Jump to content

Practical Solutions To Many Of The Core Problems With This Game (Especially In Fp)


57 replies to this topic

#21 LordNothing

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 17,620 posts

Posted 10 October 2017 - 08:55 PM

View PostHobbles v, on 10 October 2017 - 08:03 PM, said:


1. Separate queue was attempted, and was a dismal failure. Also it does nothing to teach new pilots about whats good in this game. So often I come across terrible builds from pugs, and even chat with pugs who love their garbage builds with 4 different weapon systems because the only basis of comparison they had was the stock build they replaced. If mechs came built with good practices in mind these pugs would be far more likely to realize how much their performance dips when they change their build to a bracket build with a weapon for every range.

2. I enjoy using the skill tree, because I am heavily into theory crafting and can happily spend hours tweaking that kind of stuff. The average player does not, and for a new player it's just too complex. I've seen a lot of pilots quit because they didn't want to learn it, seen others who have no idea what to do with their 4 types of currency involved and other new pilots struggle who have too steep a learning curve already without having to learn this mess.

3. one bucket is most certainly a problem right now. The IS side is giving up because there is no hope to tag planets anymore, previous iterations of FP had defense tags. Now tied phases result in no tags being given out. Since IS almost never wins a phase, no IS units gain tags anymore. When one side repeatedly loses all the time, that side will keep losing players. A small number of multiple queues actually gives hope that an IS team could stack one and pull off a victory somewhere. I use 4 planets as an example because 4 is the number of planets currently awarded each phase. If 4 buckets is too much 2 or 3 could work. we just need a system where both sides can make progress in some measurable way.

4. This post is about Practical solutions. Things that do not require a massive re-design of the mode to implement. Currently there is a faction imbalance, Clans have 1 more than IS. The simplest and most practical way to even this out is to add one to the IS.... Since PGI definitely wont remove any (unless we kill off smoke jag in a bulldog event).

Your suggestion of adding a two tree system is flawed, since it does not address the main point i was trying to make about loyalty trees; they dry up upon completion. The ends of trees need to have repeatable rewards or some kind of persistent progress. Something that doesn't lead a player having to bail on their current faction, or just stop playing if their universal tree stops rewarding them.

Actually just had another Idea that may be the simplest for PGI. Once someone maxes out their Loyalty tree, they start over. Loyalty is rewarded with repeated access to the easy to earn levels. instead of having to Desert for a week and move to a new faction.



I would rather just have them pay the cost of a properly equipped mech up front when they buy it. This counters the imbalance of IS mechs being initially cheaper. And a properly equipped mech gives new players an example of what works. Quirking up a stock in specific ways to try and make it better teaches them nothing and tries to prop up garbage.


1

separate queues were not tested. the way it went down could not be called a test. they were up for 2 days during an event, and then taken down before another event started, and during the scouting mode debut. scouting queues were bloated and thats where the bulk of the players were. while this was going on anyone who did play invasion had the long tom to contend with. so anyone with any negative preconceptions going into a match with long tom against them would not return. this was not a fair test. it was less science and more like trying to test a pogo stick on the deck of the titanic while it was sinking, then trying to put fourth a definitive conclusion.

second i didnt say a damn thing about a second queue. freelancers would still slotted into normal games that meet certain requirements, with the exception of when there are enough freelancers to fill a team on both sides. otherwise theres no damn reason to have freelancer.

new players are going to have crappy mechs and trials, theres no way around that problem. you cant train for fp in the testing grounds, the training area or in qp. to learn to play fp you must play fp and gain experience. having a kiddie pool is a good idea. you dont have to split the queues to do that. you just have to be slightly more selective about where noobs and casuals get placed.

2

i dont know whether the average player likes the skill trees or not and neither do you. i thought it was one of the more positive changes that pgi made to the game. the forum was very polarized. id need to see a poll to know what 'the average player' thinks.

