the case for output randomness

So, I have this boardgame and it has a problem that I am thinking of solving with output randomness - the look ahead problem.

Its an area control game that features combat with card play. Cards are bought publicly in the market and if people are paying attention and have good memories, they can determine whether an attack can succeed if both parties commit the entirety of their resources.

Which results in a standoff where an attacker, knowing he is close but not quite there, commits more resources to his attacking position, forcing the defender to do the same, and it becomes a stand off/arms race for the two players concerned. It doesnt feel good for either party. In a game with 20 actions, having one of those actions be - I commit an extra resource, and then having the defender do the same, so that 5% of both parties actions have been spent without a meaningful change in the board situation is...not great.

The other bad side effect of this determinism is that defenders, knowing when they cant win, will not commit any card resources to the defence, knowing they will be wasted in certain defeat. Either side really only needs to commit resources when they cant remember exactly what the other guy is holding, which does happen, but not nearly often enough. Very occasionally there can be a bit of nice bluff and bluster going on where an over confident attacker does not commit their resources to a sure win and is surprised by a defender that does, but it happens very rarely.

IF I introduce some small random element to combat that can tip a close contest one way or the other.., I feel like it will change everything in a good way.

rather than waste precious actions in an arms race, an attacker who is 'close enough' will probably just say 'effit' and attack. And the choice for committing resources to defence for a defender who is 'probably, but not necessarily' going to be defeated becomes much more interesting.
Is there any value to the defender, as the game is right now, in putting up a "moderate" level of defence? E.g. they lose the territory but still inflict painful losses on the attacker. It sounds like not if you say all defensive resources are wasted in a defeat.

Without knowing any more about it I'm thinking your small random element would be good but only if it was also made the case that close losses were less damaging than heavy losses. At which point maybe it wouldn't be output randomness any more!


Staff member
having one of those actions be - I commit an extra resource, and then having the defender do the same, so that 5% of both parties actions have been spent without a meaningful change in the board situation is...not great.
There are a lot of systems which need output randomness to work. XCom wouldn't really work without the output randomness. It's possible that the game you're working on now requires output randomness also. My claim has never been that "no systems need output randomness", but rather that "we should try to design systems in such a way that they don't need output randomness". I don't know that much about your system - it may be that there are a few rule changes you could make which could solve this issue without adding output randomness, but it may be that you need to. Give it a shot and let us know how it works out.
@richy, more combat resources increase casualties for the opposition, so there is a situational case there for that. But really, its how determinism encourages static lines, or "trench warfare" that is is major issue.

@keith. Im also against output randomness and I kind of shocked myself to be considering it. Basically I am trying to make the game feel more dynamic, particularly for one faction which brings the major military threat to the game. When they are static, the game changes markedly. This is not a bad thing per se, as 'something that can happen', but its become a recognizable trend. Its too easy for that faction to get into a standoff, and too hard to pull out of it once they are in it. Increasing their intrinsic mobility gives them a better chance to redeploy, which is something Im looking at, but go too far in that direction and there is not enough consequence for bad deployment, and worse, ruining the balance of the game by making them too powerful.

Im just really thinking out loud here, but 'changing a few rules' can have a lot of unwanted side-effects whereas output randomness is alluring because it breaks the 'spell' of determinism without really affecting any other part of the game that I can see. The player is still going to be making the same decisions and bringing the same amount of power to bear at a certain time and place, with the only difference being that close fights could go either way.
One idea might be to make it so large battles between entrenched forces are resolved bit-by-bit, with many small instances of output randomness instead of one big one, and the attacker able to respond to the results and e.g. abort an attack at any time if things aren't going well for them. E.g. Civs < 5, or Risk. Alternatively make a bigger rule change and disallow build-up of forces on any territory. One card per territory like one unit per tile in Civs >= 5. Or somewhere in between - max 3 cards per territory or something?
Sure! Specifically, Im looking to break a deterministic standoff where, knowing that they cant win, one side instead gathers more resources on its turn - say recruits another unit, or purchases another combat card, which prompts the opposition to do likewise. Sure, after a turn or two someone will run out of resources but the build up is static and boring. For sure a random element to combat would break that, but this is unsatisfying. A last resort.

What Im going to trial is the ability for a faction to sacrifice units to generate more combat strength. This is a way for an attacker to throw itself over the line, so to speak, at the expense of greater casualties - which is a significant cost in my game. For that matter its also a way for defenders to sally forth and force an issue immediately if they dont feel like waiting for an attacker to build on their doorstep.

So, extra options. And some more mindgames. Is this a real attack, or a light probe designed to waste my resources - that kind of thing.

Obviously I don’t know enough about your game, but something I would be looking out for is a way to hide information in such a way that the information is strategically guessable. In general, this could mean knowing the choices opponents faced without knowing what decision they eventually made. For example, maybe you saw them pick up three cards of which they were forced to pick one, but you don’t know which one card they kept and which two they put back in the deck. Then part of the strategy is getting inside the opponent’s head and trying to figure out what they thought their best bet was at the time.
Its a nice idea.

however its incompatible with the open card market and general philosophy of 'not being hit by powerful random effects' of this game. By the 2nd half, people DO invariably lose track of exactly which cards a faction has in hand. Particularly the cult faction that tends to buy and play a lot of one-use espionage cards and has the ability to meddle in the combats of other factions as a 3rd party.

Its early that the determinism is a problem, and its not really a problem for its own sake except where the side effect is a often a boring standoff between the two major military factions. Its the standoff that im trying to solve.

The thing I like about the sacrifice idea is that it is giving the problematic faction in particular the chance to break the standoff , but at the cost of being able to continue an advance over time. The defender may in fact prefer this outcome - concede the area with as few lost resources as possible. the attacker will also know this is an option for the defender and could risk not committing the extra resources in the hope of winning anyway. bluff and double bluff.
I like the idea! It makes you wonder if it's really a case of output randomness as such. Normally the discussion there is about random *events* generated by the game and how disproportionately disruptive they are. But here the initiator of the event would be one of the players, not the game. It's a chunk of randomness being thrown in deliberately by one player at a place and time of their choosing.

Also seems like it would be in keeping with a warlike faction. Peaceful factions would always prefer stately progress through the application of reason and logic. Of course they would - they're the ones set up to profit from that state of affairs. No alarms and no surprises. A nice bit of output randomness is just the thing for shaking up those fat complacent rotters!
Its more uncertainty, but its not random. making the problem space more complex and providing more options. seems to tick all the boxes, but will it work OK in practice? Ill report back if anyone is interested.
had one playtest. The result was better in some ways but had an unexpected side effect - the aggressive faction lost its position on the board. First time this has happened. The positive was the faction had the option to win close contests by sacrificing units and they took that option on multipole occasions, and created a nice moderate amount of destruction, but in the process depleted their unit position on the board until they had nothing left. This was their choice, so in that respect..OK. Another downside is that large battles took twice as long to resolve due the added complexity.
Another playtest, this time turning a little differently. Again, the nice moderate amount of destruction I was hoping for but the problematic faction didnt wipe itself out numbers wise, and there were several nice cases of gamesmanship on the part of factions - either using a feinting attack to draw an overcommitmant from defenders, or vice versa.

There is still a little tweaking I think concerning what proportion of a force a faction should b allowed to sacrifice, but that is a minor issue that playtesting will shed light on.

so far, Im considering the sacrifice rule a great alternative to throwing dice.