Jump to content

BattleMech Advancement and Skills: Essay length, so be ready


23 replies to this topic

#1 Andrew Harvey

    Member

  • PipPipPipPipPip
  • 142 posts
  • LocationCornfield, Indiana

Posted 07 November 2011 - 07:42 PM

This is a development of an idea I've had. I believe that a player's skills should be used as a factor that enhances certain facets of a BattleMech's operation rather than only as prerequisites to unlock 'better' 'Mechs. This would be a good way to guide a player's movement through 'Mech types without adding overt restriction that makes them feel confined to a certain 'path' or 'tree'.

What I envisioned was a list of Major Skills that a player's pilot could develop over time. Every important part of a 'Mech's operation would be included with subsets to each skill:

-Gunnery&Weapons
[This is a dual-skill type. The Gunnery Skill governs the targeting of enemies at different ranges (Short to Long) and different modes (Direct Fire, Indirect Fire), as well as affect general accuracy (resistance to recoil, zeroing-in of reticule). The Weapons Skill governs specific weapon-type use and accuracy (LRMs, PPCs, A/Cs).]
-Piloting
[This governs the ability to maneuver a 'Mech through different speed ranges and different terrain types. Also includes Engine Operation which measures how quickly the engine responds to speed changes and how quickly the 'Mech accelerates (partly dependent of 'Mech weight).]
-Jump Jets
[This governs the use of Jump Jets in different 'Mech weight classes through Neutral situations (traversing of terrain obstacles and elevation changes) and Combat situations (avoiding weapons fire, attempting DFA) with Combat use being more demanding. Jump Jets could be included as a subset of Piloting, but I hesitate to since Jump Jets are such specialized equipment.]
-Electronics
[This governs the operation of Sensors, different Sensor types, and target acquisition/detection. Included is the operation of advanced Sensor packages (like TAG, ECM, possibly NARC Beacon). It also includes the use of comms through environments that could disrupt it (urban environment, electronic interference, etc).]

A more tentative skill type; it's more general to BattleMech Operation as a whole, but includes important components that aren't as easily included in the other 4:
-BattleMech Control
[This governs the severity of: weapon impact/ramming effects on 'Mech balance, the negative effect of speed on maneuvering, the effects of a damaged gyro, the speed of 'Mech startup/shutdown, and the effect of heat on the Pilot. These are Pilot specific points that relate to personal endurance/stamina and how well they interface with a neurohelmet. A more seasoned pilot would be less affected by heat, more able to keep a 'Mech's balance, and etc.]

I didn’t include any skill involving armor because there shouldn’t be a skill that governs that. Aside from differences between Standard and Ferro Fibrous and etc, 1 armor point should offer the same protection as any other, and take the same amount of damage as any other.

#2 Andrew Harvey

    Member

  • PipPipPipPipPip
  • 142 posts
  • LocationCornfield, Indiana

Posted 07 November 2011 - 07:48 PM

With what I think of as the 4 Major Skills, I created this chart to organize BattleMechs by their skill use, weight class, and utility.


Posted Image

only an example, didn't have the space to recreate all possibilities and etc


I imagined that some 'Mechs would have 1 or 2 Major Skill types listed as Primary, with other skills and skill subsets listed as Secondary Skills. Supporting skills/skill subsets would be listed as Ancillary Skills.

A Whitworth would read like this:
Primary Skills - Gunnery/Weapons
Secondary Skills - LRMs, Jump Jets
Ancillary Skills - Piloting, Lasers

While a Phoenix Hawk would read like this:
Primary Skills - Piloting, Jump Jets,
Secondary Skills - Gunnery/Weapons, Electronics
Ancillary Skills - Lasers, Machine Guns

And a Cicada:
Primary Skills - Piloting
Secondary Skills - Electronics
Ancillary Skills – Lasers

The Primary/Secondary/Ancillary skill designations define their roles. The Whitworth is an LRM fire-support platform that uses Jump Jets to overcome a low maneuverability focus, and Lasers to backup the LRMs. The Phoenix Hawk uses speed and maneuverability coupled with a decent focus on weapons to act as a 'heavy scout' BattleMech. The Cicada is a more traditional and simple scout 'Mech, with high speed and maneuverability but little provision for weapons. Compared to the Phoenix Hawk, it is arguable less useful. Because the Cicada is simpler though, the investment in skills required to make full use of its performance will be smaller compared to the Phoenix Hawk.

