

Kanajashi Hits The Nail On The Head.
#21
Posted 13 February 2017 - 03:05 PM
As for the video. Need to watch it when I get home. But I heard he has a good point about the survival scaling by class leaves the lowest weight of each class pretty screwed. Which I guess means they should scale by ton instead?
#22
Posted 13 February 2017 - 03:13 PM
But sadly, Since blanket changes are the PGI theme, mechs like the dire wolf will always be screwed because of the KDK-3 or NTG or TBR have to be "normalized"
also, inb4 someone mentions 8xUAC2 dire
#23
Posted 13 February 2017 - 03:29 PM
naterist, on 13 February 2017 - 03:04 PM, said:
9 games of FW will generally get you enough cbills to fully spec it out. sounds OK to me.
Except for the fact that CW/FW is a hot flipping mess right now with all the top teams on Clan side, and therefore nigh impossible for IS teams to actually win. Oh I know it happens, but it's only like, what, 1 to 10 out of 100 or more games or something else equally ridiculous?
If it was still open to IS vs IS where the tech imbalance wouldn't come into play cause the top teams would still be Clan side, then you might have a point, but then the other problem is splitting the player base into a dozen different buckets once again and how do you handle loyalists and mercs?
Well, for one thing, the Clans shouldn't have the option of mercs at all, they should essentially be the same as IS Loyalists.
You pick a faction.
AND STICK WITH IT!
No "Ifs", "Ands", or "Buts".
Period.
CW/FW should be the big, grand event that everyone wants to get into, so why doesn't everyone get into it?
Well the seal-clubbing by all the Tier 1 Ultra-Meta-Comp-Teams certainly doesn't help, but that's a problem with the Player Tier system to begin with.
Simple fact is there's no good way to fix such a completely screwed up system, which could be said of the game as a whole, it's so far gone is there any chance of saving it at all?
#24
Posted 13 February 2017 - 04:09 PM
I'm beginning to believe that I'm being gaslighted. I see a completely different PTS than most others in the community. I don't understand it. I tend to really like Kanajashi's videos. I think he's mistaken on this issue though.
I think we can agree that the grind has been increased significantly. The cost and time requirements are a bit much...
It is obvious, however, that the primary (yet unstated) design goal with the current PTS is increased time-to-kill (TTK). Practically everything points to and revolves around this. Cooldown has been reduced by 70%, heat dissipation by 33 1/3% and heat containment by 25%. Significant armor and structure "enhancements" have been included in the tree. The buffs from modules have been reduced to inconsequential levels. Crit chances and damage have been reduced on several weapons. If you look at all the change notes, almost everything has the direct effect of reducing damage output across the board and increasing 'Mech hit points; hence, increasing TTK. If increasing TTK is the primary objective, it would explain why all these changes were included in the PTS simultaneously. If we analyze the PTS with these metrics involved, we may see a very different picture.
There is little worth spending SP on in the Firepower branches that will make a noticeable difference in actual game play for a far majority of players. Let the actuaries and number crunchers deal with the spreadsheets. It appears that min/maxing this system is mostly a masterbatory proposition.
4-5% cooldown is a matter of a tenth or two of a second.
5% heat gen is a reduction of 1.8 on a 42 pt damage, 6 cER-MLas alpha that produces 36 heat. 1.8 savings on 36 heat?
10% range on the same cER-MLas is a matter of 40m on the 405m nominal max. Really?
How many of us are truly skilled enough to take advantage of these slight bonuses in real-time game play, where so many other factors come into play?
The 15-20% quirk-like enhancements may be worth looking at. But will a 20% increase in velocity encourage a KDK-3 pilot to boat AC or U/AC 5s if he doesn't already do so? (What else should one do with a KDK-3?) These bonuses are minimal compared to the quirks we are currently used to seeing on 'Mechs (which are being significantly reduced or eliminated). There are now 'Mechs with 50% (ER)PPC velocity quirks, so why worry about a 20% enhancement. The significantly reduced heat management is enough to discourage energy boats as it is. Heck, my SCR, ACH, HBR and HBK IIC-A will not be the performers they currently are, as they ONLY have energy hardpoints. Today's energy boaters will feel encouraged to diversify their builds, if they can, with cooler weapon systems simply to get heat under control, while trying to maintain DPS.
