THIS IDEA WOULD BE MUCH BETTER IF IT WAS AN ALTERNATE 4v4 SCOUTING MODE.
edit: and some minor changes.
1. The data collection point is destructible
2. Either side can gain intelligence from it (net differnce in intel collected)
3. You only get rewards for gathering and returning intel
4. Mechs can only hold so much intel before returning it to the dropship (because they are running potato powered hard drives)
The issue I have with this is it again creates a precedence on medium mechs and makes lights almost useless, and cannot be used in quickplay. Basically take as many Stormcrows with streaks as you want, enjoy the fragfest.
The issue I have with this is it again creates a precedence on medium mechs and makes lights almost useless, and cannot be used in quickplay. Basically take as many Stormcrows with streaks as you want, enjoy the fragfest.
I would imagine that 4 ERLL pirates banes would laugh at the skillcrows. Everyone gets their scoop of intel, smashes the cap point and runs away. Keep in mind most of the domination maps are a decent run to get to the circle, enough that the speed difference would matter.
Remember that killing and damage doesn't pay AT ALL in my version.
Alistair Winter, on 10 August 2016 - 06:47 AM, said:
I understand that, but this doesn't cover the whole situation in this game mode. Your example only applies if Team A is wiped out, but got more data in the process. Otherwise, you would have to come up with some crazy far fetched explanation. And what if Team A got 0 data and got wiped out. Would Team B win automatically?
It was only meant to cover the edgecase of one team having more success in the objective but losing the battle, aka "winning the war but losing the field."
In the case you provided, Team B has effectively won. At that point destroying the enemy dropship is only worth higher rewards. (Hence the reason for multiple levels of victory and defeat).
Quote
But it still seems like a better idea to just go ahead and kill the enemy team. Because if the enemy team is dead, then they can't prevent your dropship from escaping either.
How you go about it is intended to be up to you. Is it better to kill them all and take your chances of killing a dropship with canonical values of 14 front (1,400 armor to the front), 12 side (1,200 armor to the left and right sides) and 10 rear (1,000 armor on the rear) within the span of 30 seconds or less stationary time [depending on how quickly you kill the last four enemy 'Mechs] and the time it reaches -- gradually -- its maximum thrust of 3G? Because even with the gradual acceleration I'm calling for which will be like a semi truck getting its way up to 60 mph, this thing will travel across the map faster than a Firemoth at 200+ kph with Unlimited MASC and unimpeded by terrain.
In which case what reason is there not to play the Sniper Bait game?
There is nothing arbitrary there except that the time they wait is 2 minutes, rather than each dropship pilot dynamically deciding when to say "**** you, I'm not paid to die" or "I won't leave you behind" and shifting course to rescue the last 'Mech from destruction at the risk of being destroyed itself.
Quote
Think of it like this.
There are 20 data points total.
After 5 minutes, Team A has 3 data points, Team B has 2 datapoints.
Team A has 5 mechs left, Team B has 1 mech left
In this scenario, if the last mech on Team B is killed, then the dropship leaves and you have to destroy the dropship. But that's not really a deterrent. Because even if Team A gets wiped out trying to destroy the dropship... Team B still can't prevent Team A's dropship from leaving either.
True. Again, in the end it is another scenario of multiple levels of victory and defeat.
In that scenario, if both teams are wiped out and both still have dropships, it would be a draw if they have even intel, and if one has greater intel then it is a minor victory for one and a moderate to major defeat for the other.
Are you beginning to get what I was saying?
Quote
Furthermore, even if Team B has 4 datapoints and then gets wiped out and their dropship leaves with 4 datapoints while Team A only has 3 datapoints... there's still 10 minutes on the clock. Why wouldn't Team A just keep collecting data and have 16 datapoints after 15 minutes?
It's just very convoluted and arbitrary, it doesn't make sense unless you imagine some kind of crazy story to go along with it. Furthermore, it seems like it would be the best strategy to just get kills anyway. The more kills you get, the harder it becomes for the enemy to stop you from gathering data. Once the score is 10-2, the winning team can just send 5 mechs back and forth to collect data. So just like Conquest, just like Domination, just like Assault... the winning strategy would normally be to get kills instead of caps.
I would set the amount of data to be less than 20. And an uneven number to boot. Say 5 or 7. You'll never see 7 lights on a single team. Even 5 would be unlikely to get all the data in a single go. A low number makes every single piece crucial.
Further limiting it to "one" to obtain per 'Mech (at a time) would also make the traffic all-the-more-important along with each trip having that much more meaning. Do you go now before the enemies are there and risk potential long range death? Do you wait for a tide in the assault/heavy stalemate? Do you go in under the protection of a Heavy damage soak? What happens if you "Drop" the data aka are destroyed, will they be able to retrieve data from your corpse?