#3 Andrew Harvey

    Member

  • PipPipPipPipPip
  • 142 posts
  • LocationCornfield, Indiana

Posted 07 November 2011 - 07:53 PM

These designation sets don’t mean that a player would be restricted from piloting a BattleMech for lack of skills though.

For example, let’s say a player is getting bored of playing light ‘Mechs and wants more powerful weapons. Through various means, he gets enough C-Bills to purchase a Hunchback, which he perceives as having ample firepower. However, it seems much slower than it ought to be, his A/C-20 fire is inaccurate, and the ‘Mech in general is very hard to him to operate. Now maybe our Pilot will stick with it and fight through the obstacles, but most likely he'll dismiss it as a bad ‘Mech and get rid of it without understanding why he had so much trouble.

Now let’s see a different pilot. She spent a lot of time in the Urbanmech and Blackjack, gaining skill in slower, Ballistic-heavy ‘Mechs. Moving to the Hunchback, she finds a design and role very similar to what she’d been previously piloting. Having developed experience in Piloting with her ‘slow’ ‘Mechs, and in Gunnery/Weapons with an Autocannon specialty, she is well prepared for the Hunchback. Compared to our other pilot, the Hunchback’s shortcomings disadvantage her less, while she makes better use of its advantages. So while a BattleMech can be used without considering skill sets, a Pilot will only receive 100% of its performance after they’ve achieved some amount of competency in the skills related to the ‘Mech and its role.

So, with a chart like this, BattleMechs could be organized in a way that guides and makes use of a player's skill advancement. As a Pilot improves their skills it suggests a 'Mech that would make better use of them. A person whose Pilot had developed excellent Electronics and Piloting skills could pilot a Panther if they chose, but a 'Mech in the Raven tree would make the best use of their skills and make them more effective. Conversely, a lack of skills in Weapons/Gunnery or Jump Jets would make them a weaker and less capable Panther pilot.

There are several advantages and disadvantages to the use of this kind of system.
Pros:
-Skill system uses good aspects of RPG elements without the bad design associated with leveling and grinding tedium.
-Results in a more organic method of BattleMech progression that guides players without forcing them.
-Helps retain utility of lighter designs over heavier designs.

Cons:
-Harder to design
-Harder to balance
-Increases learning curve and difficulty
-Makes game less open to new players

This is a big ball of complexity, I know. But hopefully, anyone who’s read it completely will agree there are few good ideas in there.

Edited by andrew harvey, 07 November 2011 - 08:02 PM.


#4 Raj

    Member

  • PipPipPip
  • 66 posts

Posted 07 November 2011 - 11:43 PM

That is quite interesting. I like the idea of gathering experience in a certain type of piloting and bringing that skill to a new mech. This would eliminate the annoying pilot grind that WoT has where as soon as you get a new tank even of the same class you have to retrain gunnery, driving, radio operator(>.<), etc..., but keep the leveling gameplay.

The only thing I can think to suggest is adding a weight class skill to the piloting since moving a 80 ton mech around would probably be quite different than a 20 ton.

#5 Astaroth

    Member

  • PipPipPipPipPipPip
  • 298 posts
  • LocationColorado

Posted 08 November 2011 - 06:37 AM

Mad props for such a well thought-out and detailed concept. Very well done.

It's difficult to find much to argue with here. I think you've got a terrific idea that would lend itself very well to an, as you said, organic methodology of developing pilot skills and mech progression. This would be a great way to help players choose their next mech by showing them what is best suited to their play style. It's a lot like the character skill system you see in The Elder Scrolls series, where your character naturally gets better at the skills you actually use, thus eventually creating a character who is focused in a way that best suits you, the player.

Unfortunately, as you pointed out, designing and coding such a progression system is probably quite a challenge, quite possibly prohibitively so. It's a shame because I really like this idea and would love to see it included in a game.