I still contend that "meta tryhards" will be the biggest losers under the proposed system. There is basically nothing in the Firepower branch that is "must have" to effectively pilot a 'Mech. In fact, one can argue that our SP are better spent elsewhere on the tree. Under many circumstances, assigning SP to the Firepower branches may put us at an overall disadvantage. If Firepower nodes are as weak as they appear, why assign any at all? This would allow us to experiment with our builds, with little to no consequences.
As I see it, the current PTS accomplishes these things:
-Significantly increases the grind...
-Significantly increases TTK.
-Allows pilots to lightly customize their 'Mech in ways we couldn't before.
-Slightly encourages diverse builds, while slightly discouraging boating.
-Takes small steps toward further chassis/tech balance.
-Achieves most of what the ED system was after in a more elegant, less intrusive way.
Are there other little tweaks that could be made? Sure, there are. But with the exception of the increased grind, I think we can all live with the above changes. We may, indeed, discover new favorite 'Mechs. Isn't that kind of exciting?
I would like to see what PGI comes back with next, on a PTS with 12v12 games. This is not a minor addition to the game. It should be mostly wrung out before it goes live. I'd also like to see more plots playing the PTS.
I wish someone could show me where I'm completely off with my analysis. I read through much of the feedback here and don't get where many pilots are coming from. I don't normally have much to say on these kinds of forums. But I'm really trying to understand, believe me. I'm simply at a loss...
Edited by Spunkmaster, 13 February 2017 - 04:34 PM.
#25
Posted 13 February 2017 - 04:39 PM
naterist, on 13 February 2017 - 03:04 PM, said:
9 games of FW will generally get you enough cbills to fully spec it out. sounds OK to me.
Only if you win all 9 matchs in huge landslide. When you lose, the bonus cbill from the faction is not applied since you lost.
Not to mention the waiting time for FW is far greater than QP. Solo players have far better chance and get better rewards to just play QP instead of FW where solo players vs group happens, resulting in far less profit for far longer time.
#26
Posted 13 February 2017 - 05:11 PM
One thing he pointed out was how the trees don't promote diversity of weapons on a mech. This is an obvious conclusion tons of other people came to, but his suggestions here still fall short.
I do completely agree with his assessment on respec'ing a mech and the costs associated with it, and how players will feel about it. If the trees fail miserably with design goals, it's the areas of weapon diversity and build diversity being promoted. Also, not to mention time taken to fill out the skills of a mech.
#27
Posted 13 February 2017 - 06:35 PM
The 9.1 mil is a pretty big cost required to master on the PTS.
But I agree with Kanajashi that there are way too many skill points.
It's too easy to get everything we had for 50 skill points, then add on effectively 8 modules worth.
He suggested 70 points which effectively gives us the current mech tree plus 4 modules which would not really a change from what we have now in terms of how powerful the mechs are.
I would suggest less, 60 points or lower.
Not only does this really force some choices in terms of which skills take which is better for balance, but it also reduces the cost to 'complete' a mech using the skill trees.
If we look just at the cost to get a single node, it's probably about 1 or 2 battles to be able to afford it with c-bills. Maybe a few extra to get the XP needed.
Having the opposing trees for the weapons seems a reasonable approach. It was raised some time ago that the modules should have a negative. You want range? Then you sacrifice duration. Same idea and it makes sense, there needs to be a trade off.
With regards to the ECM nodes, it would be interesting to measure that effectiveness vs mechs without any sensor boosts. However because we have so many skill points, it's unlikely players will not be picking up nodes from the tree as they seek out Radar Deprivation and Seismic Sensors.
