Jump to content
  • Advertisement

Archived

This topic is now archived and is closed to further replies.

Wavinator

Feedback on skill system?

This topic is 5179 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

I've stolen the combat calculation method of Civilization and turned it into a skill system. I'd like to hear what you think. As it goes with these things, it's kind of long, but I've formatted it for easy reading. The Basics Skills are a linear range with a start and end number dependant on what is using the skill (human, alien or machine). Human skills top out at 1000. Machine skills top out at about 10x. As with most skill systems, skills are based on stats, so if a character doesn't have a skill, the stat is used instead. In any contest, two numbers representing either opposing skills or one skill versus a challenge rating are added. The percentage of the total each occupies results in percentage chance for success or failure. Example: Ensign Jones is trying to navigate a wormhole. His Navigation skill is 100. The wormhole's difficulty is 100. The total is 200. 100 / (100 + 100) = 50%. Jones has a 50% chance of success. Now if he navigates another wormhole, with a 300 challenge rating, the calculation is 100 / (100 + 300) = 25% Basically, skills are either opposed or a challenge rating is randomly generated in the game world. (Challenge ratings and enemy difficulties go UP the farther out you explore, btw.) Skill Slots and Skill Decay The player and NPCs may only have a number of skills equal to their Intellect / 5. Since the range for humans is 10 - 100, this means between 2 to 20 skills. Other races have different limits, except the playable Rogue AI class, which has no limit. Players can choose to "deactivate skills" and learn new ones in their place. The deactivated skills decay by 1 point per day. Example: Swapping out Navigation 500 would result it being lost in 500 days (almost 2 game years). For balance, it's more difficult / expensive to raise skills the higher they are, so players who swap out skills constantly will be paying a hefty price. This enforces the limits of the Intellect skill and encourages players to enlarge their pool of hirelings. Skill Tests & Complexity Skill tests have multiple trials. So Jones has to make not just 1 successful Navigation check to transit a wormhole, he may have to make a dozen or more, depending on the complexity of the challenge (which is always displayed, by a scan in this case). To be successful, a character must win a number of tests based on their level of experience (see below). So in a Navigation 100 vs. wormhole 100 test, Jones may have to win the majority of 10 trials. Experience Level The number of trials needed to win is based on experience level: Green, Regular, Veteran and Elite.
  • Green must win 75% of all trials.
  • Regulars must win 50% of all trials.
  • Veterans must win 33% of all trials.
  • Elites only have to win at least 1 trial.
Skill Test Chart The skill test is compared to a chart of percentages and results for each percentage. The challenger, attacker or one who attempts a skill test occupies the upper part of the chart, the defender the lower part. Example: In a Nav 100 vs. wormhole 100 test, Jones occupies 100% - 51% of the skill test chart. The challenge occupies 1 - 50%. Example 2: In a Nav 100 vs. wormhole 300 test, Jones occupies 100 - 76%. The challenge occupies 1 - 75%. Degrees of Success or Failure For each test, a second number is generated to determine the degree of success or failure. The lowest number out of all tests is used. Here is the breakdown:
  • 1 - 5% Critical Success (3x damage done/resource gain)
  • 6 - 20% Great Success (1.5x damage done/resource gain)
  • 21 - 80% Standard Success(Full damage done/resource gain)
  • 81 - 95% Mild Success (1/2 damage done/resource gain)
  • 96 - 100% Neglible Success (1/4 damage done/resource gain)
(Replace "success" with "failure" and "resource gain" with "resource loss" to get the failure chart) Example: Let's say Jones wins 2 out of 3 trials on Nav 100 vs. wormhole 100. The first degree is 99%, which is a neglible success. The second degreee is 5%, which is a critical success. The best result is used, and Jones gets a critical success based on the Wormhole Navigation Success Chart. (Had it been a failure, the worst result would have been used-- it's only fair! ) Each test has it's own customized chart. In this case, Jones, who is an Elite, pases throught the wormhole with no damage, no fuel loss and in record time, which is a Critical Success as wormholes go. Skill Improvement Skills can be improved by use; assigning characters to training facilities or having your character use them; by paying to use facilities at colonies and space stations; or by tutoring. Use: Every critical success raises a skill by 1 point. Winning all skill tests in a contest which is equal to or above a character's skill gains points equal to the number of tests. The tradeoff here is luck (real and in-game as a character stat) and time. Training facilities: Various pods, equipment and facilities raise skills based on their quality. The tradeoff here is expense of investing in facilities, occupied crew, the amount of energy and space they consume aboard your ship / station, and wear and tear with use (requiring time spent in maintence and possibly even supplies if you let maintence drop too low). Paying: Neural RNA therapy and cortical interfaces can raise skills instantly. This is expensive, however. Every 20 points raised costs an additional 15 credits over the base price of 100. So going from 0 to 1 costs 100, 1 to 2 another 100. At 21, the price goes up to 115. Examples: Going from 1 to 100 costs 13k. 250 to 500 costs 93,400. Going from 900 to 901 costs 1535. Expensive, but not bad when you consider its a college education in minutes. Tutoring : Characters can train other characters up to their skill level. The number of points per hour is based on the Intellect/25 of the character being taught. The tutor must make a successful Instruct skill test versus the level of the skill to be taught. This means that it's easier to teach someone a skill if they know a little bit than if they know nothing at all. Tradeoff here is time and idled crew. Skill Specialization Every skill has a subskill. For every 50 points invested in a skill, players may choose for their own character or crew a specialization in a subskill. The options are: Specialize In Subskill : Add +5 each time taken to skill tests involving this specific category. Example: Navigation Specializations
  • Piloting (small ships)
  • Helm (big ships)
  • Launch / Land (fuel use and possible damage)
  • Gravigation (wormholes and planetary flybys)
  • Stellar Cartography (misjumps, getting lost)
  • Manuever (docking, evasion and combat)
  • Efficiency (engine wear & tear and fuel use)
Failsafe: Retry any failed roll for this skill test Precautionary Measures: Seasoned characters can soften failure. The entire success / fail table for this skill is improved by 1%. If taken 5 times, players can eliminate catastrophic / critical failure in the skill category by adjusting the entire table downwards. Example: Taking Navigation (Failsafe) five times yeilds no critical failures!
quote:
  • ------ ==== NO CRITICAL FAILURES POSSIBLE ===
  • 1 - 15% Great Failure (1.5x damage taken or resource loss)
  • 16 - 75% Standard Failure (Full damage taken/resource loss)
  • 76 - 90% Mild Failure (1/2 damage taken/resource loss)
  • 91 - 100% Neglible Failure (1/4 damage taken/resource loss)