3

again population is everything. you cant fix population by adding buckets. you cant even fix it by taking them away. the only way to get people back is to make it a mode people want to play. its a chicken/egg type problem. this puts this one at a very low priority since you need to introduce a 3rd thing to break the stalemate.

four buckets was my ideal solution because it would have great population balancing capabilities. but 3 might be viable. clans would have trials and is would have wars among the houses, and then the clan is war. this two might have load balancing abilities. if you dont get to choose your bucket then when you go to fight if there is any population disparity you can just fold the players on the over populated side into itself and create is v is or clan v clan game. of course this creates situations where one side cant fight a civil war type scenario if they dont have the majority. but thats not a huge problem.

4

i dont think the two tree system is that bad of an idea. one tree provides incentive to play fp, the other gives you incentive to stay with your faction. both are incentives that fp needs. the rewards tree would keep paying out even if you complete it. the faction rank tree becomes a money multiplier that directly affects your mission payouts. its also a cut and paste job so its totally practical. they add a tree every time they add a faction, and they have been doing that a lot lately. its a cut and paste job otherwise they wouldnt be doing it. you could just as easily roll all rewards into a single tree and then have a rank tracker which is more or less the same but without goodies. sounds quite practical.

#22 LordNothing

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 17,620 posts

Posted 10 October 2017 - 09:07 PM

View PostKwea, on 10 October 2017 - 08:08 PM, said:

I think having stock IS builds be better is a HUGE deal, and probably causes a lot of the horribad new players on the IS side. Also, sometimes it is 5 mill to buy the mech, and 1 mill to fit it out.


maybe faster iteration on the community designed trial varients would help. problem is you put a noob in a meta build and they have 6 points of armor in the back an experienced player can just come in behind them and take them down. you need to practice dealing with back stabbers before you can really crank down on the back armor. heat management is another area new players get hung up on. it takes practice to use those 80 point alphas without killing yourself and that is a very advanced skill. im still terrible at it, which is why i favor ballistics. you are better off giving noobs lbs and mrms so they can trash objective while the rest of the team deals with mechs.

#23 Hobbles v

    Member

  • PipPipPipPipPipPip
  • The Sureshot
  • The Sureshot
  • 354 posts

Posted 10 October 2017 - 09:22 PM

View PostLordNothing, on 10 October 2017 - 09:07 PM, said:


maybe faster iteration on the community designed trial varients would help. problem is you put a noob in a meta build and they have 6 points of armor in the back an experienced player can just come in behind them and take them down. you need to practice dealing with back stabbers before you can really crank down on the back armor. heat management is another area new players get hung up on. it takes practice to use those 80 point alphas without killing yourself and that is a very advanced skill. im still terrible at it, which is why i favor ballistics. you are better off giving noobs lbs and mrms so they can trash objective while the rest of the team deals with mechs.


They dont need to be meta. They need to not be 4 different weapons with different weapon ranges and velocities and not run at 50% of max armour.

I would rather a noob be in an alpha vomit mech learning to manage heat than trying to manage 4 weapon groups. I would rather they learn that they need to protect thier backsides instead of running a 60/40 front back split on thier armour and thus lose every single trade from the front.

#24 McGoat

    Banned -Cheating

  • PipPipPipPipPipPipPip
  • CS 2019 Gold Champ
  • CS 2019 Gold Champ
  • 629 posts

Posted 11 October 2017 - 05:06 AM

These topics with solutions to the community seen problems, not just bitching, need to be bumped every day. This is the first one that didn't have people jabbing back and forth before the first page ended - great post, Hobbs

#25 Nightbird

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 7,518 posts

Posted 11 October 2017 - 06:06 AM

Keep in mind also changing stock builds is not free, the ugrades get rolled into the price. If a better build requies an XL or LFE engine, prices can easily double whereas now you can reuse the same engines. Even a DHS upgrade means nothing compared with that.

#26 Hobbles v

    Member

  • PipPipPipPipPipPip
  • The Sureshot
  • The Sureshot
  • 354 posts

Posted 11 October 2017 - 06:12 AM

View PostNightbird, on 11 October 2017 - 06:06 AM, said:

Keep in mind also changing stock builds is not free, the ugrades get rolled into the price. If a better build requies an XL or LFE engine, prices can easily double whereas now you can reuse the same engines. Even a DHS upgrade means nothing compared with that.