I would also suggest that the skills where we have only 1 or 2 nodes to complete them do need to be broken up and given 5 levels so it is consistent. There are also a few missing skills that could be added such as skills for the Active Probe. (Unless I missed that somewhere). We could also expand on some of the options and I think there is a great opportunity to move the set of 8 quirks and provide 'lore' bonuses for weaker mechs in their own unique tree.
For example: The Atlas is traditionally seen as a very tough mech. Why not have a 'lore' tree as a separate option that players can devote points to so it combines with the defensive tree and emphasizes this aspect for players that choose to do so.
I think that's an important point actually, it should always be our choice.
Anyway...
More skill options + less points to spread around creates diversity and therefore different specializations/roles.
Not sure about scaling the bonus as if the initial bonus is too high, everyone ends up back in the same situation by all taking the initial skill but ignoring the rest. It's just the same power creep but a bit more subtle. Yes it reduces the gap from new to old players, but if there are less points overall, a player who has mastered a mech will only be better in some areas as they simply can't get everything. If they haven't and have gone for a more general approach of a bit here a bit there, then the gap isn't that big anyhow.
Otherwise, solid ideas and a very good analysis.
#28
Posted 13 February 2017 - 08:04 PM
Spunkmaster, on 13 February 2017 - 04:09 PM, said:
I'm beginning to believe that I'm being gaslighted. I see a completely different PTS than most others in the community. I don't understand it. I tend to really like Kanajashi's videos. I think he's mistaken on this issue though.
I think we can agree that the grind has been increased significantly. The cost and time requirements are a bit much...
It is obvious, however, that the primary (yet unstated) design goal with the current PTS is increased time-to-kill (TTK). Practically everything points to and revolves around this. Cooldown has been reduced by 70%, heat dissipation by 33 1/3% and heat containment by 25%. Significant armor and structure "enhancements" have been included in the tree. The buffs from modules have been reduced to inconsequential levels. Crit chances and damage have been reduced on several weapons. If you look at all the change notes, almost everything has the direct effect of reducing damage output across the board and increasing 'Mech hit points; hence, increasing TTK. If increasing TTK is the primary objective, it would explain why all these changes were included in the PTS simultaneously. If we analyze the PTS with these metrics involved, we may see a very different picture.
There is little worth spending SP on in the Firepower branches that will make a noticeable difference in actual game play for a far majority of players. Let the actuaries and number crunchers deal with the spreadsheets. It appears that min/maxing this system is mostly a masterbatory proposition.
4-5% cooldown is a matter of a tenth or two of a second.
5% heat gen is a reduction of 1.8 on a 42 pt damage, 6 cER-MLas alpha that produces 36 heat. 1.8 savings on 36 heat?
10% range on the same cER-MLas is a matter of 40m on the 405m nominal max. Really?
How many of us are truly skilled enough to take advantage of these slight bonuses in real-time game play, where so many other factors come into play?
The 15-20% quirk-like enhancements may be worth looking at. But will a 20% increase in velocity encourage a KDK-3 pilot to boat AC or U/AC 5s if he doesn't already do so? (What else should one do with a KDK-3?) These bonuses are minimal compared to the quirks we are currently used to seeing on 'Mechs (which are being significantly reduced or eliminated). There are now 'Mechs with 50% (ER)PPC velocity quirks, so why worry about a 20% enhancement. The significantly reduced heat management is enough to discourage energy boats as it is. Heck, my SCR, ACH, HBR and HBK IIC-A will not be the performers they currently are, as they ONLY have energy hardpoints. Today's energy boaters will feel encouraged to diversify their builds, if they can, with cooler weapon systems simply to get heat under control, while trying to maintain DPS.
I still contend that "meta tryhards" will be the biggest losers under the proposed system. There is basically nothing in the Firepower branch that is "must have" to effectively pilot a 'Mech. In fact, one can argue that our SP are better spent elsewhere on the tree. Under many circumstances, assigning SP to the Firepower branches may put us at an overall disadvantage. If Firepower nodes are as weak as they appear, why assign any at all? This would allow us to experiment with our builds, with little to no consequences.
