Author Topic: Game Pacing  (Read 12201 times)

Nomad

  • Sr. Member
  • ****
  • Posts: 453
Re: Game Pacing
« Reply #15 on: May 21, 2017, 06:43:10 AM »
lol...
Sorry. I thought wealth*crime was what you put in your previous reply, so I thought that was how you were doing it.

I like your idea of making it abstract to reduce people gaming the system.
However, if you are thinking of making it simplified down to:
Wealth = Gear
Crime = Numbers
Danger Level = Enemy toughness

Well, that works. We get this information through surveillance. These values are how we are quantifying the gangs we are surveilling. We look at their gear and determine that we think they are doing pretty well to afford the gear they have.
We see the number of gangers hanging out on the corners and know the kind of numbers we're up against.
We see some tough looking enforcers and decide what danger level the area is.

It doesn't really have to be abstract, seeing as the premise is that we are actively scouting the area out.
Really, if you wanted to make it a little abstract, you could fill in the three details on our first surveillance run with either random or all low values. These represent your initial impressions. Colour them red or perhaps have them say "Maybe" before their values, until you've done enough surveillance to be sure of your impressions.

=)

I like making more work for other people.  =p

Nomad

  • Sr. Member
  • ****
  • Posts: 453
Re: Game Pacing
« Reply #16 on: May 21, 2017, 06:47:13 AM »
If you two insist on equipment tier jumps between wealth (or any other variable), it's also possible to set the medium score (eg. med. wealth) as the one at which the gang by its current shape resides and then based on this variable go up like +1 rank (eg. high wealth) or down like -1 rank (eg. low wealth), or more.
The gang having better gear sounds good. If they're wealthy, then having better gear makes sense. The more money they have, the more they can afford and the better gear they'll want in order to protect that wealth.
However you want to implement it is fine. Start at the bottom and increase equipment based on level and wealth. Start at the middle and go up or down from the baseline.
Six of one, half a dozen of the other.

However, although I was unable to express well, sorry, there is reason why tried abstracted continuous modeling translated to concrete discrete only at the end. By no means wanted to have outside rules forced like this looks. On the contrary, was about what we see in intel to be linked to the gang's AI dealings in the background.
I don't think something this simple needs abstraction.
We are getting the information through surveillance. The values represent what we are seeing and noting.
« Last Edit: May 21, 2017, 06:48:50 AM by Nomad »

ushas

  • Hero Member
  • *****
  • Posts: 661
Re: Game Pacing
« Reply #17 on: May 21, 2017, 08:55:47 PM »
Uff. Let's forget I said word abstraction. Just made us talk apples and oranges (budget & AI here), sorry.

Haha, so I first thought you omitted crime, but then took it seriously and realized that wealth is the Resource after all, and went making different budgets to accommodate that local greediness:)

Gangs have various priorities, including but not only protect their income in the form of wealth*crime. So would make sense to observe this via correlation wealth <-> gear or so. The trick is to make it clear enough without a need doing statistics, but not straightforward. So you can slightly nudge gang AI into one way or the other by weights, but this doesn't mean always forcing them to use +1 equipment tier for troops on high wealth*crime tile and such...


Ok, will try muse over equipment differently -- from the AI perspective:
The point is one can express (when programming) how things are relative costly by something like Gang Units (GU, previously troop units). You can define GU for each troop category, eg. mafia rank 1 = 1 GU, sniper = 2.5 GU, underboss = 5 GU... let it vary across ranks and gangs so it feels approx. ok (not exact).

Globally, there is a decision to make how much you want to invest into equipment. Perhaps initially you you don't need to spend more than some minimal amount, and just add enough to accommodate inflation of troops. If player doesn't bother you, may as well save money for facilities even if doing well. But if he starts to be a threat (eg. easy last 3 battles), time to react by investing more money, etc.

