Topic: AOE Analysis
Started by: John Kirk
Started on: 12/3/2005
Board: Acts of Evil Playtest Board
On 12/3/2005 at 9:26pm, John Kirk wrote:
AOE Analysis
Paul recently asked me to perform an analysis of Acts of Evil from a design patterns perspective. If you are unfamiliar with design patterns, a discussion of using them for RPG design can be found here. Specifically, Paul wanted me to look at several things:
"Paul wrote:
...I need to make some decisions about the relationships between the various formulae. In particular:
I need to make the occurrence of one particular outcome more frequent across the whole universe of formulae.
I need to determine if there are any dead ends in the formulae, where a player can become trapped by their attribute values with no way to progress.
I need to make certain types of NPCs more effective as antagonists in the early game.
I need to consider ways of inter-weaving the storylines of the individual characters (particularly in the early game), but Acts of Evil has a competitive component, and so I need to do this in a way that doesn't shorten gameplay by making characters too vulnerable to each other.
Before getting into specific design goals, though, I felt I needed to get a handle on the current state of the game. Toward that end, I created a collection of Gauge Diagrams for Acts of Evil that you can download here. These diagrams were created using a diagramming technique somewhat different from that presented in the RPG Design Patterns book. A discussion of this improved technique can be found here. A document succinctly describing the diagramming technique modifications can be found here.
Before getting into the criticisms, let me say that the game has a lot going for it. It has some interesting reward systems that should produce some interesting interactions between the various contests. I don't think I've seen any game with as many reinforcing and balancing loops as this one. The "Resolution against Underlings" contest actually contains four Reinforcing Loops alone! Wow!
(Note: the currently available RPG Design Patterns book does not list Reinforcing Loop and Balancing Loop as design patterns, but I've come to realize that these are very important patterns that will be included in future revisions. If you want to read a little about these kinds of structures, take a look here.)
The Gauge Diagrams tell me several things about the game:
1) Most of the various contests have strong reward systems for the player choosing the contest (i.e. player[sub]1[/sub] in the diagrams).
2) Overall, the game is an exercise in gathering and spending resources. It lacks somewhat in strongly Conflicted Gauges. (I think Capacity would be used as a last resort when a character has no Power and so is somewhat weakened in its ability to influence play.) My guess is that this lack means the game has an optimal strategy. The introduction of more Conflicted Gauges would help here in giving players a wider variety of strategies to winning.
3) In none of the various contests are any real rewards given to other players participating in the scene (i.e. player[sub]2[/sub] in the diagrams). I believe players will therefore only be interested in scenes in which they choose the contests. At the very least, they will probably be noticeably less interested in other scenes. This also feeds into the issue of NPC's needing to be more effective. In some contests, the other players may expend resources to make the NPC's effective. To encourage this, you need to provide rewards to the other players if the NPC's win. In some contests, the other players have no input at all, and so their interest in those scenes will be zero. In my opinion, you need to provide a way for player[sub]2[/sub] to participate in all scenes and reward him for doing so.
4) The “Change Nobody to an Underling” contest is weak. No rewards exist other than the gaining of Underlings, which only helps out when a player takes on the role of player[sub]2[/sub] in some contests. Since no rewards are given for player[sub]2[/sub] in those contests, I doubt anyone will have a strong desire to choose the "Change Nobody to an Underling" contest.
Here are some suggestions I have for the game. Use them or ignore them at will:
1) In my opinion, trying to analyze the various formulae to see if there are any dead ends (or, in mathematical terms, “local-minima”), is the wrong approach. Such a process is not robust, since any future changes in the design will force a full re-evaluation of the formulae. Given the number of formulae involved, this analysis would be difficult. Instead, you need to add an accumulating failure reward so that the more a player loses the more leverage he accumulates in finally defeating his opponents. For example, you could consider changing the “Power” resource into a “Humiliation” resource. Any time a player fails in a contest, his Humiliation increases. Humiliation points could then be spent to increase the chances of success in future contests as his frustration and desperation increase. If you do something like this, though, you will need to make sure characters get sufficient Success Rewards where they previously got Power for succeeding. Raising attribute values like you have in several of the contests is good in this regard.
