "first stations" but you also want to build them in desirable locations, adjacent to two or three high-value worlds.

And third, the resource-gathering. This used to be an auction for random batches of five cards. The problem was that the batches were too random. A good batch would produce a lively auction, no problem. But a bad batch would result in somebody just getting stuck with it, which wasn't fun for that player and didn't promote interesting gameplay in general.

I tried a number of schemes for improving the batch quality, but none were satisfactory. Some of them took too long to play; some were too "fiddly"; some were just stupidly artificial, obvious "patch" rules.

So I got radical and tossed out the auction in favor of a competitive draft, as follows.

A pool of four random cards is turned face-up. In clockwise order, players take turns taking a card until each player has successfully taken five cards. There are two ways to "take" a card. The first way is to just grab one from the pool; the empty slot in the pool gets refilled, and you're done. The second way is to challenge an opponent for a card the opponent has already taken. Each player gets two "challenge tokens." To challenge, you give an opponent one of your tokens, and indicate one of his cards. The opponent takes your token and either gives you the card, or refuses by giving you all of his own remaining challenge tokens. If he has none of his own tokens left, he cannot refuse you the card. To minimize abuse and non-productive "take-backs" you are forbidden to challenge anyone whose token you hold, or who already holds one of your own tokens: that is, you get at most one such interaction with each of your opponents.

This draft-and-challenge system seems to work fairly well. Players get to try to build their hands instead of just choosing among a few pre-made hands. There's tension in simply taking an obviously valuable card, because someone else might try to take it from you. There's tension in challenging for a card, because you might get refused. There's tension in using your challenges too soon, because you cannot refuse your opponents' challenges if you run out of tokens; but there's pressure to go ahead and challenge because sometimes you really don't want any of the cards in the pool. And there are tactics. It can be worthwhile to challenge for a card even if you think you will be refused, because if you are refused you have essentially lost your turn: so you haven't been forced to take a card you don't like, and you can hope the card pool will have something better in it when your turn comes again.

Results so far

I've done just one playtest with these rules. It went well, I think, but I don't understand the dynamics I've created well enough to decide how much I like it yet. So far it seems encouraging.

* I definitely like the card draft.
* If one test is any indication, the winner of the game is not clear until the game is over, whereas in the old rules you could spot the winner by the end of round 2 or 3 (of seven!).
* The old turn-order bias may be gone as well, eliminating the need for mechanisms to compete for turn order in every round.
* The final scores were so tightly bunched that I worry that the game is now too balanced!

The game seemed to be over after just five rounds, because all possible new builds and new deliveries had been accomplished. A sixth round felt superfluous. This is probably a good thing, as the gameplay is too light to justify a long game; but it may be that there's room for one final round of cashing in on what's already been built.

I'm concerned about fiddliness. When you make a delivery, you must:

(a) Surrender an appropriate Goods card;
(b) Place a Sales Marker on the hex (so no one else can deliver there this round)
(c) Take the first-delivery token, if it's still there
(d) Take your VPs for the delivery

Written out like this, the fiddliness actually doesn't seem too bad. Maybe it just seems like a lot because I'm trying to play four players at once in my solo tests. :-)

I'll post again after I've had a few more playtests.

Originally posted on Sun 08 of Apr, 2007 [05:50pm]

Throw off your shackles!

Two constraints of the GDS that inspired Spatial Delivery were that the board should be built up by the players during the game, and that the bidding would be for strictly limited numbers of items. I've had a lot of time this weekend to work with the game and run solo playtests, and I've discovered that those two constraints were causing a lot of the problems I was having. By dropping them both I think I've got a much better game.

First, building up the board: this sounded good in theory. Players' initial Trade Routes would be only to the visible Settled Worlds. As more Settled Worlds appeared, there would be fresh competition to reach those newly-revealed destinations. This would impose a natural story arc on the game, and also provide a handy game clock: the game would end in the round when the last hexes were placed.

But ii didn't work that way. Because players couldn't see the whole layout at the start of the game, they could easily be "marooned" by building toward the best currently-available locations, only to find later that they were dead ends. The simple luck of the hex-tile draw could doom a player's game.

Furthermore the first round of auctions was almost pointless: with so little of the board visible, nobody had a clue what might or might not be a valuable batch of cards.

Finally, the placement of the newly-revealed hexes was a problem. The board areas where the final hexes were placed had little effect on the game because they didn't exist for long enough. I experimented with "spread-out" placements that wouldn't favor one direction over another, but nothing worked well and all of them would have been difficult to explain in the rules.