Skill Combination The skills are linear partly to allow NPCs and the player to combine skills. Skills can be combined either for a single task, such as creating a special project, or to operate a piece of equipment. Skills are combined with the designated lead character getting 100% of his/her skill rating, and all other characters adding 50% of their rating. For projects and tasks, the maximum number is related to the Leadership skill level of the leader you designate (else NPCs nominate the highest ranking character, whether most skilled or not). For workstations and panels, the maximum skill is the number of characters that can stand directly next to it or network with it if you have Networking Subprocessors installed in your ship / base's main Computer Core. For equipment, it is a fixed number depending on what the equipment is. Example: Let's say you assign Ensign Jones to work with Ensign Rho on a tough upcoming jump. Their skills are 100 and 100 respectively. By combining skills, they total 150. Now the contest is 150 vs. 100, which improves your odds a bit. The exception to this rule is a game mechanic called Synergy. Two characters with similar personalities, likings and traits can produce Synergy whenever they work together. This allows them to blend their skills together at full strength. If Jones and Rho were in Synergy, their rating would be 200; others could still add their talent, but would only do so at 1/2. Skill combination takes time on the order of 1 game hour per every 50 points to be combined for projects and 1 second per 50 points for real-time challenges. For projects, the tradeoff is that the characters must be willing to work together, and be together in a meeting area. Appropriate meeting facilities (chairs, table) builds up the "work" faster. You also might want to have replacement crew to take up the slack while the huddled characters brainstorm. Equipment Requirements and Skill Caps You're only as good as what you have to work with. Carmack would have trouble programming Quake using an ENIAC, but could probably work miracles with a Cray X1. Most objects have skill requirements to successfully use. This is done to encourage skill improvement, as better items have higher requirements, and to provide challenges requiring multiple teams (such as a mysterious alien artifact). Workstations for such functions as Helm, Security, or Engineering also have skill caps representing the quality of the workstation. This is done to help encourage interior ship / base upgrades and create a tradeoff in investing in a system with a cheap price but more frequent replacement needs versus one with all the bells and whistles that crew can grow into. For simplicity, the station limit is simply a skill number. So a Nav workstation of requirement 15 / limit 250 will only be good so long as a character is between 15 and 250-- then they will outgrow it. (They could still use it, but it would only confer 250 points to all Navigation checks). Versions Some skills have versions which effectively duplicate the entire skill. The best example is the skill Language, which has an Anglex or Galactic version, both of which occupy a skill slot each. Other versions apply to the equipment of the different races, so that some skills have a Species version as well. So Jones may be an ace Navigator of Terran ships, but is not automatically qualified to pilot Kovaunn or Zelenae ships. Some skills span multiple versions for the sake of other skills. Language, for example, allows characters to use alien workstations and some equipment up to 1/2 their normal skill level, the higher Language is. If Jones had Language skills in Kovaunn or Zelenae, he could at least understand the readouts and dials! The Culture skill and its subspecializations (technical, transport, military) allow a character to apply skills at full level. So if Jones has a high Culture knowledge of both Kovaunn and Zelenae, he'd get Navigation up to his Culture skill in either area. (Culture cannot be raised higher than Language, btw, so there's no worries about Jones being illiterate but skilled in flying). Whew! Okay, that's it for now. Comments as always greatly appreciated. [edited by - wavinator on April 5, 2004 4:25:48 AM]