2) I see a lot in the contests concerning what evil acts do for the perpetrator. But, I see very little concerning what evil acts do to the perpetrator. You need something like Self-Loathing where the corruption of the perpetrator’s soul aids him in performing horrific acts, but hinders him in other ways. You mention Lovecraft in your text. Perhaps you could alter Clarity to be a little more like the Cthulhu Mythos skill in Call of Cthulhu. That is, the more “Clarity” a person has about the true nature of the universe, the more it affects him detrimentally somehow (such as lowering Sanity in CoC). Or, maybe you could add an Arrogance attribute that aids in contests against “inferiors”, but hinders in contest against “superiors”. The Rage attribute is just screaming out to be conflicted. Rage could help against Victims, but hinder Clarity in other situations somehow.
3) I see no mechanical reason why a player wouldn’t just focus on raising one Aspect and ignore the others. In tactical games, the choice of attributes is forced on a player by the game environment. But, there is no corresponding coersion in AoE. Players simply choose an Aspect to use. If you want to encourage players to use different Aspects in different scenes, you could force them to “recharge” an Aspect before they re-use it. For example, you could tell players to put a check mark next to an Aspect whenever it is used. A checked Aspect cannot be used. When all four Aspects are checked, erase the checkmarks. Of course, that's a rather brute-force approach that may not appeal to you. It's only meant as an example.
4) Somewhere, you mentioned that you wanted Victims to last longer so that more diabolical acts could be perpetrated on them. If you want to do that, then you need to reward players for “cultivating” Victims. For example, you could exponentially increase the value of a Victim according to how many scenes he has existed by making him worth a number of Power points equal to the square of the number of scenes he lives. If he is killed in the first scene, he is worth 1 Power point. If he is killed in the second scene, he is worth 4 points. If killed in the third scene, he is worth 9 points, etc.
I believe this analysis has addressed the second and third of Paul's stated design goals. I need more clarification on the first, though. Which of the outcomes are you trying to make more frequent?
The fourth goal is problematic. You want to interweave the character storylines without making them too vulnerable to each other. Personally, I think you have one too many win conditions in AoE. Players are supposed to 1) kill each other and 2) attain Godhood.
Which is it? You state in "Ending the Game" that the game ends when there is one last surviving character. Does that mean he wins? If so, attaining Godhood is actually superfluous. In such a game, the best strategy is to kill your opponents with the greatest expediency possible. Screw Godhood. Your goal of wanting to make the game about Cosmic power strongly indicates that you want players to strive for Godhood, though. If that is the case, then make the first player to attain Godhood the winner.
You might even want to pushish players for killing each other and/or victims. You could consider killing an act of mercy, since the murdered person would no longer suffer from the evil acts that would otherwise be inflicted on him. If you took this approach, you might just want to eliminate the possibility of killing altogether. The point is to humiliate and denegrate your opponents, not kill them.
I hope this analysis helps. It is entirely possible that I misunderstood or overlooked some things. If so, my analysis could be completely off-base and needs to be adjusted accordingly. Let me know if any of it requires further clarification.
Forge Reference Links:
Topic 16990
Topic 17355
On 12/3/2005 at 11:56pm, Victor Gijsbers wrote:
Re: AOE Analysis
I need to reread and digest this post, but it contains some awesome advice, John!
On 12/4/2005 at 3:42pm, Victor Gijsbers wrote:
RE: Re: AOE Analysis
Comments on the diagrammatic analysis
Great work, but complicated - I think a few errors may have made their way into your analysis. (Yet perhaps I'm wrong.)
I might be wrong, but it seems to me that your representation of the Capacity gauge is wrong. First, in the "Character Makeup", Capcity should be linked to Used Capacity with a solid line leading to a double line (the 'correlated with the maximum' arrow). Then, in the other diagrams, it is always Used Capacity that should be linked to Spent Power, with a solid line.
Then, it is always the # of rolls in the scene made by player 1 that are counted, not those made by player 2. Also, you may wish to complicate things a bit by adding a player 3, the GM, since it is he who rolls the dice, not the players who spend Power on making the roll more difficult.