That is precisely a result i want. The mech should already be servicable when they buy it and not require them to spend another 6 million in the mechlab just to have somethimg worth a damn.

#27 Nightbird

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 7,518 posts

Posted 11 October 2017 - 06:40 AM

Yep, but other people missed that. I think many want the upgraded equipment at the same cost. They're gonna be shocked at the price of a Raven :)

#28 Aldodrem

    Member

  • PipPipPipPipPip
  • Bridesmaid
  • Bridesmaid
  • 100 posts

Posted 14 October 2017 - 12:04 PM

View PostNightbird, on 11 October 2017 - 06:06 AM, said:

Keep in mind also changing stock builds is not free, the ugrades get rolled into the price. If a better build requies an XL or LFE engine, prices can easily double whereas now you can reuse the same engines. Even a DHS upgrade means nothing compared with that.

Especially evident in the lower tonnage Mechs. My favorite mech to play is Locust 1V/3V which cost 1.3 Million base. Anyone who plays Locust knows DBL HS+Ferro+Max XL engine is a must. Now the Locust price rose to an astounding 6.2 Million!

Over a 475% Base Price Increase!!!
Imagine you went to your auto dealer and wanted a new car with all the options, power steering, windows, sunroof, park assist, chrome etc. and the base of that $32,000 dollar car now with all options costs you $152,000

For the fun of it go to an auto maker's site and use their build your car app thing they have and check the base price vs the model you made with all the options added in, does it come close to near 500% markup? The highest I got was 285% with a limited Edition Pickup version compared to the standard base model Pickup but interesting enough when I took the time to upgrade the base model with all the added extra it was in fact a few hundred dollars cheaper then the AIO Limited edition model!.

The pricing here is batshit crazy. It should scale with the tonnage of the Mech so while adding an extra 3 Million to a Commando for max upgrades would be insane as it is that 3 Million for say a King-Crab would be more reasonable. idk how they could make that work since engines are universal across all mechs but the DBL HS cost+artemiss cost should most certainly scale with tonnage like the armour upgrades does (those need to be toned down on the tonnage as well!).

Edited by Aldodrem, 14 October 2017 - 12:35 PM.


#29 50 50

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,145 posts
  • LocationTo Nova or not to Nova. That is the question.

Posted 17 October 2017 - 04:27 PM

View PostAldodrem, on 14 October 2017 - 12:04 PM, said:

Especially evident in the lower tonnage Mechs. My favorite mech to play is Locust 1V/3V which cost 1.3 Million base. Anyone who plays Locust knows DBL HS+Ferro+Max XL engine is a must. Now the Locust price rose to an astounding 6.2 Million!

Over a 475% Base Price Increase!!!
Imagine you went to your auto dealer and wanted a new car with all the options, power steering, windows, sunroof, park assist, chrome etc. and the base of that $32,000 dollar car now with all options costs you $152,000

For the fun of it go to an auto maker's site and use their build your car app thing they have and check the base price vs the model you made with all the options added in, does it come close to near 500% markup? The highest I got was 285% with a limited Edition Pickup version compared to the standard base model Pickup but interesting enough when I took the time to upgrade the base model with all the added extra it was in fact a few hundred dollars cheaper then the AIO Limited edition model!.

The pricing here is batshit crazy. It should scale with the tonnage of the Mech so while adding an extra 3 Million to a Commando for max upgrades would be insane as it is that 3 Million for say a King-Crab would be more reasonable. idk how they could make that work since engines are universal across all mechs but the DBL HS cost+artemiss cost should most certainly scale with tonnage like the armour upgrades does (those need to be toned down on the tonnage as well!).


It would if you had the option to change the engine from something like petrol to electric or change the panelling and the frame to different materials.
The equivalent of the 'cosmetic upgrades' on the car for the mechs would be the cockpit items, decals, camo scheme.and colours.

