General > Discussion

Feedback and impressions after 1 campaign

<< < (3/7) > >>

H-Hour:
On the rock-paper-scissors approach, you can see a proposal for balancing the Damagetypes. I hope to start working towards this in the next development cycle, but we will need more diverse aliens before we can fully implement this.

Jon_dArc:

--- Quote from: Hertzila on January 27, 2012, 08:46:12 pm ---It's the same for bolter rifle (rail gun, another pure terrestial weapon). Its basic idea is that of a crossed medium range sniper and assault rifle but it lacks the crucial accuracy of snipers, so it usually lacks users.
--- End quote ---
The ability to deliver substantial damage through walls makes it useful, but yeah, the sniper/assault role implied by the fluff doesn't pan out.

~J

ShipIt:

--- Quote from: Bashar on January 24, 2012, 11:57:53 pm ---Hello.  I should admit right up front that I haven't yet played UFO:AI for myself. 

--- End quote ---

I stopped reading at that point. Sorry, but how can you rate a book without having read it ?

Bashar:
Thank you for replying.  Hertzila, all your responses sound great.  Glad to see that the terrestrial side of research is being represented and that each tech class of weapon has its own unique characteristics.  Can't really call them tiers anymore.

One point on which I wanted to clarify myself.  When I talked about research into refitting alien weapons to what which could be used by humans, I was suggesting it as a separate line of research from the study of the weapon's technology.  Let's say we come off the battlefield with a nice, not-quite-so-shiny alien plasma gun.  The player would have a choice between researching the theory of how plasma weapons work or instead learning how to refit this specific type of gun to human physiology.  The difference is that the former, while costlier, would open up more avenues of applied research and the opportunity to manufacture the weapon, while the latter is an inexpensive means of fielding new technology, though narrow in its scope and limited in available to the quantity of alien materials captured.

This goes in tandem with what I was saying about upgrading equipment, though this in some way shape or form may already be implemented.  Let's take for example a blaster.  There might be four different types of plasma blasters in the game as follows:

Alien Blaster: The weapon that the aliens use against the humans.  Effective in the hands of aliens but either useless or incurs substantial penalties when wielded by a human.
Refitted Alien Blaster: This is the alien blaster after researchers have learned how refit it for human physiology.  Limited in availability by the number of alien blasters captured and clunky to handle (manifested by increased time cost and reduced accuracy) but this is a relatively inexpensive means of getting high tech equipment onto the field quickly.
Human Blaster: Human manufactured weapon only available once the player has invested costly research into the theory of the weapon.  Being a young technology, it might not be as effective as the alien blaster in the hands of an alien (reduced firepower) but it's more effective than the refitted alien blaster because it was specifically designed for human physiology (normal time cost and accuracy).
Human Blaster mk.2: If the player so chooses, he can continue research into blaster technology and develop an even more advanced version that can match, or perhaps surpass, the alien technology.  The ratio of research cost to the increase in combat effectiveness might be high, making it an uneconomical choice except for those players who have a reason to to specialize in this particular technology.

So you have a single weapon but four different models depending on how the player chooses to progress on the tech tree.  This is what I meant by lateral research.  Depending on how the player values a particular weapon, they may choose research it differently.  It also presents another opportunity for the player to react to variations in the AI's activity (what we were talking about before in terms of making the game more dynamic).  If the AI is throwing a lot of one type of weapon at the player then the player might decide that it would be better in the long run to invest research only in refitting that weapon and save on research points at the expense of limiting the weapon's effectiveness.  By contrast if the player recognizes that a particular weapon is dearly acquired, or if they plan on investing heavily in that particular technology, they might forego researching how to refit the weapon altogether and invest in the costlier but more pertinent theory technology.  Or the player might want to research both, refitting to get the weapon on the field quickly and then theory to get access to all the additional applications; however, this redundant investment would come at the opportunity cost of researching other technologies.

So as opposed to each technology being "fund and forget", this would introduce an additional level of complexity where the question posed to a player isn't just a matter of whether or not to invest in a particular technology but to what degree and whether the player is focusing on short or long-term application.  Whether that's too much complexity is another question but I wanted to suggest this as a means of enhancing the strategic part of the game.  I think it would also address the "intermediate techs" that Samuel was talking about in another thread without having to introduce new equipment but rather different versions of existing equipment.  I completely agree with the philosophy that fewer, diverse types of equipment are preferable to many, similar equipment types.

---

I looked at the stats proposal, H-Hour.  I'm glad you brought that up because I had some thoughts I wanted to share on that as well. <g>  I'll add that later to break up my post into smaller(ish) chunks.  Is there a dedicated thread to which I should write about stats or would here be fine?  I did a search but didn't find a recent thread and was reluctant to bump something that was out-dated.

Also, I saw that high level of alien activity has been reduced so I plan to download the latest version and look forward to trying it later this week.  Not sure that'll reduce the quantity of things I want to say, but I'm sure it'll improve the quality. <s>  Thanks again for the opportunity.