In resolution against Teachers, it seems a bit misleading to state that there is a third reinforcing loop: once player 2/player 3 has won a single contest, the round is over and the 'number of conflicts already played in this round' gauge is reset to zero. That's more of a balancing loop, as far as design goals are concerned, not a reinforcing one - although I can see where, formally, your claim is coming from. (The same goes for the other places where you make this claim.)
In "Change Teacher to a Rival", I get the impression that there is a single loop that you claim to be both balancing and reinforcing?
In "Change Rival to an Underling", the reinforcing loop you claim exists does not. Think about it carefully, and you'll see that being forced to pay down your used capacity has zero impact on the game. (That is, it is always advantageous to pay down your Used Capacity as soon as you have the Power to do so. So if you are forced to do so, it doesn't disadvantage you in the slightest. Since paying down UC and incurring UC to get more P are instantaneous, it doesn't even affect the play of a competent player at all.)
I do not understand your comments about "Change Nobody to Underling". Neither of the effects you mention are clear to me.
On 12/4/2005 at 3:50pm, Victor Gijsbers wrote:
RE: Re: AOE Analysis
Additionally, in the first to Status Change diagrams, "# Underling 2" should be "# Underlings 1". This perhaps somewhat softens your first point 4), since having underlings does provide a mechanical effect. However, since this number is always subtracted from Resistance, it is more useful to lower Resistance than to try and get many underlings. So I do believe that the reward of a status change from Rival to Underling is pretty low. (Although it is good to have one Underling, to open the possibility of rolling Resolution against Underling, which is the only way to decrease Resistance.)
Could you explain the conventions of arrows emanating from the "conflict" symbol to me? I'm not sure I understand how winning and losing are represented.
On 12/4/2005 at 3:58pm, Victor Gijsbers wrote:
RE: Re: AOE Analysis
And one additional post to state that I basically agree with every point John makes. Some of them have already been uttered by others (I'm thinking of the "use only one aspect" point and the "cultivating victims" point - although the x^2 approach suggested by John is too strong for my tastes - even though it's not, as John claims, exponential :) ). I'm especially sympathetic to the idea of more Conflicted Gauges, the idea of rewards for other players, and the claim about the ambiguity of the win conditions.
On 12/4/2005 at 4:44pm, Paul Czege wrote:
RE: Re: AOE Analysis
John,
Damn. That's an amazing analysis. Thank you. I'm still processing it. So I'm going to start with one question.
I see no mechanical reason why a player wouldn't just focus on raising one Aspect and ignore the others.
The threshold for advancement is the value of the lowest of a character's Aspects. In light of that, does your point still stand?
Paul
On 12/4/2005 at 7:50pm, John Kirk wrote:
RE: Re: AOE Analysis
You can download an updated version of the Gauge Diagrams, with fixes for all of the issues discussed in this post from here.
Victor wrote:
I might be wrong, but it seems to me that your representation of the Capacity gauge is wrong. First, in the "Character Makeup", Capcity should be linked to Used Capacity with a solid line leading to a double line (the 'correlated with the maximum' arrow). Then, in the other diagrams, it is always Used Capacity that should be linked to Spent Power, with a solid line.
After re-reading the rules on Capacity and Used Capacity, I still think the diagrams are correct. Both Power and Capacity can be spent in exactly the same way for the same reasons. But, spending Capacity raises Used Capacity and Power can be spent to lower Used Capacity.
I think the confusion is coming in here because the current way in which I represent resources in Gauge Diagrams radically changed from what is in the book. Bear in mind, a solid arrow no longer represents "increases target as source increases" and a dashed arrow no longer represents "decreases target as source increases". Now, a solid arrow represents a direct relationship ("increases target as source increases and/or decreases target as source decreases") and a dashed arrow represents an inverse releationship ("decreases target as source increases and/or increases target as source decreases"). The and/or ambiguity is resolved by the triangles placed on the arrows. If, while travelling along the arrow from the source to the target, a triangle widens, then the relationship can increase the target. If the triangle narrows while travelling along the arrow, then the relationship decreases the target.
