Replace leaving players with AI

[maik2012]

Hello

I always find it frustrating that players leave and after that the game is basically over. I thought about this and came across different things one could do, but auto-replace leaving or kicked players with strong ai is the mosdt easy, logical and benefitial solution in my opinion. Please implement this.
It has many benefits:

- Leaving players are replaced and even if the bot is often stupid, most of the times the players has already built defensive structures or they are part of the base from the beginning and so you don't get hole in your front.
- Hosts who abuse the Kick function can't do too much damage anymore.
- Players who really have to leave aren't so much seen as the bad guys.
- Playability increases because you can kick laggers more easy, this point has disadvantages as well, but hey it's 2012, anyone who is not downloading and isn't sharing his internet connection with 10 guys or is on wlan should be fine.
- Playability also increases because you can also kick rude, inactive or total inexperienced players. This point has disadvantages as well but the benefits should outweight the dangers here as well.

I think it's a great idea and really something to take into consideration. Besides the eternal Beta Phase, which is splitting the already small community effectively in half, this should should be the most important thing.

I'm not playing often, and i won't play again until these issues are solved, because it's too frustrating and i have better things to do with my time. I really like the strategy part, i think it's a great game and the beta looks very good.

If there are better solutions, fine, but i doubt it.

My other ideas were, neutral game hosts as server and a more sophisticated rating, but it's very difficult and not so benefitial like replacing leaving players with AI.

Regards

[Per]

I'm not sure that is a good solution. However, I think allies should get any power held by the player who is leaving, and a tiny bit of power for each building / unit that is destroyed belonging to this player. This should balance out things a bit.

Another alternative is to gift all units and buildings belonging to the leaving player to a random ally. Not sure what to do if that would push him over his unit / building limits, though.

[maik2012]

I think turning over all units and buildings/structures to the remaining players is a good idea as well. But you should be able to use them, so that the remaining players might have more than 5 research or more than 10 power.

Correction to first post: I don't have much time anyway and like this game, so i will come back playing no matter what.

[JJjopando]

and what about their ability to produce power and units? I would rather have the AI.

[NoQ]

When the AI becomes clever enough, i'd agree to that. Probably a special "Replacer AI" could be made.

One of the main problems is that the AI shouldn't go against the research path already picked by the team. So it should either accept the hint as a chat command or focus on fundamental research only.

[Lord Apocalypse]

Or if nothing else set an AI in Defensive mode. The idea of replacing a dropped player with an AI is pretty good but at present I don't know how well the AI could handle getting dumped in the middle of a game. Think there is a way to have an MP AI check its own prior research as well as its allies and make an attempt to guess what kind of path it should take?

[NoQ]

Think there is a way to have an MP AI check its own prior research as well as its allies and make an attempt to guess what kind of path it should take?

I guess i could make a version of NullBot that will look through his different research paths and choose the one that is most finished on the phase of subpersonality choosing.

But right now, unless this "replace myself with AI" feature is at least technically implemented, i have no way to test this behaviour.

Making chat command that forces him to switch subpersonality is an easier idea (cause this is useful and testable regardless of player replacing feature), so i will start working on it anyway once JS chat is fixed.

[Lord Apocalypse]

Will be a nice addition either way

[Giani]

But right now, unless this "replace myself with AI" feature is at least technically implemented, i have no way to test this behaviour.

Maybe with T2 or T3?

[NoQ]

Emm, 1) T2 and T3 have nothing to do with real game research statuses (nobody will ever research all branches uniformly, this is crazily bad), 2) i still need to test the technical part of it (how the AI starts when it happens, what extra code i need to add for that to work at all).

[Emdek]

Both possibilities, especially designed AI or randomly giving away stuff to allies (including checking if it could be done, due to limits - stuff that couldn't be given away should explode like in current solution) are nice.
But probably the latter will result in more fair result, maybe except case when all or most of players will reach limits and units of player that left will be destroyed instead of being given to allies which would result in sudden change in proportions of units between teams. But still, in worst case it would end up with the same result as currently.
AI would be best replacement if it could analyze how player behaved until left and try to behave in similar way. But it would be very hard to make it transition smoothly... At least it could try to determine how often he tried to attack etc. and try to simulate it.

[Avestron]

Would it be possible for a player who starts a game be assigned a random entry key that Warzone would log both on the player's machine as well as on the server side?

If so, then would it also be possible for a player to rejoin a game being given AI babysitting support? This would reduce the need for an AI to be able to conclude a game - although it would be able to do so to some degree.

As for the research dilema, why not simply go with random probability with weighting of different research options based on power invested to date? If a player has invested 1000 power on machine guns and 500 power on flamers then he is twice as likely to invest in the former - and the AI adjusts accordingly. The exception to this rule is the introduction of sub-tier weapons, which would be given an appropriate weighting to encourage their research.

For instance, if a player invests 2000 power on cannons and is suddenly offered the chance to invest in rail gun technology then that technology should automatically gain an appropriate boost in probability to encourage development.

/\/\/\/\/\

Another way to go about this would be to classify different areas of development:

- Weapons Technology > With variable favouring of Anti-Air Technology depending on the primary path chosen (the ability of the core weapon vs Air).
- Vehicle Development > Body and propulsion enhancement
- Infrastructural Development > (Engineering, Power, Research, Manufacturing, Repair, Rearming, Sensors)
- Defensive Development > (Defensive Structures, Base, Hard Crete development)
- Advanced Development > (A further advantage for cutting edge techs - seeks high level techs available and semi-randomly selects)

[Andrie]

It would also be good if players can then build more unit's.