Armando, on 30 May 2016 - 01:44 PM, said:
@Danjo San:
Instead of using a butcher knife (make every unit with more than 48 members split up), use a scalpel (factor in the enemy teams win% into the rewards)? If a team has a collective 10.00 w/l ratio, take the base rewards for a loss x10 for the payout.
Give those facing insurmountable odds EXTRA incentive; If the enemy team has a collective 10.00+ W/L Ratio, take the base rewards for a win x10 as well.
This would turn a loss against a team YOU SHOULD lose to into a serious c-bill boots, allowing the people who need it the most to buy/build/module-up mechs that can compete FASTER.
Raise you hand if you have always been the pilot who is getting stomped by 12 man units, you have never be IN a 12 man+ unit drop, and you HATE not getting $%!@t for rewards because you got curb stomped!
If you raised your hand, tell me if xN payout where N is the collective difference between your teams W/L Ratio and the enemy teams W/L Ratio is paid out for both wins AND losses.
Example:
Enemy Teams Collective W/L Ratio: 10.00
Your Teams Collective W/L Ratio: 0.50
(N) Collective Difference: 10.50
For a loss: Base Payout (current payout) x10.50
For a win: Base Payout (current payout) x10.50
@Danjo San: Your thoughts on turning obvious losses into C-Bill boosts for those you need the C-Bills the most?
Even better - you simply use the existing PSR ranking system to value each team and have a separate PSR (seeded from QP) for FW.
So if MS fights 228 they are both going to get a crap ton more payout than if MS fights skittles. The skittles MS fights will may way more losing to MS and would make more money if they beat an MS 12man than if 228 beat an MS 12man.
So you get a payout based on both the relative 'value' of the other team as well as a bonus on the relative difference in value between your teams.
This rewards units for trying to play other units and good teams for playing good teams and it strongly rewards everyone else for playing against them.
Remove Attack/Defend queue so there's only one queue on each front, this way if 228 is in Kurita and MS is in Davion and there's fighting on the Davion/Kurita border you don't have both 228 and MS in the attack queues stomping puggles who clicked on defend because most people still, 2 years later, don't understand how (stupidly) FW works. With units going to the front of the line this will generally help shuffle 12mans into fighting other 12mans.
Remove tagging planets from any sort of value. Make it 100% participation based. If a MS 10man drops for 4 hours on a world, every match they win has 2 pugs in it. Suppose there were 30 total winning matches in flipping that world for the winning side. Everyone in those matches, regardless of unit membership or not, gets 1 MC for each winning match and an extra 200k cbills for each match they were in that was a win. Then they get 1 MC and 500k cbills for every full 24 hours it stays in their factions possession.
Just random numbers but the result is that every pug who plays with that MS 10man in winning matches gets the exact same reward as every MS member in those matches for taking the world. This makes playing with units more rewarding for pugs than playing solo rambo. Dropping with a unit or a group now pays them more and if that group drives wins it pays them a *lot* more to be actively trying to flip a world.
Doing these things would completely remove the hostile dynamic between big/good units and smaller units and pugs. Everyone benefits from those units and playing with them as well as against them. It puts the biggest reward on playing more often and playing to your team to win - the two behaviors you want to drive, in a unit or not.
Make LP rewards cycle back to start over when maxed.
We also need to add in a ton of bonuses for Loyalists. Say 15% MC gain bonus, 30% Cbill bonus, 50% LP bonus. Every cycle in which a Loyalist faction takes planets they get a bonus, offset by a negative for losing planets. So a net loss results in a net negative, etc. Loyalists can vote on a 7 day cycle to reduce any/all of those bonuses to provide new bonuses to mercs who take a new contract in that new cycle timeframe.
So you have the vote screen for who to attack then you have another with MC gain, Cbill, LP. You could vote to, say, reduce CBill bonus by 10% to give mercs a 30% bonus and LP gain by 10% to give LP gain a 20% bonus for mercs up to a max of X merc players. If more mercs than that sign up then the Loyalists lose a litlte more bonus % in the next cycle.
Mercs then see what rewards are offered by what factions and the unit leaders choose which to apply for. So suppose a new cycle starts June 1st. The last 7 days of May the Loyalists decide by vote how much of their bonus they're going to give up in the first week of June to attract mercs (if they want). On June 1 to June 7th any merc starting a contract in that faction sees the offering and if they sign a contract they get 7 days from their start date of that reward if they're accepted. They have to try to accept a contract 48 hours in advance.
HOWEVER. They click 'accept' and Loyalists see who applied. Over the next 48 hours loyalists who care can vote to accept or decline. Mercs have to meet a certain threshold to get their contract accepted. The more mercs who are accepted the steeper the cost to Loyalists bonuses for that week.
A merc can always take a contract anywhere with no bonuses.
Essentially this creates a system where Loyalists get big rewards but are stuck in a single location and faction. No running off to play their Kodiaks. They can sacrifice those rewards when needed to attract mercs who get bonuses for the actual drops they make so mercs that don't really play won't see big payoffs. It strongly rewards mercs with a good reputation as the more mercs who accept the more it reduces the loyalist bonuses so it's ideal to get the smallest number of most effective mercs. This makes groups like KCom and 228 every bit as desireable as, say, MS because they're smaller, however MS is a juggernaut that can really move the map for their employer.
Make voting results public so Davion can SEE that Marik, for example, is going to offer *big* rewards to attract mercs next week and as such decide if they want to try and out-bid them or at least try and attract mercs to help them in turn.
This strongly rewards merc units communicating with loyalist units to build ties and trust to increase the odds of mercs getting accepted when applying for a contract there. It rewards merc units for building a good reputation. It rewards loyalists for actively participating in voting that directly affects their own rewards. It helps create the framework for an involved semi-political system for driving actual Faction Warfare.
Anyway. Yes. No question, I am confident I could make FW way better than it is now. The above ideas would all need tweaking but the concept would create the environment we want. Neither it nor anything like it will ever see the light of day.