In the end I find this a far better option than the 'die once per planet' approach put up elsewhere.
Here's the thing. The most a 'troll unit' can cost a planet in tickets is however many each player gets on world. They'd be functionally wasting time trying to 'drain tickets' when they'd be better off playing AGAINST and just killing mechs. I don't see this being a big issue.
You have pugs able to play and count towards tickets but you adjust the ticket totals to win/lose accordingly, as in the total needed not the value per player. You can't split the value between pug/units or people will just game the system by syncdropping.
I'm fond of the idea of 16 total mechs per world per 8 hour stretch. That's 4 total wipe losses. In the end this system creates the best possible balanced IMO between zerg numbers and skill performance.
1
24 replies to this topic
#21
Posted 09 January 2015 - 02:48 PM
#22
Posted 10 January 2015 - 06:51 AM
I think to start to keep it simple PGI just gets a ticket system up. Once it's in place they could play with it and make it more complex.
#23
Posted 10 January 2015 - 02:38 PM
Vas79, on 10 January 2015 - 06:51 AM, said:
I think to start to keep it simple PGI just gets a ticket system up. Once it's in place they could play with it and make it more complex.
The advantage of a 'ticket system' is exactly that. Easy to modify once in. How many tickets per player per world can be adjusted later. They could just put it in and track the results before even implementing limitations on it. This is, IMO, a mechanic to put in a viable version of the other suggestion Gut put in *coughafteryoudidcough* and make it work.
Could even have 'major worlds' have a huge ticket bucket to represent being worth fighting over and minor worlds a small ticket bucket. This would even make it viable to open more than 1 world on each faction front.
Think about that for a second - each front may have 2 worlds open; 1 small ticket world and 1 big ticket major objective that could take several days to flip.
It may be that you need to take several smaller worlds to open a single bigger one and then the bigger one will likely be fought over, HARD, for extended periods. During which time the smaller objective worlds with relatively minor ticket buckets can be fought over or not as each faction decides where to spend its soldiers time.
On the Dav/Liao border for example; Capella and Sarna and St. Ives would be major objectives. If either faction is within 1 world of them they are opened up and have huge available ticket amounts and should require a massive number of collected wins to flip, ideally over several days. Beid and Ranar would be small ticket objectives; 2-4 drops per player per 8 hour period worth of mechs to be lost. Each unit and faction may focus only on major objectives and ignore the smaller ones or they may try to balance the two or however they want to split it up. This gives people something to continue to work towards if they get burned off a smaller objective.
Make sense? It's a more complex, logical outgrowth of logistics and tickets and gives faction vs faction wars a more epic feel for me.
#24
Posted 12 January 2015 - 03:42 PM
The number of tickets could be easily adjusted to represent important worlds. It would be simple to start but could easily grow in complexity as time when on and changes were made to CW.
#25
Posted 12 January 2015 - 04:40 PM
Vas79, on 12 January 2015 - 03:42 PM, said:
The number of tickets could be easily adjusted to represent important worlds. It would be simple to start but could easily grow in complexity as time when on and changes were made to CW.
Exactly. What we want is a system that is easy to expand to include logistics and varying targets and other changes in CW over time, which this would let us do.
Hmm... how about if we make the tickets per player and not per world? That prevents a dedicated group of bads from eating up all a worlds drop tonnage? So each player gets, for example, 16 total mechs they can lose on a single world per 8 hour cycle on a minor world and 36 they can lose on a major world? Maybe half that, I don't know - depends what the telemetry looks like.
That would largely fix turret drops, zerging and spamming bads, even reduce the relative value of tar pit tactics or at least turn it into one of 'wanting to ensure you do the most possible damage' rather than 'take the most possible time', as the loss of mechs is more relevant than the loss of time.
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users