#6 Charles Okonkwo

    Rookie

  • 5 posts

Posted 08 November 2011 - 08:23 AM

Wow! Respect for such a great concept!
When I first read that the new MechWarrior will have persistent skill sets, I thought to myself:" Nooooo! Please don't do this, it will make the game suck BIG TIME!". I've had horrible visions of messages saying "you do not have the required skills to use the BAP." while the player goes "***? When I read they want to make the game focused on piloting skills, I didn't know they meant THAT!". I think a player should always be able to use a certain piece of equipment or any kind of mech they just aquired. I always found RPGs stupid where you spend skill points on some branch of a skilltree and from one second to the other your character knows how to use bows. X_X And then after a while you find out that you actually don't like bows at all and you need a respec (tell me, how do u actually UNLEARN a skill in real life so u can learn something else? O_o), which you often have to BUY with real money in many MMOs. Such concepts don't feel like "role-playing" at all and make the player have to plan out his whole character progression before he even creates his character in order to have the best possible game experience. He has to make game making- or breaking decisions before even playing. Do you want to be a scout forever or do you want to be an in-fighting pilot forever? Tell you what: I don't want to be ANYTHING FOREVER! It's like asking 12 year olds what job they want to have for the rest of their lifes. I want to be able to change my style of playing as it suits me right now and not have to follow the tracklines of the skill path I chose at the beginning of the game. Pilots also need to be able to adapt to the current situation: What if there's a big battle taking place in a rocky mountain range with many canyons and yet my Merc Corp has no dedicated scouts or only 1? (suppose most players will prefer heavy mech combat, as blowing stuff up is simply more fun) What if a Merc Corp wants to take place in a raid but has mostly defenders in its ranks? Mechanics like these don't make a character seem like a human able to learn, evolve and adapt, but rather like sculpture carved out of wood or stone.
Andrew's concept may be a way to avoid these problems while still implementing a skill-tree. After all character skills can make the game more realistic and add a new and fun role-playing side to the game which past MW-games have been lacking. I really hope it gets the developper's attention. But no matter if his ideas make it into the game or not, there MUST NOT be a limit to how many skills a player can learn as to not prevent players from changing and adapting. A man will always be able to learn new things in the real world, even if he's already an expert at mathematics. And that should be the same for any game with role-playing aspects.

#7 Andrew Harvey

    Member

  • PipPipPipPipPip
  • 142 posts
  • LocationCornfield, Indiana

Posted 08 November 2011 - 08:00 PM

View Postraj, on 07 November 2011 - 11:43 PM, said:

The only thing I can think to suggest is adding a weight class skill to the piloting since moving a 80 ton mech around would probably be quite different than a 20 ton.

Right and right. I'd forgotten somewhat about weight classes. Rather than make it a subset of Piloting though, I had the idea to keep it separate from the 4 Major skills. I would take the Weight Proficiency skill [thanks again raj ], a skill called Tech Capability I'll elaborate more on later, and the BattleMech Control skill, and rope them all into a dominant class I'll call Pilot Competency.

Pilot Competency would include all those skills that are not directly used in battle/combat situations, as well as Pilot-specific skills that relate to their cockpit and neurohelmet familiarity.
--BattleMech Control: (The same as before in that it governs skills that depend on a Pilot's stamina & fitness, how 'at home' they are in a cockpit, and how strong of a neurohelmet connection they've developed.)
--Weight Proficiency: (Governs the success a Pilot has in operating BattleMechs of differing weight classes. It does not prevent outright the use of any weight class, but a lack of a proficiency would leave a portion of that weight class's performance untapped.)
--Tech Capability: (Among other things, this skill would govern the ability of a player to easily make repairs to their 'Mech and control a weight-class organized ability to modify a BattleMech's loadout and specs.)

One reason I added the Tech skill was because I realized that while my Skill Tree had a lot of advantages, it would be pretty much scrambled by customization. And I support the ability to customize BattleMechs; I don't think a player should be limited to major variants. The Tech Capability skill comes as a kind of compromise. I do think smaller modifications (like exchanging a laser for a ton of armor and etc) should be available from the start with an investment of time and C-Bills. Following that though, I think that the successful completion of more sweeping changes should depend on the player's Tech Capability, with the difficulty and chance of failure growing as the modifications become more complicated. The skill would advance through the player's repair/maintenance of their 'Mech and through their time spent operating 'Mechs in that class.