M = sum of all Gang Units across all troops. So money value per 1 GU is
MeanValuePerGU = Money-in-gear-property  /  M.           (if you invested more money -> more valuable gear)

Gear-property can be variously distributed though. For example, locally putting higher or lower weights on tiles, but the total amount has to be conserved. If a tile's troop total is N GU and weight W (eg. W=1.4 on high wealth or W=0.8 on low), then
summing [W x MeanValuePerGU x N] over all tiles = MeanValuePerGU x M = Money-in-gear-property.
The most challenging is to first well define what is important for AI at various stages. Inferring then the weights placements and balancing is just an exercise.

So by weight adjusted local money value per 1 GU is   
TileMeanValuePerGU = W x MeanValuePerGU.      (if you care more/less about this -> more/less valuable gear than mean)

Till now, what's inside gear-property hasn't been defined yet. It will only need to be translated into concrete state for a group of gangsters at the time of a mission (ie. observer's effect). All one needs are tables (array) for each troop category (3 x boss, all underbosses, all specialists, 3 x tier 1) to map  values of TileMeanValuePerGU  into equipment tiers. Naturally, sniper has different and more expensive equipment than rookie mafia guy (for the same TileMeanValuePerGU).

When the gang gets access to new stuff, tables can have a second column with an alternative gear to read from. The same if it's a special encounter (eg. all guys having knives or katanas) or a local gear deviation (facility?) - just add a column or condition (eg. exchange main weapon)...

All that is easy to do because, apart from making it look like progression and using common sense, you don't need to care about the cost of gear for gangs at all nor doing any purchases. So yeah, looks like a lot of tables but easy to balance and add special complexity (making tables vs. programming). One may as well have all equipment tables to be read from external xml file, to edit and test on the fly...

Btw. if we defeat and loot an enemy, those items will have specific value balanced from the player's pov, but it's the  [TileMeanValuePerGU x troop's GU] what is substracted from the gang's Money-in-gear-property, and his GU from M and N.
« Last Edit: May 21, 2017, 10:49:17 PM by ushas »

Nomad

  • Sr. Member
  • ****
  • Posts: 453
Re: Game Pacing
« Reply #18 on: May 22, 2017, 06:41:08 AM »
It appears we have differing views on the subject.
I do see where you are coming from. You're looking at it from a business perspective. Only spend what you need and then spend reactively.

I'm looking at it from an observational perspective.
The player is observing the gang in a particular area and evaluating their threat level. Then the player writes down three values. These are observational values.
Why can't they just apply?

Eg: As it IS surveillance, we could just change the values to:
Danger Level          Same as current Danger Level - Proposed to be Enemy toughness/skill
Strength                 Was called Crime - Proposed to be numbers of enemies
How well equipped    Was called Wealth - Proposed to be Gear level

Could just call the last one Gear, as an evaluation of their gear.
Currently it is called Wealth because that's the value to determine how much money the gang makes in that area... however... if the values are related then it doesn't matter. If their Wealth determines their gear, then we can just look at their Wealth and know how well they're going to be geared up (because the more money they have, the more they spend on their gear). The player just thinks he's evaluating their gear. Then when more money drops... well... that makes sense. They must have more money if they can afford this stuff.

The Gear value that the player sees is taken from the current Wealth (we just rename it on the Surveillance screen). It doesn't change. That value is still used by the gangs to determine their earnings.
programatically, nothing changes. If it was Crime x Wealth before, it still is. It's just that now the player sees Crime as Numbers and Wealth as Gear.
Observational.

Even if there's a heap of additional programming under the hood to determine how well the gangs are doing, to the player if could just come down to:
How tough are the baddies?
Roughly how many would I be facing if I went in?
How well equipped are they?

Daithi

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 1250
Re: Game Pacing
« Reply #19 on: May 22, 2017, 01:56:54 PM »
@ Nomad - The problem with linking the equipment entirely to crime rate & wealth is that the highest levels of equipment will be available to them very quickly. A small modifier based on wealth will work.

@ Ushas - The hope is to have different gangs emphasise different priorities - the survivalists want better gear, the churchers will focus on economy, with the mafia in the middle. More ground work (including rackets, permanent upgrades, re visiting basic Gang AI code (a basic implementation, which was written more than a year ago *shudder*), is needed though, before this can be realised).

