Wednesday, July 07, 2021

Proposal: Get Smart, Richardo

Self-killed. Failed by Kevan.

Adminned at 08 Jul 2021 16:52:28 UTC

Add a new subrule to the rule The Crypt of Dracula, called Threat:

Each Room has a score for Threat, which must be an integer that cannot be lower than 1, and which is not publicly tracked, although it may be noted in the public tracking page for convenience.

A Room’s Threat is calculated as follows, with effects applied in the order that they are listed (any effect that would reduce a room’s Threat below 1, even if only temporarily, instead sets it to 1):
* Each Room has a base threat of 1.
* For each Threat Feature whose Effect is in a Room, add 2 to its Threat.
* For each Denizen in a Room, add 1 to its Threat. For each Denizen in that room whose combined Cunning and Brutality is greater than 5, add an additional 1.
* For each Opportunity Feature in a Room, subtract 1 from its Threat.
* If a Room is a Sepulchre, add 1 to its Threat.
* If a room is Daunting, or is the room from which Richardo most recently moved, add 3 to its Threat.

When Richardo Picks a Destination, he must:
* Evaluate the rooms to which he could legally move (the ‘pool’), which by default are the Lit rooms that are orthogonally adjacent to his current room;
* For each room in that pool, calculate its Threat;
* Secretly-randomly select a room from that pool, which each room’s chance of being selected being inversely proportionate to its relative Threat (specifically: for each room, divide 1 by their Threat, and then allocate outcomes proportionately; eg if Richard were Picking a Destination and his choices were a room with 3 Threat, 5 Threat and 1 Threat, then the first room would have a roughly 22% chance of being selected, the second would have approx 13% chance, and the third would have 65% chance.)
* The room so selected is the Destination for that Move.

In the rule Richardo’s Journey, replace the text “Secretly randomly select a Lit room that is orthogonally adjacent to his current room and change his location to be that room. When randomly selecting a room for this step, Richardo may only select a Room which is Daunting and/or was his immediate previous location if it is the only option in the set of rooms he is randomly choosing from” with the following:

Pick a Destination, and change Richardo’s location to be that Destination.

In the rule Features change “Richardo von Nestor’s next Move in this Enter the Crypt action must have his immediate previous location as its destination” to “When Richardo von Nestor Picks a Destination from this Room, every room other than his immediate previous location is considered to have a Threat of 10.” In the same rule, in the Threats table, change each instance of “is blocked” to “gains 2 Threat”. Update the effects of any Rooms that have any of these Features accordingly.

In the rule Treasures, change the text “Richardo immediately uses this item and must randomly select from any Lit room to be his next destination” to read “Richardo immediately uses this item. The next time Richardo Picks a Destination, the pool includes all Lit Rooms; this supersedes the impact of any effects in his current room, and the effect of any other Items Richardo might be holding.” In the same room, for the four Compass items, change the text “Richardo selects that option and ceases to hold this item” to “Richardo uses this item and the next time he Picks a Destination the pool only includes the room to that direction; this supersedes the impact of any effects in his current room.”

Please check my maths on inverse relative threat, I am very stupid and had to think about it for a long time. (I think I’m basically doing: any room has an inverse threat score of 1/x, which then get scaled up to whole numbers - does that sound about right?)

Also this might be a disruptive change to the way the game works relatively late in the day. I have never really liked how random Richardo’s movement was so this is attempting to smooth out some of the lumps.

Comments

Kevan: he/him

07-07-2021 10:12:42 UTC

Inverse thing looks good. I don’t know if your working needs to be spelled out more literally for players who aren’t exactly sure what you mean by “inversely proportionate” (which is probably ambiguous from both extremes of how much maths someone knows) - I expect it’s enough that it’s checkable from your 22/13/65 example, though.

Some ambiguity in the “cannot be lower than 1” and whether the bulleted calculation list is meant to be applied in sequence, or all simultaneously. If a Daunting Sepulchre contains an Opportunity, is its Threat four (1-1+1+3) or five (1-1 can’t be done so it remains 1, +1+3)?

Josh: Observer he/they

07-07-2021 10:51:53 UTC

Thanks; I’ve tightened up the language on both points.

Kevan: he/him

07-07-2021 11:02:10 UTC

“cannot be lower than 1” might need another turn of the wrench. Does a boring room with a single Opportunity cost 1 (because its total Threat of zero would be less than 1), or is it illegal to create (because doing so would be giving that room an impossible Threat score)?

Josh: Observer he/they

07-07-2021 11:05:18 UTC

Better?

Kevan: he/him

07-07-2021 11:07:07 UTC

Yep.

ais523:

07-07-2021 15:35:44 UTC

against I fear that this will make things much more random, rather than less random; it changes absolute effects into mere probability-changers. In particular, it’ll make it almost impossible to route Richardo into squares which would gain you influence, because those squares will naturally be very threatening, and your only option for making him more likely to go there will be to make the other rooms around there also a little threatening, but it’ll be very expensive to try to match the threat level – and even then, it just sets things back to random.

As a consequence, the only way players are likely to get substantial Influence under this model is via a complete fluke in which Richardo randomly happens to decide to go somewhere dangerous and dies as a result.

There’s also a wording issue, in that this creates a feature that attempts to give Threat to exits, but Threat is a value that only applies to rooms.