So let's revisit one of our Pilots. She wants to replace an LRM-10 in her Valkyrie with a pair of LRM-5s and another ton of ammo. Since the weapon type of the components she's swapping are identicle, and are of similar size, the difficulty is not as high. She's spent enough time operating and repairing Light BattleMechs that the chances are high that the modification will be successful. Initiating the customization process, the Pilot comes back after a few hours to find her project was a success.

Now let's see our other Pilot. He just made the transition from Light to Medium 'Mechs, and wants to customize his Clint by removing the A/C-5 and installing a Large Laser, 3 Heat Sinks, and a ton of armor. These modifications are much more complex, but he thinks his Tech skill will be high enough. However, his Tech Capability hasn't been developed enough to start making a difference in Medium 'Mech modifications, and the radical changes he is proposing only increase the difficulty. Disregarding the consequences of failure, he starts customization. His homecoming is much less happy than our other Pilot; not only did the modification fail, but he lost some of the equipment he was trying to install in the failure.

A system allowing Battlemech customization, albeit with a low chance of success without some skill in technology, would be good in a number of ways. It is more realistic in that a rookie Mechwarrior would have no idea how to jury rig weapons, while a more seasoned Pilot would have more knowledge of ad hoc modifications. Balance-wise it would also be beneficial in that it would prevent random customization and instead promote the development of skills that will benefit a player most. It will also create renewed interest in Light BattleMechs a while after launch, because while operating Medium and even Heavy designs may be becoming more feasible by that time, Light Mech modification will be viable long before heavier classes.

#8 Charles Okonkwo

    Rookie

  • 5 posts

Posted 09 November 2011 - 08:11 AM

The skill Tech Capability would only make sense if it was the Mechwarriors themselves doing the changes or at least directing the personel at every single step. However, I always thought that the Mechwarrior only told the staff of the hangar how he would like his Mech customized and anybody working at the hangar would have to have the required capabilities to modify mechs in every way possible. Aside from that, while I do see how this skill would benefit the use of lighter mechs, as every player starts out with a light one, I can also imagine that it will force players to play with configurations that don't suit their play stayle. Say a pilot who's been using a light mech with lasers, and thus is skilled at heat management, buys a medium mech that comes with mostly ACs. Because he's scared of losing his expensive equipment he can't just get rid of an AC, add a laser, armor and maybe more speed. He'll have to play several missions with the base configuration which he's not comfortable with at all. But not only that, it would prevent players from customizing all game long because there's always a small chance that a customization process fails. Imagine the frustration when you loose your precious and expensive new gauss cannon to a 5% fail chance. I think this would be too much restriction.

#9 Andrew Harvey

    Member

  • PipPipPipPipPip
  • 142 posts
  • LocationCornfield, Indiana

Posted 13 November 2011 - 02:37 AM

View PostCharles Okonkwo, on 09 November 2011 - 08:11 AM, said:

The skill Tech Capability would only make sense if it was the Mechwarriors themselves doing the changes or at least directing the personel at every single step. However, I always thought that the Mechwarrior only told the staff of the hangar how he would like his Mech customized and anybody working at the hangar would have to have the required capabilities to modify mechs in every way possible.

You're right to bring up maintenance personnel, it'd be surprising if they were ignored completely in-game. I just think that a players skills will have to make an impact at some point in order to guide the kind of customization that can be done. In the canon, given a good enough tech and enough money, you could make innumerable changes to your 'Mech. But for the sake of balance, I think that the game should deviate somewhat in this respect.


Quote

Aside from that, while I do see how this skill would benefit the use of lighter mechs, as every player starts out with a light one, I can also imagine that it will force players to play with configurations that don't suit their play stayle. Say a pilot who's been using a light mech with lasers, and thus is skilled at heat management, buys a medium mech that comes with mostly ACs. Because he's scared of losing his expensive equipment he can't just get rid of an AC, add a laser, armor and maybe more speed. He'll have to play several missions with the base configuration which he's not comfortable with at all.

Part of the idea was to make the choosing of a Battlemech a much more serious and thought out process. After being acquainted so long with lasers, as in the case you put forward, it would be logical for the player to choose a 'Mech that complements those skills. There's enough 'Mech types at the 3049 point in time that you could have a progression in 'Mech weights that would have similar types of weapon loadouts. He's welcome to choose a 'Mech with different weapon types, like A/Cs as you mentioned, and the consequences are his to deal with.