So I've now tried three games with the entire board laid out at the start, and all of those problems magically went away. Players could see the entire board, form long-term strategies at the start, use the entire board at will, and intelligently evaluate the auction batches. Phew!

The "limited supply of goods" was another kind of problem. I had planned the card deck of auctionable Goods to be exactly enough to last the game. This required a deck of 120 cards, which is a lot to manufacture (see "The Death of 1,000 Cuts" below) and an awkward handful to shuffle. So I broke the deck's association with the game clock: reduced the deck to a much more manageable 80 cards, increased the game length by one round (which added a surprising amount of depth without adding much more playing time), and just recycle the discards when the deck runs out.

Fixing the auctions

I've been searching for a way to pep up the later auctions in each round, and I've found it, and it was simple: instead of five cards, the final auction each round is for only four cards. This makes even the last two bidders interested in winning the penultimate auction; yet it is not so harsh a punishment as to be a disaster for the loser.

The "Big Bomb" fizzles

A couple of posts back, I was excited about a "big bomb" concept: Planetary Upgrades that players could purchase to improve the payoff of selected Worlds. I thought it through very carefully and was sure it would work, but it didn't. It was too "chunky" somehow. Players able to take exclusive advantage made too many VPs, and I couldn't see a smooth way to tone the upgrades down without making them pointless. I had hoped that they would foster competition, since the upgrades would be available to anyone making a delivery to an upgraded world; but players simply wouldn't purchase upgrades unless they were pretty sure they'd get most or all of the benefit. So I had to throw that idea out.

Seeking a new balance

Today I had a new idea, and it seems to work fairly well (in solo testing, anyway). I've been worried about having runaway winners and hopeless losers, and I've been seeking a way to provide a bit more balance. The runaway winner is a player who gets some really good Stations on the outer ring, early in the game and pretty much to himself. The hopeless loser is the player who gets the worst Stations and can't ever seem to catch up. The new idea is the Nonesuch (which really needs a better name), and it will take a bit of explaining. It works like this:

The board is 2/3 Settled Worlds and 1/3 empty Space. Previously, Space hexes were worthless. But now a species of deep-space alien has been discovered: they avoid planetary systems, and can be found only where three empty Space hexes meet. Players who build Trade Stations there can obtain a cargo of Nonesuch to take back to Earth. Delivering a Nonesuch to Earth gets you 3 VP, the same as the highest-valued normal Delivery; but the catch is that Earth, like any other world, will accept only one delivery per round. The first Nonesuch delivery pays off; all others rot with no payback. Delivering the first Nonesuch therefore usually means being late to deliver normal Goods to other worlds; it is therefore (usually) profitable only to players who have nothing better to do: in short, the losers. It offers a kind of second path to victory: for the right player it can be a guaranteed 3 VP per round, which is quite significant. In test games it raised a player who would have been dead last up to a position at least in the middle of the pack, and in one case only one point behind the winner. It also added new tactical decisions for players to make: should they try to get the jump on delivering their normal Goods, or go for the Nonesuch this round?

Playtesting on the Cheap

Hey Seth: I think you could maybe playtest with a Settlers set now, if you've got an expansion with extra hexes. You still need 5 hexes in each of four colors. For cards you can use a deck of 19 of one kind of resource (for XC) and 5 each of all others. You'll be recycling the deck every other round, instead of just once in mid-game. Now get some money and VP counters and Sales Markers, and you're ready to go!

Originally posted on Wed 04 of Apr, 2007 [12:25pm]

Jonathan Degann wrote four great articles under the heading of Game Design 101. I strongly recommend them to anybody with any interest in Game Design:
http://www.thegamesjournal.com/articles/GameTheory1.shtml
http://www.thegamesjournal.com/articles/GameTheory2.shtml
http://www.thegamesjournal.com/articles/GameTheory3.shtml
http://www.thegamesjournal.com/articles/GameTheory4.shtml

In the second article, he discusses a feature he calls a "big bomb": a disproportionately large reward that players can attain by gaining enough of an edge on their opponents, sometimes by taking a leap-of-faith risk. Big bombs can introduce new tension to a game, and offer alternate paths to victory.

I've been worried that Spatial Delivery was a bit too bland and straightforward, and I've been wanting to add something to spice it up. Seth (sedjtroll) has been after me to add some kind of "upgrade" feature to the game. He's right; the game needs a big bomb, and I think it's fairly clear how to do it.

I plan to introduce a set of unique tokens, each representing an "upgrade" that can be applied to a selected Settled World. Players can purchase and apply upgrades during the Exploration phase. Prices will be high; they will vary depending on the likely value of the upgrade, and will be paid in combinations of cards and money.