Because of this, a resource relationship can be represented as an inverse releationship (dashed arrow) with an "only increases target" triangle on the arrow. As the resource goes down, the target may go up. But, if the resource then goes up again in the future, it does not cause the target to go down. Does that make sense?
Incidentally, in these diagrams I was a little lazy. Any relationship that does not have any triangles adorning it essentially means it can both increase and decrease the target.
Victor wrote:
Then, it is always the # of rolls in the scene made by player 1 that are counted, not those made by player 2. Also, you may wish to complicate things a bit by adding a player 3, the GM, since it is he who rolls the dice, not the players who spend Power on making the roll more difficult.
I see your confusion, but the diagram is actually correct. In this case the "# of rolls in the scene" is actually a Shared Gauge that is not really associated with any particular player. In such cases, my choice of whether I use a filled circle or empty circle is arbitrary. Generally, I use filled circles in those cases and the subscript merely indicates "good for who". I could have used an empty circle and a subscript of 1, indicating player[sub]1[/sub] wants the value to be low. In either case, the actual relationship between the gauge and the die icon does not change. It is a direct relationship either way.
Victor wrote:
In resolution against Teachers, it seems a bit misleading to state that there is a third reinforcing loop: once player 2/player 3 has won a single contest, the round is over and the 'number of conflicts already played in this round' gauge is reset to zero. That's more of a balancing loop, as far as design goals are concerned, not a reinforcing one - although I can see where, formally, your claim is coming from. (The same goes for the other places where you make this claim.)
You are correct. Those would more properly be termed Balancing Loops. I'll fix that in the diagrams.
Victor wrote:
In "Change Teacher to a Rival", I get the impression that there is a single loop that you claim to be both balancing and reinforcing?
No, actually, that was a typo. The second reinforcing loop is through Capacity, not Clarity. I fixed that in the diagram note.
Victor wrote:
In "Change Rival to an Underling", the reinforcing loop you claim exists does not. Think about it carefully, and you'll see that being forced to pay down your used capacity has zero impact on the game. (That is, it is always advantageous to pay down your Used Capacity as soon as you have the Power to do so. So if you are forced to do so, it doesn't disadvantage you in the slightest. Since paying down UC and incurring UC to get more P are instantaneous, it doesn't even affect the play of a competent player at all.)
Oops. I got too carried away with copy/paste. You are correct. I fixed that in the diagrams. "Change Rival to an Underling" does not contain any loops.
Victor wrote:
Additionally, in the first to Status Change diagrams, "# Underling 2" should be "# Underlings 1". This perhaps somewhat softens your first point 4), since having underlings does provide a mechanical effect.
Doh! You're right. That should be "# Underlings[sub]1[/sub]" and that should be an inverse relationship between it and the die icon. I fixed that in both places. So, you're right. "Change Nobody to an Underling" is not as weak as I originally thought.
Victor wrote:
However, since this number is always subtracted from Resistance, it is more useful to lower Resistance than to try and get many underlings. So I do believe that the reward of a status change from Rival to Underling is pretty low. (Although it is good to have one Underling, to open the possibility of rolling Resolution against Underling, which is the only way to decrease Resistance.)
I'd like some clarification on the rules text here, because I do not interpret the rules the same way Victor does here. The rule states:
Paul wrote:
The GM rolls an opposing pool of d6's equal to the occultist character's Resistance minus the total number of Underlings he has dedicated to this specific rivalry in prior scenes, and reduced by one die for each point of Power the player elects to spend (and also possibly increased or reduced by other players spending Power).
Does this mean that the number of Underlings is subtracted from Resistance or from the opposing pool of d6's? If the former, does this imply that the minimum value of Resistance-Underlings is zero? (That is, it cannot be negative, which would make the question moot.)
Victor wrote:
Could you explain the conventions of arrows emanating from the "conflict" symbol to me? I'm not sure I understand how winning and losing are represented.
I originally meant to explain that, because it is not explained in any of the documents I referenced. My apologies.