Quote

But not only that, it would prevent players from customizing all game long because there's always a small chance that a customization process fails. Imagine the frustration when you loose your precious and expensive new gauss cannon to a 5% fail chance. I think this would be too much restriction.

You have a point that I went too far here. Losing a piece of equipment entirely wouldn't be acceptable. For it to become damaged might be more acceptable, but you're correct that complete destruction of components and equipment shouldn't take place outside of combat.

#10 Amechwarrior

    Member

  • PipPipPipPipPipPipPip
  • Legendary Founder
  • 729 posts
  • LocationHawaii

Posted 13 November 2011 - 03:53 AM

Wow, great ideas. I like how it is designed to not deny any one path and allows us to change up our styles. About the Tech thing and maintenance crews, just change the name to something like "Crew Capability(or Competency/Clout)/etc" to represent him bringing in the same 'mech in to the repair and modifications bay and the tech crew becoming more familiar with fixing and modifying his personal 'mech. Like having your own mechanic, they develop a relationship with your vehicle over time, you bring them a new one and it is like you mentioned. The skill and know-how is there, the familiarity is not. This way the Techs get the credit, not just the guy piloting the machine and all you have to do is change the name of the skill.

#11 GreyGriffin

    Member

  • PipPipPipPipPipPipPip
  • 792 posts
  • LocationQuatre Belle (originally from Lum)

Posted 13 November 2011 - 03:57 AM

I very, very much approve of skills being intrinsic to the pilot and not the 'Mech, and allowing skills to transfer, to a degree. A mechwarrior who knows his way around an autocannon should know his way around an autocannon.

I would not be opposed to specific 'Mech specializations, though - optional skills that allow a player to get really good with his chosen 'mech, either in addition to or in place of increasing his general skill, much like specialties in Shadowrun.

#12 Andrew Harvey

    Member

  • PipPipPipPipPip
  • 142 posts
  • LocationCornfield, Indiana

Posted 14 November 2011 - 03:48 PM

I took my BattleMech Skill Progression idea a little further, chartering 'Mechs from 35 tons to 60 tons. Still only a barebones chart, but I think it provides a good framework that makes sense. I've half a mind to continue developing the chart, but I'm a little worried that if I keep going, I'll be tempted to create a comprehensive 20t to 100t chart, and that'd drive me insane.

Posted Image


View PostAmechwarrior, on 13 November 2011 - 03:53 AM, said:

"Crew Capability(or Competency/Clout)/etc"

You make a good point. I guess it all depends on whether they model 'Mech repairs/modifications/etc through a tech crew or not.

#13 Charles Okonkwo

    Rookie

  • 5 posts

Posted 15 November 2011 - 12:05 AM

I still think this idea of failed mech customization is too restrictive. A game must not prevent palyers from trying out new things.

#14 collosus

    Member

  • PipPipPip
  • Knight Errant
  • 77 posts
  • LocationArgentina

Posted 15 November 2011 - 12:16 AM

This is a very interesting idea and I like it, much to the TES series (like astaroth said), and it complements much to the way of playing of different people (in my case E-WAR ftw!).

#15 Kudzu

    Member

  • PipPipPipPipPipPipPip
  • 769 posts
  • LocationSomewhere in the SEC

Posted 15 November 2011 - 12:58 AM

Very well thought out!

What changes would you make to this now that we know you can start with any weight class you want rather than working your way up from light to assault? Would you keep it as a straight mech to mech flow or would you change the skills to be more tech-oriented (such as unlocking a certain skill from the autocannon tree allows you to upgrade any of your mechs from a AC/10 to LB-10X)?

#16 Mezzanine

    Member

  • PipPipPipPipPip
  • 106 posts
  • LocationMinneapolis, MN

Posted 15 November 2011 - 07:16 AM

This is a great concept, and hopefully elements of this have made their way into the game design.

This actually might help new players, since the skills and progressions provide a path from mech to mech.

If I were brand new to the BattleTech universe, I probably wouldn't know that a Hunchback can be the medium equivalent of an Urbanmech, but with that extra guidance it gives me something to work towards and keeps me engaged with the game.