The list goes on. My post wasn't meant to be the "be-all-end-all" to the discussion, just an opening of the avenue of conceptualization by answering a single edge case that to you, didn't make sense.
Anyway, as explained across 3 posts now.. Incremental levels of victory and defeat.
Get data, dropship survives, you survive, enemy team destroyed, enemy dropship destroyed. Total Victory! All the moneys! It's raining money!
Get data, dropship survives, you survive (as the team mind you), enemy team destroyed, enemy dropship still alive. Major victory. Lots of money!
Get data, Dropship survives, you survive, enemy team survives, enemy dropship destroyed. Major victory! Lots of money!
No one got data. Dropship survives. You survive. Enemy team dies. Enemy dropship dies. Major Victory! (assume you got data later).
Get data, Dropship survives, you survive, enemy team survives, enemy dropship survives. Moderate victory! Decent money.
Fail to get data. Dropship survives. You survive. Enemy team is dead. Enemy dropship destroyed. Moderate Victory! Decent money; got paid.
Failed to get data, Dropship survives, you survive, enemy team survives, enemy dropship is destroyed. Minor victory! Decent money.
(Note one dropship being destroyed but both teams have forces; mission isn't over yet).
No one got data because both dropships die. Gotta duke it out then. Minor victory to draw.
Both teams wiped out, no data superiority. Draw. Got paid.
Get even amount of data, draw up to moderate victory/defeat depending on other outcomes.
Fail to get data. Enemy team failed to get data. Dropship survives. You died. Enemy team died. Enemy dropship survives. Minor defeat. I showed up, gimme some moneys.
Fail to get data (enemy team got data). Dropship survives. You died. Enemy team died. Enemy dropship survives. Moderate Defeat. This is what I get paid?
Fail at everything, other team gets all the conditions for total victory. "25,000 cbills + some for what what I did?!"
The list goes on and on for each edgecase.
Insert a scenario for every single edgecase and award Total, Major, Moderate, Minor, or Draw victory/defeat to each. The higher the level of victory, the higher the reward.
In other words you are paid based on the objectives you complete.
Just killing the enemy but not getting the data, not destroying the dropship is basically a minor victory. Kinda low pay. Thus, Skirmish only is not the way it would fluctuate unless it's a team that doesn't care about earnings.
Much like old Assault, back in closed beta the amount of money your side could earn through capturing the base was significantly more than you could earn through killing everyone yourself. So what was the natural flow of combat? Earn moneys.
And the natural defense? Protect the base.
This was until capturing the base was only worth "25,000 cbills" instead of 200,000 cbills (at a time when total earnings could reach 500,000 without premium time, that cut total possible earnings to barely over 300,000, further cut by other changes to economies). At that point objectives became meaningless and we got this:
Thus. Money.
Better level of victory, better earnings.
I kinda like the fact that fast mechs are going to have a very specific role, while the heavies\assaults suddenly lose their importance and are actually providing cover fire for the light mechs.
However sticking a USB to the HQ instead of just transmitting over some wireless method seems kinda weird when it's 1000 years advanced from now
Not if you think that info is being monitored and intercepted.
They still use physical runners in a modern military for certain reasons.
I would imagine that 4 ERLL pirates banes would laugh at the skillcrows. Everyone gets their scoop of intel, smashes the cap point and runs away. Keep in mind most of the domination maps are a decent run to get to the circle, enough that the speed difference would matter.
Remember that killing and damage doesn't pay AT ALL in my version.
That might work, but we know PGI won't remove "kill" and "damage" payments. Not to mention the huge uproar of very unhappy players. You also have the issue of "Farming" where both sides just abduct equal amounts of items and won't even fight because easy money.
Whenever a developer comes up with an idea for something, the first step of the developer is to pick it apart to see how players can break it. Removed Link on incentive systems and politics Pay attention to the incentive systems aspect of this discussion made by game developers, for aspiring game developers.
You may see a number of ways to break the 4x4 idea, including how much easier it is to get 4 enemies to agree with farming the game than it is with 12 enemies.
(Edit: Removed link; it isn't the specific video I wanted to reference; give me a moment to watch through the series for the specific part I want you to see.)
Edit 2: That video is really educational, but for reasons that don't relate directly to this discussion.
This one however, does and is the one I thought I was linking.
There is a kids party game like this. There is a pail of water in the center and a bucket for each team. The teams are made up of 4-5 kids. Each kid has a spoon and they must spoon water out of the pail and carry it to the bucket. The first team that fills their bucket wins.
It is rather hilarious with 10 kids trying to spoon water out of the same pail and tripping over one another as they try to get the spooned water to their bucket without spilling it.
We never let the kids kill each other to win though.