I'm experimenting here a bit, because it is sometimes difficult in Gauge Diagrams to tell what is a Success Reward and what is a Failure Reward. In these diagrams, prior to a contest being resolved, the value of the contest is neutral, which can be represented as a value of "0". After the contest, it is positive for one player (+1) and negative for the other (-1).
The arrows leading out, combined with the triangle adornments, can now tell you whether a relationship represents a success or failure reward. For example, in the "Resolution Against Teachers" contest, the contest output to Clarity is a Success Reward because the triangle specifies that the relationship can only increase the target and a direct relationship can increase the target only when the contest goes from neutral to positive. In the "Resolution against Victims" contest, the contest output to Rage is a Failure Reward because the triangle specifies that the relationship can only increase the target and an inverse relationship can increase the target only when the contest goes from neutral to negative.
I hope that clarifies things.
Paul wrote:
I see no mechanical reason why a player wouldn't just focus on raising one Aspect and ignore the others.
The threshold for advancement is the value of the lowest of a character's Aspects. In light of that, does your point still stand?
Yes, I'm afraid so. The root of the problem here is the ambiguity of the win condition. If I can win by killing all other players off, that makes attaining Godhood pointless. If Godhood is pointless, then so is advancement toward it. So, building up my Aspects to attain a minimum threshold is pointless as well.
On 12/4/2005 at 8:41pm, Victor Gijsbers wrote:
RE: Re: AOE Analysis
John wrote:
After re-reading the rules on Capacity and Used Capacity, I still think the diagrams are correct. Both Power and Capacity can be spent in exactly the same way for the same reasons. But, spending Capacity raises Used Capacity and Power can be spent to lower Used Capacity.
This is where formalisation really shines: making ambiguities explicit. I take it we are reading the rules differently, and Paul should tell us which is the right reading (and make this very clear in his text). Here is how I think you read the rules:
Player A has 0 Power, 0 Used Capacity and 3 Capacity. He really needs to have some Power, so he spends 2 Capacity. Now, he has 2 Power, 2 Used Capacity and 1 Capacity. Capacity cannot go below 0.
Here is how I read the rules:
Player A has 0 Power, 0 Used Capacity and 3 Capacity. He really needs to have some Power, so he uses 2 of his Capacity. Now, he has 2 Power, 2 Used Capacity and 3 Capacity. Used Capacity can never exceed Capacity.
So Paul, which of the two is it?
On 12/4/2005 at 11:20pm, Paul Czege wrote:
RE: Re: AOE Analysis
So Paul, which of the two is it?
You have it right:
Player A has 0 Power, 0 Used Capacity and 3 Capacity. He really needs to have some Power, so he uses 2 of his Capacity. Now, he has 2 Power, 2 Used Capacity and 3 Capacity. Used Capacity can never exceed Capacity.
Capacity is like a credit limit. And Used Capacity is the balance you're carrying against that limit.
Although I expect those 2 points of spent Capacity don't actually sit around as 2 points of unused Power for very long. It doesn't make sense to convert your Capacity to Power until the moment you actually need it.
Paul
On 12/4/2005 at 11:43pm, Paul Czege wrote:
RE: Re: AOE Analysis
The root of the problem here is the ambiguity of the win condition. If I can win by killing all other players off, that makes attaining Godhood pointless. If Godhood is pointless, then so is advancement toward it.
I see. I don't think I consider a player who kills off the other players' starting occultist characters to have won the game. The game doesn't end until all starting occultist characters are dead. So this last character will either be killed by the efforts of other players via their replacement characters, or he'll confront Ephactha and either beat him or not. And then the game will end. Or maybe one of the other players, using a replacement character will confront Ephactha before you manage it, and then him or someone kills your character and the game ends. If so, you should hang your head in shame.
You have to beat Ephactha to win (with maybe an honorable mention if you don't beat him, but yours is the only character to ever confront him during the game).
Paul
On 12/5/2005 at 1:12am, Victor Gijsbers wrote:
RE: Re: AOE Analysis
I took the liberty to talk about this analysis in the RPG blog I started today.