Providing guidance and information about the relationship between different mechs can help new pilots find a chassis that fits their playing style, without spending hours poring over the different mech designs.

#17 Havoc2

    Member

  • PipPipPipPipPipPipPip
  • Mercenary
  • 505 posts
  • LocationBarrie, ON

Posted 15 November 2011 - 08:50 AM

While I respect your ideas and the thought you have given to this, I must respectfully disagree.

Firstly (as you've said already) it makes it very difficult for new players to pick up the game. Once they choose their starting 'Mech(s) they're pretty much pigeonholed into their choice. Even for veteran players who prefer 1 type of combat or another they are pigeonholed into fire-support/recon/urban brawler whatever they start out choosing.
This type of progression is a lot like many MMORPGs where you choose your "class" (assassin, healer, tank, whatnot) and are pretty much set in that role unless you re-roll your pilot or create a new one.

Secondly, (again for the new people) much of this would be very difficult to properly train for. You're talking about 5 major skill categories plus subsets to think about training into. While many of us already have our favorite 'Mechs we could just look up and train for those required stats, newbies (and I'm guessing many vets too) may be torn between 2 different "classes" of 'Mechs. I'm not opposed to thinking about what I want my pilot to be able to do (God knows I spent enough time in DAOC speccing and re-speccing my characters depending on what I wanted to do) but unless there is some type of re-learning feature in the game, I don't want to pour hours and $$ into a pilot and be either trained towards something that I was loving a few weeks ago but is now getting stale, or something that doesn't benefit my lance.

I like where your ideas are starting from but would suggest that many of the subsets of your major skill categories be included into that category. So instead of training in Electronics and Jump Jets, maybe combine them into Support Modules. Battlemech Control could be included in Piloting and your Weaponry and Gunnery class could consist of the 3 weapon class: Energy, Ballistic, Missile.
This could allow pilot specialization without pigeonholing a pilot into certain 'Mechs.


Also, this all works out great but we're assuming that there is no 'Mech customization allowed (which after reading the FAQ I'm confused as to whether it will be allowed or not).

Edited by }{avoc, 15 November 2011 - 12:56 PM.


#18 wolf74

    Member

  • PipPipPipPipPipPipPipPip
  • 1,272 posts
  • LocationMidland, TX

Posted 15 November 2011 - 12:52 PM

You have the core of what I would also do if I has the software skill need for a game. Each Tree Level cost about 1.5-3x more than the one before it the “General Skill" Effect all skill in the current level. Each Branch of the Tree is Given the boost from it core. AKA Level 2 Mech Piloting 5% would go into all Mech Piloting that are link to it from level 3 up. Personal I would say CBT Spec are the Average Base line. A pilots skill in the operations of the mech can strongly effect it operation. So take a CBT spec drop it by 25% for a Starter. I’ll try to explaine more at the end of the basic skill tree how I see. Just going up the Light Piloting branch

Level 1 Tree: (1st one is free)
Mech Basic (Give 5% to all mech based skills)
Vehicles Land based - For Future
Aircraft - For Future
Aquatics craft - For Future
Large Transportation - For Future (AKA dropships)

Mech Basic Level 2 Tree
Mech Piloting 5% (Over is now 10% or 13% with Operation)
Mech Weapons Operation 5% (Laser, Ballistic, Missiles, Heat Control)
Mech Electronic 5% (Rader, Thermal Sensor, ECM, ECCM, BAP, TAG, Etc)
Mech General Operation 2%

Mech Piloting Level 3 Tree
General Piloting 3% to all Weight Classes
Light 10% (Over all is 26% if they Also got the General Piloting, 23% without)
Medium 10%
Heavy 10%
Assault 10%

Mech Light Piloting Level 4
General Light Piloting 2%
Turning 10% boost
Speed 10% boost (36% over all or 38% with General Light)
Stability 10% boost

Mech Light Piloting Speed lvl 5
Speed Operation General 5%
Top speed 10% (By now you get the Idea)
Acceleration / Deaccelerate 10%