#30 Hobbles v

    Member

  • PipPipPipPipPipPip
  • The Sureshot
  • The Sureshot
  • 354 posts

Posted 21 January 2018 - 09:47 PM

Bump

#31 ANOM O MECH

    Member

  • PipPipPipPipPipPipPip
  • Survivor
  • 993 posts

Posted 22 January 2018 - 12:52 PM

View PostHobbles v, on 21 January 2018 - 09:47 PM, said:

Bump


My game and numbers have improved significantly since I spent about three hours in the mechbay fixing earlier skill tree decisions. Luckily I have you around in TS who has actually spent the time figuring out some really basic stuff I missed such as investing in cool run when the build doesn't have enough heat sinks to benefit.

PGI is unlikely to back off the skill maze. For those of us who don't like spending time thinking about puzzles and would rather be dropping, there needs to be something that clearly and very simply and directly lets players know what exactly will effect their build. Sure lots of it is simple enough by itself, but it is very easy to just do what I had done and come up with more of a default tree that was less than ideal.

Would be even better if there was something like metamechs for skill tree so guys like me can just go and skill stuff out quickly and not really have to think about something that is not fun for us.

I also honestly at this point think PGI would be doing themselves a huge favour working on tier system despite it not being applicable in CW. Part of it, and not to be elitist, but I honestly believe that it would make a lot more sense if CW was greyed out until a player at the very least reaches tier 3. A player should reach the theoretical middle ground before jumping into 'end content' that is absent any form of match maker. The warning is somewhat lazy and almost too Canadian in letting people who will never have a shot or much of a good time participate anyway. Why would anyone be allowed to hope into CW under these conditions if they don't even have a few mastered mechs?

#32 Willard Phule

    Member

  • PipPipPipPipPipPipPipPipPip
  • The 1 Percent
  • The 1 Percent
  • 3,920 posts
  • LocationThe Omega Company compound on Outreach

Posted 22 January 2018 - 01:20 PM

Here's an idea....apply the Tier system to FW, but only in a minor capacity. If you're T5 or T4, then the only drops that are available to you are the QP map/modes and DEFENSE on Invasion. You can't drop on offense until T3.

Perhaps this will stop the 10 LRM boats on the attack teams a bit.

#33 r4zen

    Member

  • PipPipPipPipPipPip
  • The Tip of the Spear
  • The Tip of the Spear
  • 309 posts

Posted 22 January 2018 - 03:10 PM

View Posttker 669, on 22 January 2018 - 12:52 PM, said:

A player should reach the theoretical middle ground before jumping into 'end content' that is absent any form of match maker.


I earnestly believe we would have fewer issues with pugs/solos if there were a higher barrier to entry for FP. I'm not sure what that answer specifically is, whether it's tier, owned/mastered 'mechs, # of QP drops, a combination thereof, or something else, but there needs to be a harder barrier than just a popup you can wave away.

New player experience in FP is atrocious, and a higher barrier won't fix that on it's own, but I think it could improve it more than marginally.

#34 Eirik Eriksson

    Member

  • PipPipPipPipPipPip
  • Philanthropist
  • 201 posts
  • LocationIn the deep forests of Småland

Posted 23 January 2018 - 03:10 AM

View PostHobbles v, on 04 October 2017 - 12:29 PM, said:


The simplest solution (albeit a temporary stopgap at best) is to release Comstar as a playable faction for IS. This may curb some of the mass exodus to clan side rewards.


A more permanent solution would be to scrap the mercenary tree and payout system. As it does not reflect the active populations anyways. Mercenarys should sign up for merc contracts with each specific faction like they did in phase 1 and 2. And earn rewards with each specific faction they are curently working for (at a much slower rate than current loyalists).


Finally there should be no hard cap on loyalty earning. The end of each loyalty tree should have some kind of reward that can be unlocked repeatedly after earning a certain amount more LP, give players at least some reward for sticking with their current faction, and more importantly keep their respective community hubs active.



Thanks for the post, very interesting to read.



Whatever solution to the merchs payout I think there really need to somehow be a one of reset of all player contracts, or at least of all player contracts that have not played in x amount of time which should be simpler to do. If that is done the rewards should at least represent the current actives a bit better. And I think their rewards should somehow be cut lower than loyalists. When the present rewards system was cut out it includes all the "right thinking" in form of steering merchs to factions with lower numbers, but failling to reset the org base of course makes it impossible for such an implementation to work.


