25 replies to this topic

[Blakkstar]

Personally I'd like to see them concentrate on CW and refining the core gameplay. PvE content that's not developed as part of the core game is usually pretty underwhelming. Look at Hearthstone...there's some kind of fun (and short) content that's mostly built around cheating AI, and sitting behind a paywall (or major F2P time investment). It's also pretty much mandatory to keep up with the meta because you get new cards for playing it.

Then there's the EVE model, where you try not to die from boredom while auto-firing your way through a dozen cookie cutter missions while hoping for RNGesus to give you a magic loot box.

So how much are you willing to pay for MWO PvE?

[Tombstoner]

There is a significant amount of money to be made via PVE. Personally i think its far more then PVP. for example for $10 you could get 5-10 mission pack covering a short story arch and be rewarded with cash and salvage based on performance. multiple play thoughts can be achieved by higher difficulty levels for more rewards. top level being a new unique mech/ weapon.
In fact new player experience / training would fit in perfectly with a PVE environment and be the first PVE package for free. However PVE is a very different sort of beast then PVP.

AI skill is first and foremost IMO the number one issue. Weapon stats and lance mate skill level being second/third. Some may think going 1 vs 12 and winning is a satisfactory experience. i think the player is buffed far too much and it becomes unrewarding.

If skill level of the AI is dynamic players can deliberate miss lowering AI skill level and then start alpha head shots

Building a fun PVE game i think is a greater challenge then PVP.

Edited by Tombstoner, 19 November 2014 - 06:23 AM.

[Killstorm999999]

Tombstoner, on 17 November 2014 - 10:57 AM, said:

It is very simplified because thats the bottom line. Your right in that you can construct different probabilities via multiple factor inputs for the player and AI. The primary point is fun becomes the illusion of competition not actual competition because your facing a scripted encounter your going to win x% out of y times if your performance is greater then c. But giving both a player and an AI a COF weapon aimed at your head. The hit probabilities become the guns hit probability and the players ability to be on target at the moment of fire. The AI would have a perfect ability but this would matter very little since the difference between the AI and the human is much closer to parity. Then meaning full game play decisions can be made without an over whelming twitch advantage inherent to a non handicapped AI.

But in the end can it still be fun knowing that the outcome is just some form of 2d6. well since the economy of Las Vegas is build upon that premise a lot. Just don't fool your self thinking your actually competitive with an AI. Its like playing hoops with Michal Jordan or boxing with Mike Tyson.

I know what you mean, but I feel like it's the same thing when you consider 2 humans fighting it out. You know one or the other has to win, and you can probably figure out which one has the greater chance. It's all based on player skills, such as accuracy, ability to use terrain, calmness under pressure, and so on. These skills could be programmed into an AI.

Of course with the two humans there is a chance one of them will make a mistake, or take a long shot risk, but these same things could be programmed into an AI. The AI could have it's accuracy lowered when it has critical damage to represent panic, or a particular 'personality' of AI might always attempt headshots even though the chance to succeeding are very low.

I think the key is not make a challenging AI (which is very easy, you could just make its accuracy perfect) but to make an interesting AI, that behaves unpredictably. In any case, we just need something that we can put into all these potential PvE scenarios.

[LordKnightFandragon]

MWO AI: Give them a CoF, centered around aiming at our upper CT. Obviously, twisting would be very effective lol, spreading damage all the same.

And in some cases, an AI is harder then a person because an AI will not cower, if its got numerical advantage, it will use it, as its orders are simply to follow a waypoint or move to a certain area.....and in some cases that is highly problematic. I know in Battlefield games, being swamped by AI is alot harder to deal with then 1 or 2 humans. Plus, AI will always be there to support you. Players might cower or wander off. We just need a nice, 4 man coop campaign...even if it is just a series of capture and destroy objectives using just hte mechs and maps we have in game, an AI and bam.....it would be awesome lol. Make a sort of story for the mission progression and go at it.

[Blakkstar]

Is anybody here familiar enough with Cryengine to field a guess as to how well it could be adapted for PvE?

[LordKnightFandragon]

Blakkstar, on 19 November 2014 - 06:59 AM, said:

Is anybody here familiar enough with Cryengine to field a guess as to how well it could be adapted for PvE?

Cryengine? The same one that is used in Crysis games? id vote pretty well then....