Bashar:
2A) Generic Super-Stats

Speed and Accuracy, while they function differently, both result in improving an agent's overall effectiveness.  Speed increases the time units available while Accuracy increases the chance of scoring a successful hit, both of which directly and universally improve the.  As they stand, these stats don't really contribute to an agents unique identity except as an indicator of longevity or innate ability.  In other words there's no motivation for the player to groom an agent as "the speedy guy" or "the accurate guy" because these are things that the agent will develop naturally just by participating.  To put it another way, Speed and Accuracy don't contribute to specialization whereas other stats are defined in their scope my a limited but clear set of parameters.

I would suggest limiting the scope of these stats so that agents will value them differently based on their designated role.  Take Speed for instance.  Instead of granting the agent additional Time Units, perhaps instead use it to reduce the TU cost of movement, either running or crawling.  This would define Speed as a stat useful for front line agents who are responsible for scouting and weaving between cover but be less attractive to an agent in his sniper nest or a machine gunner providing suppression fire.  Thus rather than being a general skill that is desired and attained by all agents, the player can focus agents on becoming "the speedy guy" and have that contribute to the agent's specific role in the team.

I haven't looked at the code yet so I am basing my information on what I've read on on the forum but Accuracy seems to be mainly a measure of overall effectiveness.  Whatever weapon the agent is wielding, however the agent is wielding it, accuracy imparts a linear bonus.  As a stat it doesn't provide any functionality that isn't already provided by specific weapon skills.  As such it's redundant except as a measure of "cross training" between weapons.  There are more direct means of doing that, such as giving individual weapon skills a floor increase at the end of a battle, based on a percentage of the increase that was accrued by the skill that incurred the greatest improvement.

What I would suggest is changing Accuracy so that it impacts the player based on the fire mode of their attack.  Automatic fire would receive marginal or no benefit from a high Accuracy, burst fire, snap shot, and aimed shot would receive a respectively progressive benefit, and head shots (which should probably be renamed something like "critical" since there's no guarantee that alien physiology will model terrestrial life) would receive maximum benefit of a high Accuracy score.  This would allow Accuracy to supplant the proposed Sniper skill while being versatile enough to apply, to varying degree, to most weapons and still provide a unique benefit that will allow agents to specialize in certain roles.

The stat would be less a matter of overall effectiveness as versatility.  For example an agent with low Accuracy would be just as effective as a trained sniper if all they're interested in is laying down firepower; however, if a sniper wants to take a deliberate shot, even with a short-ranged weapon such a pistol, he'll see some degree improvement in his chance to hit over his trigger-happy team mate.  Should an inaccurate agent choose to pick up a rifle, the fact that he was a low Accuracy score would mean that while he's effective at burst fire, he won't see as much of an improvement if he used that rifle to take an aimed shot.  Thus this proposal of implementing Accuracy would help define an agent's combat role rather than act as a generic improvement skill.

The algorithm could look something like:

--- Code: ---(base_accuracy + fire_mode_bonus * agent_accuracy) * agent_weapon_skill
--- End code ---

This is not to say that there shouldn't be a stat that improves an agent's overall combat effectiveness, which is what Speed and Accuracy currently do, but I would suggest that such an effect be consolidated into a single stat, possibly called "Experience" or "Discipline" if you wanted a label with less of an RPG connotation.  Have it based the number of missions the agent has participated in and the quantity of aliens encountered and felled in those missions.  Basically, make it as simple as possible but no simpler.

This would also seem to correspond a little bit with the proposed Mind stat, since it was states that this would correspond with the agent's level of experience.  In my opinion, and I'm just throwing this out there for your consideration, but I would have two separate stats: Discipline and Psionics.  In terms of defending against psionic attacks (mind control, hallucinations, panic attacks, etc) the game could pick the higher of the two (perhaps unevenly favouring Psionics) to determine resistance while normal shocks to morale would be handled solely by Discipline.  For instance, a psionic warrior might prove to be a pansy when bleeding from a wound while a battle hardened veteran might be able to withstand the effects of a pisonic attack even though he has no such gifts himself.

2B) Encumbrance

I admit to personal bias here but for my part I generally don't like encumbrance rules.  It's not that I don't like the concept but often the implementation results in tedium.  I spend enough time managing inventory for my Skyrim avatar, the thought of doing the same for eight Phalanx agents when they go in to battle makes me wary.  The proposal page was kind of non-specific in terms of how encumbrance would be implemented so I'm assuming you were just leaving this for another date.

From my perspective, the most straightforward approach would be to add more inventory space based on the proposed Strength stat.  Of course, this is more a measure of volume than weight; however, it has the benefit of being very easy to telegraph to the player and provides a demonstrable benefit without increasing complexity.  You could then base movement modifiers on the space left remaining in the agent's backpack.  Leave so many tiles open and you get full use of your Time Units at the start of the turn, for each space filled beyond that the amount of Time Units diminishes.