Think the LP reward system needs to be redone and be as indifferent as possible for loyalists so that changing factions would be close to worthless when it comes to LP rewards. This should be possible to do by putting the same reward on equal steps, for example 10000, 20000, 30000 LP etc is equal to the specific reward x and are the same in all factions. This could also be done on the merchs side only requirering a bit higher level to get the same rewards.


The LP cap should of course be removed and why not just simply roll it forward so that when max rank is reached, the rank is of course kept, but the rewards are starting over from the begining again. This should be applicable even to some of the achievements in QP by the way.


Edited by Slow Speed, 23 January 2018 - 03:13 AM.


#35 Joshua McEvedy

    Member

  • PipPipPipPipPipPip
  • The Ogre
  • The Ogre
  • 491 posts
  • LocationDuchy of Oriente, Free Worlds League

Posted 23 January 2018 - 03:30 AM

View Postr4zen, on 22 January 2018 - 03:10 PM, said:


I earnestly believe we would have fewer issues with pugs/solos if there were a higher barrier to entry for FP.


Yeah, let's make Faction Play inaccessible to most of the casual community and reduce the mode's population even further. Sounds like a winning strategy to me (after sitting in queue last night for over an hour and not getting a single drop before bedtime). Posted Image

#36 Hobbles v

    Member

  • PipPipPipPipPipPip
  • The Sureshot
  • The Sureshot
  • 354 posts

Posted 23 January 2018 - 06:53 AM

View Postr4zen, on 22 January 2018 - 03:10 PM, said:


I earnestly believe we would have fewer issues with pugs/solos if there were a higher barrier to entry for FP. I'm not sure what that answer specifically is, whether it's tier, owned/mastered 'mechs, # of QP drops, a combination thereof, or something else, but there needs to be a harder barrier than just a popup you can wave away.

New player experience in FP is atrocious, and a higher barrier won't fix that on it's own, but I think it could improve it more than marginally.


Achieve tier 4 or have over 75 qp matches under your belt.

Edited by Hobbles v, 23 January 2018 - 07:20 AM.


#37 CARNA6E

    Member

  • PipPip
  • The Blazing
  • The Blazing
  • 23 posts

Posted 23 January 2018 - 07:38 AM

https://mwomercs.com...t-quirksskills/

^^^ This could piggy back in on #4.^^^

#38 Hobbles v

    Member

  • PipPipPipPipPipPip
  • The Sureshot
  • The Sureshot
  • 354 posts

Posted 23 January 2018 - 08:31 AM

View PostCARNA6E, on 23 January 2018 - 07:38 AM, said:

https://mwomercs.com...t-quirksskills/

^^^ This could piggy back in on #4.^^^


No. House/clan specific quirks or tech is a horrendous idea. No matter what, one house or clan will come out with the "best" options and further throw off balance and concentrate everyone into that faction. Faction specific benefits should not effect battlefield mech performance.

At most I would consider logistical benefits like equipment and mech discounts based on each faction. But even that will have its clear "winners" and "losers".

Best to K.I.S.S and continue with only trying to balance clan vs IS.


#39 naterist

    Member

  • PipPipPipPipPipPipPipPip
  • Mercenary Rank 6
  • Mercenary Rank 6
  • 1,724 posts
  • Location7th circle of hell

Posted 23 January 2018 - 08:39 AM

Build/flush out a meta game that is connected to the galaxy map, then spend time ******* with that. Its the time between matches that should differentiate fw from qp, not in match stuff (outside siege, scouting, and the respawn aspects). Add in cool things for us to que up to take/defend, as opposed to the current, mindless requeing.

#40 Q

    Member

  • PipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 137 posts

Posted 23 January 2018 - 08:50 AM

Re: Stock builds

I know the community has designed some hero mechs (or champions?)... why not allow us to design some "stock" FP builds that are purchasable for c-bills? Maybe they cost slightly more?

That way we have the stock variants available for purchase to keep the lore monkey's happy as well as allow more experienced players access to whatever mech they want to play with... but new players can be exposed to "ready to play out of the box" builds.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users