Finally, this makes Compasses incredibly overpowered – they could now send Richardo to Daunting rooms, or override blocked exits, because all those moves are technically legal (just unlikely). So the game would mostly becomes about piling Compasses into chests and hoping that Richardo picked yours rather than someone else’s.

Josh: Observer he/they

07-07-2021 16:36:35 UTC

@ais I agree with your descriptions, but not your judgement; I think that the current ruleset is constructed around an essentially passive form of play, in which players set up deadly rooms and wait for Richardo to walk into them, while this proposal changes the direction of the game, forcing players to be more active in direction Richardo as they prefer. It is different but not more random; just more steerable.

ais523:

07-07-2021 16:45:05 UTC

I don’t think there’s any reasonable way to steer the game under your proposal (other than Compasses) – basically nothing anyone does has any strong effect. So it’ll still turn into “set up deadly rooms and wait for Richardo to walk into them”, just that walking into them becomes less likely and you have no real way to control that.

If you want to remove the incentive for gameplay to be around making deathtraps, you need to give some reasonable source of Influence other than deathtraps; changing the probabilities of Richardo walking into one isn’t going to change the fact that no other Influence source is available (thus players need to make them anyway).

I also dispute the assertion that current gameplay is essentially passive (something which is likely to become clear if we have another couple of Expeditions under the current ruleset). I think it’d become a lot more passive with these changes, rather than less so.

Brendan: he/him

07-07-2021 20:38:48 UTC

for

ais523:

08-07-2021 00:31:56 UTC

I just noticed that this also makes large rooms full of Industrious Denizens overpowered – the Denizens look very threatening to Richardo (even though he could kill them easily) so he becomes unlikely to go there, and due to how seriously nerfed all the methods of directing Richardo’s movement will become with this change, there will be no sensible way to attack them (not even Crossbows work because Richardo will never go to the room with the Denizens in the first place).

Please vote for my proposal (which reduces randomness), rather than this one (which increases randomness to the extent of making the game effectively unplayable, and will probably lead to Jumble or Chiiika winning in a couple of weeks because nobody will be able to accomplish anything without extreme luck).

ais523:

08-07-2021 02:32:11 UTC

I just realised yet another issue with this proposal: because it makes pretty much everything in the game do nothing but change the weights on secret random numbers, it’ll become impossible for anyone else to check Josh’s work.

Josh has made quite a few mistakes evaluating the Enter the Crypt action already this dynasty (such as failing to spot rules loopholes that affect how the action resolves – sometimes, which nobody spotted until later – or missing the fact that a room contains a Denizen). That isn’t really a problem; everyone’s human, and the ruleset can get quite complicated to interpret correctly (especially if people are slipping scams into it intentionally; Kevan has admitted to doing this, for example). But under the current rules, in which every legal move is equally weighted, any mistake will be visible, and a player who spots a mistake, or maybe a player who introduced a loophole that Josh missed and has taken at face value, can let Josh know that something is wrong and allow the Enter the Crypt action to be continued from the first step that broke.

With this proposal enacted, there won’t be any way to do that double-checking; as it causes almost every dynastic effect to do nothing but simply change probabilities, any mistake will look like a legal move. So for example, if Josh accidentally causes the “The north exit is blocked.” feature to change the probabilities at all (rather than having no effect, like would be the case under this proposal, or blocking the move outright, as would be the case under a sensible ruleset), there will be no way for any of us to be able to catch that the mistake happened – the atomic action will fail, but be updated on the wiki, and we’ll all be spending Puissance we don’t have and generally making a mess of the gamestate.

Even if it weren’t for the a) invalidation of what many players have been working towards, b) massive gameplay imbalances, c) causing things that are weak to be much stronger (Compasses, Denizen stacking) or strong to be useless (exit blocking), d) massive overemphasis on the centre of the map because Richardo basically just degenerates into a random walk and squares far from the centre are unlikely to be reached via one of those, and e) gameplay degenerating into having compasses pointing towards your deathtraps because there’s no other way to get any Influence, I think the problem of simultaneously making the update action much more complex and removing any ability of anyone else to check it is enough to make this proposal a disaster by itself. It’s one thing to have a dynasty that revolves around the Emperor, but another thing to require them to be utterly superhuman (never missing the existence of a Denizen, or a scam that causes a dynastic rule to not function).

ais523:

08-07-2021 02:46:56 UTC

Oh, and here’s another fun bug: you can end Richardo’s run by pointing a compass at an unlit room. The previous version of Compasses had protection against that; this overpowered version is so overpowered that it doesn’t.

lemon: she/her

08-07-2021 03:20:50 UTC

against

Clucky: he/him

08-07-2021 04:11:37 UTC

imperial

Lulu: she/her

08-07-2021 04:41:50 UTC

against

Kevan: he/him

08-07-2021 06:59:08 UTC

against as this is probably too complex a simulation. I’d marvel at Josh’s spreadsheet for it, but as a player it makes it harder to visualise what’s likely to happen.

[ais] “especially if people are slipping scams into it intentionally; Kevan has admitted to doing this, for example” - not quite, I’ve just been voting through mild loopholes from other players, intending to exploit them later if the opportunity arose.

Janet: she/her

08-07-2021 15:36:43 UTC

against

Josh: Observer he/they

08-07-2021 16:48:50 UTC

against s/k