By using a small Skill boost, This would let those in a given weight class to get the most out of there machines. At about 34% skill boost your at the Average Mech pilot make the mech work at it CBT level (Vs the 75% starting line I suggest at the start) with 55% boost you could have the mech working at 116.25% it cbt value showing your and Elite pilot. This will also make it where a Skill Light Pilot Could Take a good it from Larger mech and Not Fall. Or to be able to due that High Speed Turn to not get shot by the other mech that just came around the corner. Where someone who is having to drive a Light at the moment do to the fact his Heavy mech is in the shop could not do as well in the same Light mech.
For Weapons systems, Speed at which they converge, For ammo basic weapon possible, tweek on the reload time, With a Minor jamming chance (The High you tweek it the higher the jamming become). Energy weapon Faster reload but at the cost of extra heat in the recharge cycle. Lowering the lock-on for missile system. Are all possible options

just my personal thoughts.

#19 Andrew Harvey

    Member

  • PipPipPipPipPip
  • 142 posts
  • LocationCornfield, Indiana

Posted 15 November 2011 - 10:02 PM

Oh dang, wall of text, GIRD YOUR LOINS!!!

View PostCharles Okonkwo, on 15 November 2011 - 12:05 AM, said:

I still think this idea of failed mech customization is too restrictive. A game must not prevent players from trying out new things.

Well, my idea wasn't to restrict players from trying new new weapons, I just think that skills could be used as a way to guide customization, most especially in the early stages of a pilot's development. In the canon it was often very difficult to swap different weapons, for example some A/Cs (even those of the same class) weren't of equal dimensions and didn't always fit in the same space. We might not have customization at all, and if we do, it may be dependent on a tech crew rather than a player's personal skill. If it is based on a player's skill though, I think that skills could be used to help a player make better decisions. Outright failure might be a bit strong though, you're right.



View PostKudzu, on 15 November 2011 - 12:58 AM, said:

Very well thought out!

What changes would you make to this now that we know you can start with any weight class you want rather than working your way up from light to assault? Would you keep it as a straight mech to mech flow or would you change the skills to be more tech-oriented (such as unlocking a certain skill from the autocannon tree allows you to upgrade any of your mechs from a AC/10 to LB-10X)?

Wow, I wasn't aware of that weight class news! I still think that having a pilot start out with the more simple 'Mechs of a weight class would make sense. I think that the Assault class should start out with the four basic 80-tonners; the Awesome, the Charger, the Victor, and the Zeus. These 'lighter' BattleMechs could pave the way for more complex and powerful Assault 'Mechs; for example, a player piloting a Victor would gain experience(Autocannons and Jump Jets) that would be useful if they decided to move up to the Highlander. As for upgrades, I do think that when weapons like the LB-10X and the Ultra A/C-5 come out, corresponding experience (in the A/C-10 and A/C-5 in this case) should allow for a degree of familiarity when using them.



View Post}{avoc, on 15 November 2011 - 08:50 AM, said:

While I respect your ideas and the thought you have given to this, I must respectfully disagree.

Firstly (as you've said already) it makes it very difficult for new players to pick up the game. Once they choose their starting 'Mech(s) they're pretty much pigeonholed into their choice. Even for veteran players who prefer 1 type of combat or another they are pigeonholed into fire-support/recon/urban brawler whatever they start out choosing.
This type of progression is a lot like many MMORPGs where you choose your "class" (assassin, healer, tank, whatnot) and are pretty much set in that role unless you re-roll your pilot or create a new one.

The objective isn't to outright restrict the choices of 'Mechs and 'Mech roles, it's to provide a certain level of guidance while rewarding intelligent choices. I envision a system that rewards continued experience with certain weapons, equipment, and chassis' with improved performance. A player who finds themself liking SRMs could develop that skill and use the tree to find BattleMechs that made good use of them rather than encounter them accidentally.

My idea is that firstly, a player could potentially choose to try any role they want at any time. However, they will be rewarded with better performance if they use 'Mechs that makes use of skills that they have developed. This does mean that someone only experienced with a specific weapon type like Lasers couldn't jump into a Panther and immediately see peak performance. But I think that is how it should be. An Autocannon-only player shouldn't expect to be able to jump from their Blackjack into a Catapult and be just as proficient with LRMs as they were with their A/C-2s. But that doesn't mean that they can't use a Catapult if they wanted to. They just need to spend time developing the relevant skills (which will in turn make them better players).