Can't be certain that the idea of setting numbers, buffs, and equipment will work the way that has been discussed. It seems to make good enough sense, but a lot of things need to be implemented and thought about before any changes are made. Unlikely anything major will happen on the balance front until V19.
« Last Edit: May 22, 2017, 02:27:42 PM by Daithi »

Nomad

  • Sr. Member
  • ****
  • Posts: 453
Re: Game Pacing
« Reply #20 on: May 22, 2017, 03:58:24 PM »
Well... I think you should go with whatever you've already got (assuming you've already got something going, or in mind).

I'll be happy to see the enemies sporting different gear.
There's heaps of melee weapons for them to choose from. There's a reasonable number of guns. It will make the fights a little different for them to have a variety.

I'm happy to see what you've got in mind. I'm sure it will work just fine.

And as far as them getting gear too early...
My perception was that when I do surveillance, there are areas with low wealth and areas with high wealth. If I know the high wealth areas are going to have bigger guns/knives, then I just won't do them early. I'll do all the low wealth areas and clean up the dregs before I move up to medium.
It doesn't have to be wealth x crime, but if it is, then I'm sure it will be fine.

ushas

  • Hero Member
  • *****
  • Posts: 661
Re: Game Pacing
« Reply #21 on: May 23, 2017, 01:51:15 AM »
IIRC gangs are supposed to grow in time - in numbers, leveling up, building facilities, upgrading gear, training troops... If the gear itself would go as you say, many players will do exactly opposite than you.

@Daithi Yeah, on the base level though all gangs want the same, priorities can be re-balanced later, no? That's why was having fun over equipment, no problem if the implementation doesn't fit, its principle is general - the gang AI being robust as bare while it's easy to add nuances (factions, global progression/upgrades, local space and time variations). You have much more going on though, of course. Except, you guys still want to tie the AI's hands for gear as opposite to nudging. >:(      (btw. why is mafia always in the middle?)

I don't think anybody expects that anything will happen as imagined, on the contrary it a kind of frees...
(in case of trying to shoo the discussion, too mild tbh, hint: redirect)

I do see where you are coming from. You're looking at it from a business perspective. Only spend what you need and then spend reactively.
Not necessarily. Flexible system will let you do such if desired. Doesn't need to work like that though, probably won't.
Quote
The player is observing the gang in a particular area and evaluating their threat level. Then the player writes down three values. These are observational values.
Why can't they just apply?
Empirical... Mars does not live by epicycles, his mistress is the gravity.
And that's the beauty isn't is? When empirical laws do apply (under certain conditions), you can go by those laws and fight the enemy, and yet observing just effects of a bigger force...

Btw. a good summary from the player's perspective. Intel we're getting is a mixed bag indeed.
Let's categorize useful things to know:
  • Attributes of a tile (basic resources, doesn't change)
    • Gang's presence
    • Wealth
  • Gang's long term and short term activities in the area (strategical part) 
    • Crime rate (% of tile's wealth collected, not sure if its grow is set in stone or can vary)
    • Number of tile occupants (goes with crime, unsure how exactly, upper limit for # of enemies for a battle?)
    • Underboss / Leader (not sure how/if benefit, but their placement (hunting) is a part of gang's (player's) strategy)
    • Special Encounters (??)
    • Facility (gives local/global benefits to gangs)
  • Tactical: Opposition to be faced in battle (varies, based on gang's strength, indirectly says how much gang values the tile at the moment) 
    • Number of enemies / Guard level (either exact or gives good sense how much to expect)
    • Average level of enemies (it's tied to global leveling, not sure if useful to know)
    • Danger level (how strong/buffed relative to gang's base troops)
    • Equipment level (how well equipped, either relative to gang's current average, or exact in tiers)
    • Mission objective
    • Leader (morale mechanic, special enemy)
(things are uncertain, not sure how will correlate, took the liberty, feel free to correct)

Nomad's proposal to disentangle gear level and number of enemies is interesting, if possible (especially if helps to liberate AI:)). It makes sense next to the Danger. But I would keep Wealth and Crime too, those are good to know strategically.