What I have trouble wrapping my head around is how to improve strength over the course of the game, particularly how rationalize strength increases from combat actions.  Normally a person would improve their strength with exercise and training, lifting weights and such.  The only way I can think of reflecting that in the combat view would be to improve strength based on the agent's level of encumbrance through the course of the battle.  Each turn the agent suffers an encumbrance penalty, the more points are contributed to increasing his Strength stat at the close of battle.

2C) Experience progression

I also want to talk about how stats are improved through the experience system.  At present most stats are dependent on the frequency of hitting a target.  I'm not sure this is the way to go about it for a few reasons.  First, this method creates a positive feedback loop where in hitting a target increases an agent's skill, which in turn makes them more likely to hit their target, where upon they'll increase in skill even faster, and so on and so forth.  This means that agents would progress slowly at the start of their career but as they participate in battles and improve their stats, they progressively become more proficient and accrue stats are a faster rate.  This would seem to be contrary to what I've interpreted as the developers intentions.

The other issue with tying stat increases to hits on a target is that it unfairly awards agents experience based on their weapon choice.  Let's assume that we have a weapon that has an 80% accuracy rating and does 25 points of damage.  Compare that to a weapon that does 50 points of damage but only has an accuracy of 40%.  Statistically these weapons are equally effective, on average they will both output the same degree of firepower.  But because the experience system is tied to hits, the agent using the former weapon will improve his pertinent stats at twice the rate as the agent wielding the latter weapon.

And finally, the system unfairly awards experienced based on the Time Unit cost to fire it.  If you have two rifles of equal accuracy but the first requires fewer time units to fire a round than the second, then the agent wielding the former weapon will progress his rifle skill faster than the agent wielding the latter weapon.  This is because both agents accrue the same experience award regardless of how much time they are spending.  The agent with the faster weapon will progress faster than the agent with the slower weapon.

This leads me to what I think would be a good solution for these issues.  Take accuracy out of the picture altogether and instead base experience instead on the expenditure of Time Units.  After all, a person learns just as much from their mistakes as their successes.  It also gives the developers much finer control in shaping the experience system as well as a clearer picture of its impact.  If experience were based on time units then you could look at the end of battle statistics and divide the numbers by the duration of the battle in turns to see the average time expenditure an agent used a particular skill.  Divide that by the agent's TU pool and you get a very close approximation of what percentage an agent used his skills in relation to one another.

You mentioned balance before and I appreciate how important that is.  I think basing experience on the expenditure of Time Units would would make skills on a whole easier to balance and tweak.  It would eliminate the bias toward fast, accurate weapons and mitigate some of the guesswork in determining the "value" of individual skills.

2D) Stat Proposal

Here I've included for your perusal an alternative list of skills to that posted on the proposal page, based on what I've talked about in this post.  Note that I'd include stun-grenades with explosives since it seems a more appropriate category.  If you want a non-lethal weapon for the pistol/specialist skill then I'd suggest implementing a tazer or shock prod since they would seem to have more functional relevance to the skill in general.  Also I'd separate shot guns, giving "sawed off" variety to pistol/specialist and including long barrel or automatic shot guns with the rifles skill.  As for the flame thrower, it just seems more appropriate to include with Strength since I imagine that fuel tank to be pretty heavy and space consuming.

Discipline: Overall determination of an agent's combat experience and training.  Has a strong impact on morale but might also improve the agent's pool of Time Units, hit points, and combat effectiveness.
Speed: Determined the efficiency of movement actions.  Having a high Speed rating means an agent can cover more tiles with the same number of Time Units.
Accuracy: Determines how much of an accuracy bonus the agent receives from choosing a deliberate/slow firemode.  Minimal impact on spray-and-pray automatic fire, major impact on time consuming aimed shots.
Specialist: Weapon skill for single grip firearms and melee weapons as well as a catch-all for non-combat packages like medkits and motion sensors.
Explosives: Weapon skill for grenades, grenade launchers, and -maybe- RPGs if only to fill it out.  Includes mines and demolitions, if implemented, just for the sake of consolidation. 
Rifles: Weapon skill for double-grip firearms including both carbines and sniper rifles as well as long barrel shotguns.
Strength: Dual-purpose skill.  Determines encumbrance rules determining how many inventory slots an agent has available, and how many of those slots may be consumed before they start to earn Time Unit penalties.  Also a weapon skill to determine the effectiveness of large/heavy weapons such as the rocket launcher, flame thrower, and anything that ought to be mounted, like Rambo's M60 or Terminator's minigun.
Psionics: Mind control attacks and defense, though defense may be supplemented or over-ridden by a high Discipline stat.

Navigation

[0] Message Index

[#] Next page

[*] Previous page