As I see it, the current PTS accomplishes these things:
-Significantly increases the grind...
-Significantly increases TTK.
-Allows pilots to lightly customize their 'Mech in ways we couldn't before.
-Slightly encourages diverse builds, while slightly discouraging boating.
-Takes small steps toward further chassis/tech balance.
-Achieves most of what the ED system was after in a more elegant, less intrusive way.
Are there other little tweaks that could be made? Sure, there are. But with the exception of the increased grind, I think we can all live with the above changes. We may, indeed, discover new favorite 'Mechs. Isn't that kind of exciting?
I would like to see what PGI comes back with next, on a PTS with 12v12 games. This is not a minor addition to the game. It should be mostly wrung out before it goes live. I'd also like to see more plots playing the PTS.
I wish someone could show me where I'm completely off with my analysis. I read through much of the feedback here and don't get where many pilots are coming from. I don't normally have much to say on these kinds of forums. But I'm really trying to understand, believe me. I'm simply at a loss...
If their goal is to slow down TTK then they failed. Easiest way to slow down TTK is to slow down weapon fire. 10% reduction? Have the weapons fire 10% slower. Heck go 15% slower due to rounding some of the faster firing weapons.
Don't make mechs run hotter which disproportionately hurts energy mechs.
Don't apply a blanket boost to armor which skews the balance between all of the mech classes.
Just slow down the rates of fire. Solved.
#30
Posted 13 February 2017 - 09:12 PM
Ruar, on 13 February 2017 - 08:04 PM, said:
If their goal is to slow down TTK then they failed. Easiest way to slow down TTK is to slow down weapon fire. 10% reduction? Have the weapons fire 10% slower. Heck go 15% slower due to rounding some of the faster firing weapons.
Don't make mechs run hotter which disproportionately hurts energy mechs.
Don't apply a blanket boost to armor which skews the balance between all of the mech classes.
Just slow down the rates of fire. Solved.
It's far more complicated than that. Everything needs to be brought back together. Too much has creeped over the years. Notice that I never said that TTK was the only design goal for this patch, only the primary. The skill tree brings other goals to the mix. Simply slowing cooldown would not address the issue of 80+ point alphas. Bringing heat management back into reasonable levels does. Crit chance and damage has become more of an issue the past couple of months. Cooldown nerfs don't address that either.
What balance is disturbed by the armor/structure boost? The step between top of one class and the bottom of the next? I don't find that such a big deal. A blanket boost to all armor, regardless of weight class, would cause a larger imbalance than tiering the classes, so lights get a bigger bump than assaults. It doesn't hinder the 80 tonners, for instance. They still tend to be faster and more mobile than their big brothers. Besides, a fluid, sliding scale would fix that simply.
Because the game is too strongly rooted in a table top game (which I love to play), it will never be perfectly balanced, unless it is scrapped and rebuilt from scratch. The best that can be done is to get as close as possible. This or something similar may do just that.
As I've mentioned before, I'm interested in seeing what PGI comes back with.
Edited by Spunkmaster, 13 February 2017 - 09:15 PM.
#31
Posted 13 February 2017 - 09:33 PM
Spunkmaster, on 13 February 2017 - 09:12 PM, said:
It's far more complicated than that. Everything needs to be brought back together. Too much has creeped over the years. Notice that I never said that TTK was the only design goal for this patch, only the primary. The skill tree brings other goals to the mix. Simply slowing cooldown would not address the issue of 80+ point alphas. Bringing heat management back into reasonable levels does. Crit chance and damage has become more of an issue the past couple of months. Cooldown nerfs don't address that either.
What balance is disturbed by the armor/structure boost? The step between top of one class and the bottom of the next? I don't find that such a big deal. A blanket boost to all armor, regardless of weight class, would cause a larger imbalance than tiering the classes, so lights get a bigger bump than assaults. It doesn't hinder the 80 tonners, for instance. They still tend to be faster and more mobile than their big brothers. Besides, a fluid, sliding scale would fix that simply.