Share this post


Link to post
Share on other sites
Advertisement
To be honest, my eyes glazed over about halfway down the post. However, here are my thoughts for the parts I read:

Skill decay might make more sense if skills decayed faster at higher levels. Maybe instead 1% of the current skill, with a minimum of 1 point per day rather than a straight 1 point per day? Since it''s harder to gain skill at higher levels, it makes sense that losing skills is easier at higher levels as well.

Also, there seems to be a discontinuity in the use of experience levels. That is, if the actor is green and the target is green, shouldn''t it be a straight 50% of trials, not 75% of trials? I can see it working as it stands against static difficulty values rather than against actual opponents.

The skill test chart seems rather redundant. You''ve already specified the chances of success for each trieal. Is it really necessary to specify who''s on top and who''s on bottom?

Also, I''d tamp down the ranges for the critical success just a bit in your chart. With your best value/per trial approach, the critical successes are going to be disproportionately common with the current rates.

For skill improvement, it''d probably be better to say that every critical success in a won trial gives a point rather than a critical success in a skill check. This way you can beef up the number of trials for less often used skills to keep up the earned skill rate with more frequently used skills.

Share this post


Link to post
Share on other sites
Well, It seems like a solid system, not very intersting to read but then math isn''t suppose to be.


One thought under skill improvement, what about mentoring? You assign a more experinced crew member to mentor a less experinced, one. So they work together and the students recives a percentage of the mentors skill.



-----------------------------------------------------
Writer, Programer, Cook, I''m a Jack of all Trades
Current Design project: Ambitions Slave

Share this post


Link to post
Share on other sites
Wow. That's an amazing system you've got there. Here's what I've got to contribute:

quote:
Original Post by SiCrane
Skill decay might make more sense if skills decayed faster at higher levels. Maybe instead 1% of the current skill, with a minimum of 1 point per day rather than a straight 1 point per day? Since it's harder to gain skill at higher levels, it makes sense that losing skills is easier at higher levels as well.

I agree with this. Maybe you could tie this into the intellect stat, or to some other character trait, so that some characters will have extremely good retention while others will only be able to maintain skills with constant practice and training.

In addition, I'd like to see skills raise faster when they're being relearned. If Sgt. Sykes was a 300-level navigator before he was transferred to your craft, and he's been out of the loop for a while, he should pick it up again faster than some green cadet who's taking the helm for the first time. I see that the experience level will have an impact on this, since he's likely to succeed more frequently, but I'm not sure that will be enough to emulate his reaquisition of the skill.