A first-cut list of upgrades includes:

* +1 VP per delivery (or maybe +1 in second ring, +2 in third, or one of each)
* Two deliveries per round
* Accept any kind of Good
* Accept extra delivery of specific kind

Again, each purchased token is placed on and applies to a specific Settled World. Its special power applies to anyone making a delivery to that world. The "specific kind" type shows one of the four kinds of Good, and allows one extra delivery of that kind of good to the world it is placed on. It might also be allowed to place those on an empty-space hex, to allow deliveries there as if it were a Settled World of the token's kind.

The high prices mean that players may have to choose between trying for an upgrade token, or just plowing ahead with normal development. Tokens that cost money will obviously cut into a player's bidding power. Varying prices and powers mean that players will have to choose between grabbing a cheap token while it's still available, or holding out for a high-power one that may be grabbed by someone else first. The desire to get the sole advantage of an upgrade should also make routing more interesting: players can either strike out for less-fruitful areas of the board, hoping to monopolize an area and enhance it; or compete for the good areas, hoping to horn in on any upgrades placed by their opponents.

Other than the need to balance the prices and effects carefully, I don't see any drawbacks to this notion. Anybody who does, please let me know!

Oh, and I need a thematic excuse and title for these things. "Economic Develpment Investments" works but doesn't exactly sound exciting!

Originally posted on Sun 01 of Apr, 2007 [04:48pm]

A prototype of Spatial Delivery can be cobbled together out of Settlers of Catan parts, plus money and VP counters, and the custom card deck. In a pinch you can use a grab bag of colored cubes instead of a card deck, if you have enough of the right colors. (You'll need a lot of black.)

(Note: The rules at the link below will no longer work with the prototype PDF!)

I've posted a PDF with full instructions for assembling such a prototype. It includes two pages of card images, which (after you print enough copies of each) allow you to build your own 120-card deck. The PDF can be downloaded from here: http://www.semicolon.com/Boardgames/Files/SpatialDeliveryProto.pdf
and the rules are here: http://www.semicolon.com/Boardgames/SpatialDelivery.html

Just so you know, you'll need a Settlers expansion (Seafarers or 5-6 player) to get enough of the right kind of hexes, or else you need to do a little extra work to disguise some of the tiles from the basic game. The instructions tell how to do it without damaging your Settlers hexes.

Eventually I'll have custom hex artwork as well, for those who want to go whole hog; but that's not ready yet.

Originally posted on Fri 30 of Mar, 2007 [09:21pm]

Spatial Delivery was an entry in the March 2007 Game Design Showdown, "Future Business." My initial solo tests showed me that that entry was seriously flawed. I didn't have time to improve it for the showdown, but it seemed to have potential so I have kept on developing it. The current rules are different in a number of ways.

The first problem I found was that it was too easy to predict the value of a batch of Goods in the auction. Goods would be bid up to a point just below their expected value, so profits were low and sometimes negative. There was still a good game in trying to out-earn your opponents, but it was unsatisfying for the winner of a 90-minutes game to have only a few more dollars than he started with! To fix that, I changed the goal of the game from making money to making Victory Points (called "Market Share Points" in the game).

This change also meant that the players' starting cash was all the money they would ever have! Suddenly the budgeting got very tight, because your auction purchases would not get you more money. Your initial seed money would have to last you the entire game. This is good for a couple of reasons: among them, it prevents a runaway-winner problem because the leading player can't use his profits to bully his opponents in later auctions.

Another problem was that the board was too big; players could wander it at will, and had little need to interact with each other. To solve that, I made the board smaller and borrowed mechanics from Settlers of Catan and any number of pickup-and-deliver rail games. Players must now build up trade routes and establish nearby trade stations in order to make deliveries to planets. The smaller board and its mix of planets vs. empty space crowds players together, and makes them compete for the best routes and trade station locations.

The auction batches now include Exploration Cards ("EC") as well as Goods Cards (oh yeah, Goods are now cards instead of tokens). You need a good balance of Goods and EC to win, because you need EC in order to establish trade routes and stations, and you need Goods in order to profit from your routes and stations.

I also changed the rule that says any planet will buy any Good for a low price. Now each planet buys only one kind of Good, period. This makes it much more important to get the right Goods at auction, because you can deliver only to worlds where you have a nearby trade station. All of this adds interest and tension to the auctions.

I kept the one-delivery-per-planet-per-round rule; this means that there's also tension during the delivery phase. A Good worth 3 VP becomes worthless if other players beat you to the worlds that will accept it, and deliver their own Goods ahead of you.