Because the game is too strongly rooted in a table top game (which I love to play), it will never be perfectly balanced, unless it is scrapped and rebuilt from scratch. The best that can be done is to get as close as possible. This or something similar may do just that.
As I've mentioned before, I'm interested in seeing what PGI comes back with.
I will concede that slowing down RoF does nothing for large alpha builds. I never play assaults so I don't think in terms of more than 30ish points of damage per alpha. Seems to me if alphas are a problem then they could kick up the ghost heat a bit. A better option then a blanket heat increase which hits the mechs who don't have issues along with the ones that do.
Armor balance is a significant issue when you consider light mechs being harder to kill. Again, not much of an issue with high alpha builds but a problem when in a medium mech.
I think the solution requires an individual approach to mech balance instead of a tonnage balance. I think quirks are the balance solution even though a lot of people don't like them. I also think they need to make more weapons interchangeable similar to how the HBK swaps out the AC for lasers in the hunch. Why can't missiles swap out for a laser pack? Limit the weapons that can go there, but by having interchangeable weapons packs it reduces the impact of poorly placed weapon positions.
They don't have to rebuild from scratch, but a one size fits all answer won't work either. The skill tree makes strong mechs stronger and mediocre mechs still mediocre.
What the skill tree does a great job of doing is bringing some balance to modules and I hope they implement a module only version. But if you really want balance the answer is work with quirks on a mech by mech basis. Ideally you'd have a list of approved quirks for that mech and you pick specific quirks to fit your build. This would prevent mechs like the KDK from getting stronger and would boost mechs like the Atlas.
Honestly, if they could do something similar to the skill tree but limit options for each mech it might work, but I think quirks would be simpler.
#32
Posted 13 February 2017 - 10:50 PM
Ruar, on 13 February 2017 - 09:33 PM, said:
I will concede that slowing down RoF does nothing for large alpha builds. I never play assaults so I don't think in terms of more than 30ish points of damage per alpha. Seems to me if alphas are a problem then they could kick up the ghost heat a bit. A better option then a blanket heat increase which hits the mechs who don't have issues along with the ones that do.
Armor balance is a significant issue when you consider light mechs being harder to kill. Again, not much of an issue with high alpha builds but a problem when in a medium mech.
I think the solution requires an individual approach to mech balance instead of a tonnage balance. I think quirks are the balance solution even though a lot of people don't like them. I also think they need to make more weapons interchangeable similar to how the HBK swaps out the AC for lasers in the hunch. Why can't missiles swap out for a laser pack? Limit the weapons that can go there, but by having interchangeable weapons packs it reduces the impact of poorly placed weapon positions.
They don't have to rebuild from scratch, but a one size fits all answer won't work either. The skill tree makes strong mechs stronger and mediocre mechs still mediocre.
What the skill tree does a great job of doing is bringing some balance to modules and I hope they implement a module only version. But if you really want balance the answer is work with quirks on a mech by mech basis. Ideally you'd have a list of approved quirks for that mech and you pick specific quirks to fit your build. This would prevent mechs like the KDK from getting stronger and would boost mechs like the Atlas.
Honestly, if they could do something similar to the skill tree but limit options for each mech it might work, but I think quirks would be simpler.
I agree with you on much of this. I don't own an assault either. Chassis should be balanced appropriately on base stat or quirk level, not the skill tree. I think that's what's starting to really happen in the latest PTS. Most, if not all, of the defensive quirks remain, if not slightly reduced. Maybe PGI is telling us they think they have the base chassis balanced?
I've always believed the hard point should come with the weapon. If there is appropriate tonnage and slots, then mount it. All other rules apply.
I think it's best to keep the same tree for all "Mechs, unless we break it down by variant and even build. Not doing that could very easily discourage variety of builds. I think PGI is still balancing the damage nerf with the defensive buff. I think that we'll eventually see damage a bit higher or hit points a bit lower or a combination of both. Besides, a KDK is supposed to be a frggen beast. A medium was never supposed to be able to 1v1 a KDK; yet, with greater speed and mobility, sometimes it can.