----------------------------

Next, I have a question about the skill slot system. If a character has 20 Intellect, and thus 4 skills, can he have only four skills total, or can he only have four skills active at any time? Also, when a skill is deactivated, can that crew member not perform that task, or is it just in a steady state of decay? It seems odd to me that a character could have a skill level in conduit maintenance but be unable to maintain any conduits because he's not actively thinking about that skill at the time.

----------------------------

quote:
Original Post by SiCrane
Also, there seems to be a discontinuity in the use of experience levels. That is, if the actor is green and the target is green, shouldn't it be a straight 50% of trials, not 75% of trials? I can see it working as it stands against static difficulty values rather than against actual opponents.

This is another good observation, but it seems to me that since most of the "opponents" in skill tests will be inanimate (like control panels or astrological phenomena) they will not have relevant experience levels. But for things like combat between soldiers or psionic effects, experience levels will surely be relevant. After all, if two Elite gunmen draw and fire, it's almost certain that both will get the requisite 1 successful test, and so each would inevitably kill the other.

In all honesty, I'm not sure I fully understand the "experience level" system you've described here. I like it, because it allows for some characters to become reliably dominant, which will be a major concern in strategy, but it does seem that a clash of titans will end in a bloodbath. SiCrane's suggestion of taking that into account is a good one, and I'd like to see an algorithm for comparing experience levels in tests that pit two characters against one another.

----------------------------

I'm not sure I like the skill test system. I'm trying to see what it is supposed to be emulating, and I can't quite figure it out. It seems that the multiple trials would be bizarre in some instances, and the percentage of winning would be bizarre in others. Remember that Civilization was a turn-based game, and so the multiple trials and whatnot might not translate properly into your design.

Taking your example of the wormhole, where Jones has to navigate the complex gravities of the phenomenon, why would he have to win the majority of ten tests? Wouldn't it just be him and his helm console versus the gravity? If that's the case, then it would only be a matter of time before he got a good trajectory calculated. Maybe if you set him the task an hour before the ship had to adjust its course, he'd be able to come up with ten difference possible routes, and then you'd choose the best of those ten, which would almost always be a big success. However, if you're in combat and you need to jump the hell out of there, he'd only get time to do it once, which might be a catastrophic failure. If the number of trials was a function of time, it would seem more intuitive. You can still get multiple trials, but only if you're going to have the same guy hammering at the same problem more than once.

On the other hand, if a boarder armed with a plasma rifle and a defender armed with a laser projector are facing off, they'll never see more than one trial in their combat. They'll raise their weapons, sight in, and take the shot. Whoever gets shot first loses. Maybe if they both miss they'll get another trial, but it won't really be the same problem, and it'll be later in time. This will also eliminate the mutually assured death I noticed above. They might both die, but it'll be less common.

If you link trials to time, then you can have some crew members who have skills to solve problems and go get a beer, while greener folk will perhaps spend an hour tinkering with a system to get it to the same level of efficiency.

----------------------------

Everything else seems to be harmonious, innovative and applicable to the game. Characteristically superlative work, Wavinator.

Editted repeatedly to make it better.

[edited by - Iron Chef Carnage on April 5, 2004 4:32:03 PM]

Share this post


Link to post
Share on other sites
Guest Anonymous Poster
I must be reading something wrong.

At the very top, where you make success/failure calculations, you say that if the player has a skill of 100 and the challenge has a difficulty of 300, they have a 25% chance of success. However, by your calculations, if they have a skill of 300 and the challenge as a difficulty of 100, they have the same 25% chance of success. Clearly, this is wrong.

Share this post


Link to post
Share on other sites
I assume you''re referring to this, AP:
quote:
His Navigation skill is 100. The wormhole''s difficulty is 100. The total is 200.
100 / (100 + 100) = 50%.
Jones has a 50% chance of success. Now if he navigates another wormhole, with a 300 challenge rating, the calculation is
100 / (100 + 300) = 25%


The "300 challenge rating" refers to the wormhole''s difficulty. Jones'' skill is still 100. That''s a 25%. If the tables were turned, and Jones had 300 while facing a wormhole test of 100 difficulty, he''d have a 75% chance of success. It would look like this:

300 / (100 + 300) = 75%

See?