The question of absolute vs. relative: Globally the gang is in the state reflected on all tiles. One needs to know is how it differs between them. One can use Guard/Danger/Gear level, with scores relative to gang's base. So Low Guard means 5 Survivalists but 7 Churchers, etc. You don't see absolute numbers, but if you pay enough attention will get the grip.

Alternative surveillance list v3 (ordered by approx intel % needed)
Gang
Guard level
Danger level
Equipment level

Wealth
Crime rate
Leader / Special. Encounter
Facility

Edit: correction rackets-> spec. encounters; re-ordered list to put tactical intel first - just something to consider.
« Last Edit: May 25, 2017, 10:28:19 PM by ushas »

Daithi

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 1250
Re: Game Pacing
« Reply #22 on: May 23, 2017, 09:47:21 AM »
@ Ushas: I think the most likely outcome is that on the wealthiest tiles, the gangs will get +1 to equipment level. In the scale of 10 levels, this is pretty small. Outside this, the gang will have to purchase permanent upgrades and build rackets (some provide bonuses to local tiles, others global bonuses) to increase the quality of their gear. Each gang's AI will make the choices on what goals to pursue.

Not trying to shoo discussion - you're welcome to talk about anything. It doesn't take long to follow the discussion and a single perspective is much too narrow to design a game of this scale, so ideas are very useful for improving decision. Trying to avoid giving the impression that anything's going to happen very soon in this area, or that anything weeks or months away from being worked on is fixed.

Hadn't thought about showing equipment level, but can see certain advantages to it, particularly in terms of detecting rackets which boost equipment level in a local area.
« Last Edit: May 23, 2017, 10:07:07 AM by Daithi »

ushas

  • Hero Member
  • *****
  • Posts: 661
Re: Game Pacing
« Reply #23 on: May 23, 2017, 10:44:10 PM »
Albeit what you said about equipment would be possible with flexible AI too, I understand, will have to declare defeat.  :'(
For the better, of course, the most important is that you work towards your vision the way feels the best.

Yeah, it's good to moderate our expectations, in case translated wrong (ie. over the internet), wouldn't mind to be redirected from reasons you mentioned. On the other hand, knowing you're sticking to your schedule makes one less worry being out of line...

Hadn't thought about showing equipment level, but can see certain advantages to it, particularly in terms of detecting rackets which boost equipment level in a local area.
I smell detective approach!

To recap, from the pacing perspective it's good if tactical buffs in numbers/toughness/gear here & there vary in space & time, so players experience different setups. As secondary, it would be neat if contains deeper campaign meaning as well. You have ideas, fingers crossed.

What we seek - Leaders/bosses, (spec. encounters), facilities and high income tiles? So what about a system where each of those 3-4 categories boosts at least one observable tactical parameter on a tile(s), differently if possible, each leaving its own imprint? Perhaps it's handled individually for each gang (reflecting their priorities).

If done well and players will be able to detect or suspect via observing uneven guard/danger/equipment levels or so, no need to tell explicitly, like at all. It would be satisfying finding out the correlations and how utilize in hunting on one's own...
Well, I guess complicates balancing though...

Edit: correction rackets-> spec. encounters
« Last Edit: May 25, 2017, 10:22:12 PM by ushas »

ushas

  • Hero Member
  • *****
  • Posts: 661
Re: Game Pacing
« Reply #24 on: May 25, 2017, 10:19:29 PM »
I've watched the video (a nice map btw.), did you just said rackets=facilities? :o

I'm an idiot. All the time thinking rackets are special encounters (like the bootleggers and such). That aside facilities the gang AI is also deciding when/where to trigger those and some of them are supposed to give temporary local/global bonuses too, so you can detect... while you were talking the facilities... Well, better late than never, I guess.

Will edit some of the posts, as many things make no sense anymore. Apologies for confusion. Please don't hesitate to correct me when factually wrong.

Daithi

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 1250
Re: Game Pacing
« Reply #25 on: June 16, 2017, 01:34:21 PM »
It's all good. Will correct if notice, but didn't in this case!

Am testing the gang AI placing rackets and buying upgrades now. It's likely to be pretty unbalanced at first, but scaling of enemy threat level won't be restricted to level now, they will also have better equipment and buffs.