As for energy being hit especially hard... I've always believed we should be designing our energy builds with a 1.35 or higher heat management. Now, our 1.2-1.30 builds are going to pay for it. A big change in "meta" and tactics could be approaching.
Edited by Spunkmaster, 13 February 2017 - 11:08 PM.
#33
Posted 13 February 2017 - 11:15 PM
It's obvious PGI doesn't play or care about their own game. They just go there to work so they do something until the end of the day. Look at all this work we've done, this big tree! They're just doing something just for the sake of doing it, no matter how bad it turns out as long as you put your hours into something. They should just let Kanna design this whole thing would turn out a lot better, at least he has the passion for it and that shows. Whoever designed this thing has no passion or any idea. It's all in the video, freaking gold mine for PGI if they do it like this.
Edited by NeoCodex, 13 February 2017 - 11:21 PM.
#34
Posted 14 February 2017 - 12:01 AM
Primarily, he brings up an excellent point about the armor/structure. A Griffin (55tons) under this system has a total of 444 armor max, and a Mad Dog ends up with 462. An 18 point difference (around half a tons worth of armor?) for 5 tons. So I'm in agreement with his static percentage for armor as well as the static percentage for ammo increases.
I'm also in agreement with the incremental increases, which allows a quicker jump to parity for newer players/mechs. His point on ECM is also well thought out.
I'm not sure I like the 'choose 1 of 2' kind of limitations. If a mech wants to focus on guns - I think that's great. Let them get range and durability - but they'll need to sacrifice in other areas, like mobility / armor / sensors.
#35
Posted 14 February 2017 - 06:22 AM
Spunkmaster, on 13 February 2017 - 10:50 PM, said:
I agree with you on much of this. I don't own an assault either. Chassis should be balanced appropriately on base stat or quirk level, not the skill tree. I think that's what's starting to really happen in the latest PTS. Most, if not all, of the defensive quirks remain, if not slightly reduced. Maybe PGI is telling us they think they have the base chassis balanced?
I've always believed the hard point should come with the weapon. If there is appropriate tonnage and slots, then mount it. All other rules apply.
I think it's best to keep the same tree for all "Mechs, unless we break it down by variant and even build. Not doing that could very easily discourage variety of builds. I think PGI is still balancing the damage nerf with the defensive buff. I think that we'll eventually see damage a bit higher or hit points a bit lower or a combination of both. Besides, a KDK is supposed to be a frggen beast. A medium was never supposed to be able to 1v1 a KDK; yet, with greater speed and mobility, sometimes it can.
As for energy being hit especially hard... I've always believed we should be designing our energy builds with a 1.35 or higher heat management. Now, our 1.2-1.30 builds are going to pay for it. A big change in "meta" and tactics could be approaching.
The problem with hardpoint = weapon is the lack of variety in most builds. Yes, the builds are based on lore but in TT weapon position doesn't matter. So each variation of a mech has very little that changes. Perhaps the main weapon gets altered but the support weapons are left alone. Sometimes the support weapons change slightly while the main weapon stays.
If hardpoint = weapon then each variant would have to be unique. The difference between Enforcers shouldn't be whether or not you take an LBX, AC, or UAC. Because picking one of three autocannons isn't variety. There should be a missile build, a short range build, a long range build. We don't have that though because of lore. Which is why I would prefer to see the ability to swap out weapon packs on the hardpoints so we can see variety. 4R would be the laser build, 4M would be missile options, 4P would be a brawler with medium lasers and SRMs.
I would actually prefer a damage nerf simply because the conversion from TT to arcade doesn't take into consideration pinpoint aiming. That fact alone changes balance for every weapon system. And I'm not saying the KDK should be nerfed, but it also shouldn't be head and shoulders above the other assaults. Typically assaults are heavy armor, heavy weapons, weak mobility. The KDK is a balance of all three though which is bad because the amount it gives up on firepower and armor is completely balanced by mobility. Which is a problem due to the conversion from TT to arcade.
