Jump to content

Skill Tree Criticism And Suggestions


5 replies to this topic

#1 process

    Member

  • PipPipPipPipPipPipPipPip
  • Star Colonel II
  • Star Colonel II
  • 1,667 posts

Posted 09 February 2017 - 05:46 PM

First, it must be said:

You cannot charge players to re-spec.

To associate cost with customization is antithetical to the core principles of a Mechwarrior game. This is by far the most egregious problem with this initial iteration the skill tree system, and is extremely unfriendly to new and old players alike.

It will feel even worse when PGI inevitably re-works how certain abilities function, altering their value to players.

Second, by revising the mastery system to focus on a single variant, you are disincentivizing players from purchasing more than 1 variant per chassis. This may be desirable for certain chassis with similar loadouts, but for others it only serves to inflate the grind, since each variant now requires the equivalent time and cost that three used to.

In addition, while it might be more friendly for beginners, it immediately reduces the enjoyment for veteran players with large stables. To speak only for myself, it's taken years to master my 200+ mechs, and I'm not too keen on having to invest upwards of twice that long to master them under this new system.

This may also have the unintended consequence of undercutting PGI's own business model, as they may no longer be able to sustain themselves on mech packs and other large-bundle deals.

As someone who sincerely wants this game to be great, it's really distressing to see PGI propose a system that could potentially cripple their product and alienate their customers.

---

With the major negatives aside, I do think the new tree system has much to offer. The degree to which a player can tailor their mech is a vast improvement over the stale "skill" progression we have now. While it's not currently implemented, I see lots of opportunities in the proposed system to add variety and uniqueness at both the chassis and variant level.

A few suggestions to improve the tree system, making it more user-friendly, without compromising the concept of the tree:
  • Make all skill nodes permanently unlocked, and shared within each chassis. Having to re-acquire skill nodes is tremendously tedious and a major fun-sink. Make skill nodes cost more to unlock if that's what it takes to make the game economy work. I would much rather grind a little longer than to be punished for experimenting. It's especially asinine not to have them shared between variants of the same mech, since in theory it's the same mech you're supposedly mastering. This is essentially the same as needing to unlock modules with GXP, but per chassis, and even under my proposed system, a significant investment. Again, increase the cost to unlock skill nodes if you have to.
  • Remove GXP and HXP; use more MC. Too many currencies! By sharing experience via shared permanent skill nodes within a variant per suggestion 1, you eliminate the need for the intermediary HXP. GXP has always felt sort of tacked on, it would be better to use MC directly as an alternate for both cbill and XP upgrades and unlocks.
  • Mech XP sharing (inter-variant). To address the problem of disincentivizing more than one variant per chassis, other owned variants should receive some percentage of XP earned by other variants. This way, you aren't completely abandoning variants while you grinding a single one. Conceptually, you aren't efficiently mastering those other variants, but because they share a common chassis, there should be some common knowledge gained by piloting just one of the variants.
  • Mech XP sharing (intra-variant). Alternately to Mech XP sharing between variants, multiple purchases of the same variant should have access to the same pool of Mech XP. There's no sense in siloing each variant, especially if one is used more frequently than others. Even nicer would be for all variants of the same type to earn the same Mech XP simultaneously, allowing you to level them simultaneously, as with the current skill system.
  • Diminishing skill tree values to offset boating. The current skill tree values are access linearly. This makes it much too easy to pour points into a single weapon type, which already has the benefit of being easier to operate in-game. The total amount of bonus for a particular trait could be the same, but the greatest values are closer to the top, which has the added benefit further of incentivizing mixed builds.
  • Reinstate an "Elite" rank at the equivalent Mech XP as the current "Master" rank (57,250 XP). Throw us old farts some recognition for our years of grinding and support.
  • Reduce the number of skill points and/or skill nodes. 91 nodes seems like a arbitrarily large number. I would refit the system to no more than 50 skill points.
  • Re-evaluate skill node placement. This goes without say. Certain trees seem very OP at the moment, such as the sensors tree that permits what would have been multiple 6 million cbill mech modules.
Overall, the skill tree is an interesting concept and, optimistically, has the potential to be a big step in the right direction. However, the initial implementation is a mess that seems to be creating a lot of uncertainty for many players, including myself. If the proposed system is executed without certain critical changes, I see myself significantly reducing the amount of time and money I spend on this game.

Edited by process, 10 February 2017 - 05:02 AM.


#2 Codpond

    Member

  • PipPip
  • Ace Of Spades
  • Ace Of Spades
  • 28 posts

Posted 09 February 2017 - 06:43 PM

Nice post. I will second that the skill tree promotes boating of weapons. Why level up multiple weapon systems when you can just boat one type and level your utility to max.


#3 process

    Member

  • PipPipPipPipPipPipPipPip
  • Star Colonel II
  • Star Colonel II
  • 1,667 posts

Posted 09 February 2017 - 09:26 PM

Added a suggestion about preserving mech rank, so it's not a mastered-or-nothing system. Vanity, I know.

#4 Kynesis

    Member

  • PipPipPipPipPipPip
  • 224 posts
  • LocationSydney

Posted 09 February 2017 - 09:41 PM

The entire structure is fundamentally flawed - the linked nodes and deep trees reward specialisation at the expense of diversity, which is a totally logical outcome but it's not the only one possible and it is definitely not the best.

The cost to spec, let-alone respec is absolutely prohibitive. Charitably, one could say that it gives us the option of investing in many mechs and build those mechs for specific a purpose or role or to only purchase one mech and spend that same amount (or more) re-gearing and re-speccing that one mech.

Less charitably it strong-arms players into buying mechs or premium time with cash rather than in-game currency.

PGI obviously needs to make money, to be sustainable but its core commitment needs to be making the best game possible.
This implementation very clearly puts the businesses interests ahead of its customers.

Edited by Kynesis, 09 February 2017 - 09:45 PM.


#5 0bstinate

    Rookie

  • Philanthropist
  • Philanthropist
  • 5 posts

Posted 09 February 2017 - 09:46 PM

100% agreed on most points, especially diminishing values. Almost every game with skills does this. Look to Eve for example: they also have a punishing skill grind, but at least the first few levels of each skill can be gotten relatively quickly.

You know the saying, "Easy to learn, but hard to master?" That is a good property in almost anything. The skill tree should reflect this.

#6 Skribs

    Member

  • PipPipPipPipPipPip
  • 465 posts

Posted 09 February 2017 - 10:07 PM

I agree that the system is way too expensive. Here is my suggestion:

1) Respecs should definitely be free.
2) Mech XP required to max a Mech should be 50k-75k. I already have too many Mechs that aren't leveled to start over on everything I have leveled. There should be 0 C-Bill cost to level a Mech. I already spend that on my next Mech.
3) Do you want everything to be MC or MXP? I agree, that there are too many currencies right now.
4) I agree about the skill tree placement.





5 user(s) are reading this topic

0 members, 5 guests, 0 anonymous users