Quote

Secondly, (again for the new people) much of this would be very difficult to properly train for. You're talking about 5 major skill categories plus subsets to think about training into. While many of us already have our favorite 'Mechs we could just look up and train for those required stats, newbies (and I'm guessing many vets too) may be torn between 2 different "classes" of 'Mechs. I'm not opposed to thinking about what I want my pilot to be able to do (God knows I spent enough time in DAOC speccing and re-speccing my characters depending on what I wanted to do) but unless there is some type of re-learning feature in the game, I don't want to pour hours and $$ into a pilot and be either trained towards something that I was loving a few weeks ago but is now getting stale, or something that doesn't benefit my lance.

I like where your ideas are starting from but would suggest that many of the subsets of your major skill categories be included into that category. So instead of training in Electronics and Jump Jets, maybe combine them into Support Modules. Battlemech Control could be included in Piloting and your Weaponry and Gunnery class could consist of the 3 weapon class: Energy, Ballistic, Missile.
This could allow pilot specialization without pigeonholing a pilot into certain 'Mechs.

I can't argue that it would introduce a greater degree of complexity. I just think that it would be worth it because it would provide a method for players to plan their BattleMech progression in a way that would make sense in-game and in-canon. If a person wants to be an ace Archer pilot, he/she should use 'Mechs armed with LRMs and Medium Lasers. No-brainer. I'm worried that diluting those skill types would make the entire system less effective.

And I know that a skill-relearning feature seems logical in a gameplay sense, but what can you equate that to in the in-game universe? Unless there was some sort of psychological reprogramming, it wouldn't be right in game for a pilot to suddenly and rapidly shift from one set of skills to the next. Being identified as an expert in some skills would also have its advantages; teammates would know that you could be counted on to know what you're doing with your role, rather than just be someone trying it out.

Quote

Also, this all works out great but we're assuming that there is no 'Mech customization allowed (which after reading the FAQ I'm confused as to whether it will be allowed or not).

Well I addressed the idea of 'Mech customization somewhat earlier in the thread, although the that and the idea of a skill tree may seem like opposing ideas. Part of my reasoning in making customization skill-based was because it would take a player ''off the grid'' of the skill tree. My idea was that by the time a player could make full use of a Mech Labs customization suite, they would be knowledgeable enough about their skills that they would be more likely to make changes that would fit their intended playing style/role. Until they were fully capable of customizing their BattleMech, the skill tree could serve as a roadmap for where they want to be and what they want to be doing.



View Postwolf74, on 15 November 2011 - 12:52 PM, said:

You have the core of what I would also do if I have the software skill need for a game.
stats stats stats
just my personal thoughts.

Sorry, didn't want to copy your entire post. You've basically hit the nail on the head and stated in clear percentages what I had in my head as a vague idea. Offering incentives to developing skills in order to guide players down the paths/roles they're interested in is the main concept behind my idea. It would be a good idea to implement, because I don't think that making everyone equally skilled in every weapon, piece of equipment, or 'Mech type/weight would be a good way to garner player interest, or a good way to introduce players to the idea of specific roles.

Edited by Andrew Harvey, 15 November 2011 - 10:14 PM.


#20 Havoc2

    Member

  • PipPipPipPipPipPipPip
  • Mercenary
  • 505 posts
  • LocationBarrie, ON

Posted 16 November 2011 - 06:10 AM

View PostAndrew Harvey, on 15 November 2011 - 10:02 PM, said:

... because I don't think that making everyone equally skilled in every weapon, piece of equipment, or 'Mech type/weight would be a good way to garner player interest, or a good way to introduce players to the idea of specific roles.


I can see your points for training, however I strongly disagree with this.

I do like the ideas of training, but no one will be equally skilled in every weapon, equipment or 'Mech even if this was a simple Instant Action game like MW4 multiplayer. Certain people will always be tactically better, or better at placing their shots, or better at mitigating damage.
People who think when they fight should not be significantly outmatched by someone who has simply played longer.

One of the things I miss the absolute most about MW is the fact that when I wanted to fight, I could find a match, pick my 'Mech and kill. There was no grinding for gear, experience or whatever.
I would like to see the training ASSISTING in a Mechwarrior's success, not being the deciding factor.

Edited by }{avoc, 16 November 2011 - 06:11 AM.






1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users