I already build my energy mechs with a 1.5 heat ratio or better. I don't see the point in having so many weapons that I have to shutdown or stay hidden the bulk of the fight. If heat goes up anymore then I'm going to be forced to drop 30% of my weaponry in order to keep up my rate of fire. There's just not that room for adjustments on a medium mech. Which is why I think ghost heat is a better solution because it forces the mechs carrying a lot of weapons to make decisions instead of the mechs with only a few weapons without the tonnage, or space, for additional heat sinks.
#36
Posted 14 February 2017 - 07:16 AM
#37
Posted 14 February 2017 - 07:17 AM
Only the jump in structure bonus between weight classes is not as obvious as to bee seen in multiple thread already.
This doesn’t make the points less valid, ofc, and it’s good to see them spoken out on a prominent place. I myself was looking forward to his his vids on the topic since PTS start.
I don’t agree with his strong opinion on RNG crits though. He may have his personal aversion against anything random, nothing wrong with that, but it makes no sense him saying “it makes no sense”.
I also think he is exaggerating the problem with the ECM change a bit.
Concerning his solutions, they are (for the important problems):
- Weapon-independent bonuses for most of the weapon skills with weapon-specific ones in the corners of the tree
- Mutually excluding trees for range, duration, CD etc.
- Diminishing rewards
Number 2 sounds quite stiff. I like the general idea (and already read it on the forum, surprise) but I imagine it possible to be designed more flexible and still encourage choosing one tree over an entire branch.
Number 1 doesn’t fit with the rest of the suggestions and is kind of stand alone, or did I get it wrong?
Edited by Kuaron, 14 February 2017 - 08:33 AM.
#38
Posted 14 February 2017 - 11:00 AM
I do not think that ECM is currently over powered.
I also do not like the tier system based on class, it will force any mech with tonnage at the top of its weight class to be OP, while mechs at the bottom of a weight class will be useless.
Edited by BigScwerl, 14 February 2017 - 11:02 AM.
#39
Posted 14 February 2017 - 01:25 PM
Thanks for being sympathetic - I generally get in about 10-12 games a week so it sounds f*cking terrible to me
It'll take me at least five years to skill up the mechs I've got now and gives me zero reason to buy more
#40
Posted 14 February 2017 - 02:21 PM
DAYLEET, on 13 February 2017 - 01:12 PM, said:
It sure is when you're on mobile and you have to shift thru three keyboard pages to find the right / [ ] with fat thumbs to finish the quote tags. But thanks fo the condecension.
Quote
Well now you can add someone who owns 30 mechs and has had to sell mechs in the past to participate in all facets of the game to that list that thinks differently. But let me be clear, I'm not saying, nor have I said, that current players losing their mastered mechs is a good idea or that players like Gas Guzzler should be forced to spend billions of cbills and thousands of exp to earn back under a new system what they earned under the old. But rather the new skill tree is superior for new players over the current rule of three. Its cheaper, it allows more options, less waste, and a path to participate in all facets of the game with the starter 4 mech bays.
Quote
It's worse for new players just like everyone else. It's worse for everyone because the system now is designed so you DON'T have to pay for the modules more than once, you pay for the mech the gear and the mechbays. In the new system, everyone if forced to pay 19millions for all the mech that they want. It's insane. The cost don't need to be tuned down, or just for new mech not previously owned, but removed entirely. Remember how new players started with medium and light mech because they were cheaper? AHAH now they ALL cost more than an 100ton Clanner ffs.
Right now only module cost cbills. After PTS EVERY skill will cost cbills. A newplayer could skip modules in this system and be at a disadventage. In the next system he will have to skip everything, this will go well with the complete lack of tiering since there isnt enough people to make it happen. This will be the golden age of even more less people.