Share this post


Link to post
Share on other sites
Guest Anonymous Poster
That does clear things up.

Share this post


Link to post
Share on other sites
quote:
Original post by SiCrane
To be honest, my eyes glazed over about halfway down the post.



Yeah, sorry about that, posting in-depth designs is always a bit dry. :/

quote:

Skill decay might make more sense if skills decayed faster at higher levels. Maybe instead 1% of the current skill, with a minimum of 1 point per day rather than a straight 1 point per day? Since it''s harder to gain skill at higher levels, it makes sense that losing skills is easier at higher levels as well.



I thought about 1% per day, but that seemed a bit unfair and unrealistic. Although, I have to remeber it''s just a game, and I''m not simulating memory.

quote:

Also, there seems to be a discontinuity in the use of experience levels. That is, if the actor is green and the target is green, shouldn''t it be a straight 50% of trials, not 75% of trials? I can see it working as it stands against static difficulty values rather than against actual opponents.


Only the actor''s aim counts in this case, I''m not factoring in evasion as directly as you might think. It doesn''t matter if the target is green. What counts is the actor. Greens will miss more often than elites, regardless of what they''re shooting at.

quote:

The skill test chart seems rather redundant. You''ve already specified the chances of success for each trieal. Is it really necessary to specify who''s on top and who''s on bottom?


I''m not sure I understand you here. The thing about top and bottom simply let''s you see how the percentage chance for success or failure actually work. (Admittedly, it might have been an excess mechanical detail not relevant to a design discussion.)

quote:

Also, I''d tamp down the ranges for the critical success just a bit in your chart. With your best value/per trial approach, the critical successes are going to be disproportionately common with the current rates.


OUCH! Big flaw! I see what you mean. The chance of getting at least one critical is going to rise the more trials there are. So there should only be 1 check for degree of success not best value/per trial if I want the 5% chance for criticals to hold.

If I do that, Elites will still hit more often than Greens, but everyone will have a 5% chance of critical be they Elite or Green. Elites will just see them more often, statistically, because they''ll hit more often.

So in a 50/50 contest with 4 trials, a Green will have 12.5% chance of winning(.5^3); while an Elite will have an 93% chance in the same.


quote:

For skill improvement, it''d probably be better to say that every critical success in a won trial gives a point rather than a critical success in a skill check. This way you can beef up the number of trials for less often used skills to keep up the earned skill rate with more frequently used skills.


This is what I meant, I wasn''t clear. The way it''ll work now is that you get a point if you both win a skill test AND get a critical in the process.



--------------------
Just waiting for the mothership...

Share this post


Link to post
Share on other sites
quote:
Original post by TechnoGoth
Well, It seems like a solid system, not very intersting to read but then math isn''t suppose to be.



Yes, I''ve learned my lesson, never again will I post something so dry. (It''s funny, though, because math is the nuts and bolts of game balance and game systems-- no wonder so many of them are screwed up! They''re the least interesting thing to work on! )

quote:

One thought under skill improvement, what about mentoring? You assign a more experinced crew member to mentor a less experinced, one. So they work together and the students recives a percentage of the mentors skill.


Okay, kind of like on the job training. What do you think the cost / tradeoff of this should be? This is like training, but you wouldn''t lose staff, so it wouldn''t be a labor tradeoff. Simply requiring an extra workstation doesn''t seem to be a steep enough tradeoff either.

Maybe the requirement should be that the two individuals can''t be more than 50 or 100 points apart in skill level? Or that they have to be compatible personalities? Hmmm... I like it, but I''m not sure what the cost of having someone tag along as you do your work would be in gameplay terms.

--------------------
Just waiting for the mothership...

Share this post


Link to post
Share on other sites
I like your skill deterioration scheme. So many games let players build ''super'' characters which is completely unrealistic, IMHO. I guess its fun for some players in single player games.

However I would adjust the deterioration so that it doesn''t become a reason why players dislike your game. The skill decline should continue until the skill is half of its original value before the decline began. Once the character returns to using the skill, it advances faster than it would under normal circumstances until it reaches the original value.

Share this post


Link to post
Share on other sites

  • Advertisement
×

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

Participate in the game development conversation and more when you create an account on GameDev.net!

Sign me up!