So they're shifting the cbill cost to skills. And? There always was a cost in exp and cbills. Its just he cbills were sunk in mechs you may or may not of wanted. That you could save some money moving modules around doesn't make this system more expensive for everyone.
The cost is going down for new players allowing them a bigger window to decide if they even want to play this game or sink real money into it. If you're a new guy that played MW2 back in the day and the Timber is THE iconic mech you remember and want to get you're spending 32 million on extra varients to be "mastered" that you may or may not keep. Say you sell them cause you mastered your Timby and found the playstyle on one you like and want to move on to something else, you get half that back(actually less if you keep the omnipods but what ever), 16million cbills back. You just pissed away 16million cbills to try something else, vs 9.1 million? That math seems pretty simple to Mr. BayStrapped NewDickerson. Add in that that same 9.1 million gets you the modules and puts you on par with the establish players stats wise. That would cost 18million on top of the same 16 million thats just gone. But you can save money swapping modules! So we're arguing to keep a 16million tax for every chassis to be mastered cause we can pay 18 million for modules once? The hell? Verses a 9.1 million tax on every mech? Its more if you buy every damn mech in the game, yeah. But its a hell of alot cheaper and easier for those that don't. IE new players
But everyone knows you don't sell mechs! Like hell, if I'm deciding between fake digital money and real money for mech bays I'm spending the fake money. Its a decision I've made in the past when I was on the fence with this game, I'm sure its a decision others have made just the same. Even with free MC from events and waiting for Christmas for 50% Mech Bay sales to be frugal one isn't exactly swimming in Mech Bays. God help you if you chose the wrong mech, not cause it was bad, but because PGI brought out the nerf bat.
Gas Guzzler, on 13 February 2017 - 01:25 PM, said:
Lose SOME of our mastered mechs? More like MOST. Anyone who has a lot of mechs (I have 303) is going to be wrecked by the costs. I don't think I will have enough C-bills to master even 25% of my mechs, and as someone who has spent a ridiculous amount of money buying mechs and being an active player with premium time, for years, it is deeply insulting and frustrating. It means I'm guaranteed to not bother playing unskilled mechs that I only play once in a while to break up the monotony, and that's a shame.
Did you just stop at the sentence where I said some? Literally the next sentence in the quote is:
Quote
And well I agree with you and have tailored my response to the PTS to show that. Shafting established and loyal customers is unacceptable. That we're sorta implicetly ok with shafting new players with the current crumy system is just as unacceptable.
I_AM_ZUUL, on 13 February 2017 - 02:33 PM, said:
THE RULE OF THREE does that... so complain about how crappy it is (which for new players it is) instead of the removal of an artificial placeholder as a benefit for the dumpster fire that is this PTS.
I have complained, hell I've voted with my feet before and taken long extended breaks from this game w/ 0 money going into PGIs pocket. Its been, what, 4 years? with the rule of three. Excuse me for pointing out when something is an objective upgrade over the trash we have for new players.
9.1 million for 4 different mastered mechs(36.4 million total) is light years better than Buying three chasis, keeping one, buying three more, keeping those three(I sure hope I dont get shafted next patch meta shift!) for 4 mastered mechs. 4 sets of modules alone for your Mr. BayStrapped NewDickerson player is 72 million. Ignoring the cost in lost mechs. 36.4 vs 72 for the same performance on paper.
So we have a system where Mr. BayStrapped NewDickerson can compete, on paper at least, in every facet of the game on a even footing if planned right for 36.4 million "tax" vs one where they can do that now for 72 million+what ever they pissed away selling mechs to make space "tax"? How is one not objectively better?
My response is "Yeah this is better than what we had before. For new players. You need to fix it for established players." Not "Burn the whole damn thing, we're keeping the crap we have!" Which is what I'm seeing from people saying they'll leave the game if the new system goes live. Its not constructive, its not helpful.
Edited by LordLeto, 14 February 2017 - 02:22 PM.
4 user(s) are reading this topic
0 members, 4 guests, 0 anonymous users