Player experience.
Moderator: Forum Moderators
Forum rules
Before posting a new idea, you must read the following:
Before posting a new idea, you must read the following:
Player experience.
I think, that Wesnoth should match AI level to player level. Best way could be giving/reducing player's experience points. It could work like giving money for fast end of mission, but we must set minimum level of experience based on actual player level. When player has smaller amount of exp for mission than request, we will reduces player experience.
Player can get next level and can lost level.
We can also make Wesnoth more depend on experience level. It will be always private, not public. I think, that we can use player skills level to separate players in network game, adding rooms to network game. UI will automatically select the best room without sending any information(each room will have set minimum required level and maximum required level). Many times I played with total noobs and with very average players.
Download/addition manager could been integrated too, separating campaning too hard for player.
Player can get next level and can lost level.
We can also make Wesnoth more depend on experience level. It will be always private, not public. I think, that we can use player skills level to separate players in network game, adding rooms to network game. UI will automatically select the best room without sending any information(each room will have set minimum required level and maximum required level). Many times I played with total noobs and with very average players.
Download/addition manager could been integrated too, separating campaning too hard for player.
-
- Art Contributor
- Posts: 1700
- Joined: December 7th, 2006, 8:08 pm
Re: Player experience.
If the player is getting killed, AI should suddenly run away, and send all its units to shallow water, and let the player murder them.
I like it.
I like it.
Re: Player experience.
I don't. (You were probably being sarcastic) I kind of agree that we should have AI on different levels, but that's about it. (Right now we kind of have a medium-easy) But that's kinda changing gameplay. I think it makes more sense to let the AI have less money than you, which you can already do.
Plus, I think it's an FPI to separate people by how good they are too. First, it will make bad players feel bad, and make good players too proud. (Which we already have a little bit of even with out competition)
Also, campaigns are made by different people, who all have different ideas of what's hard and what's easy. And the players will have even more different ideas. Plus, most campaigns, (and all mainline ones) are already labeled with "Novice, Intermediate or Expert"
Plus, I think it's an FPI to separate people by how good they are too. First, it will make bad players feel bad, and make good players too proud. (Which we already have a little bit of even with out competition)
Also, campaigns are made by different people, who all have different ideas of what's hard and what's easy. And the players will have even more different ideas. Plus, most campaigns, (and all mainline ones) are already labeled with "Novice, Intermediate or Expert"
F:tGJ, Saurian Campaign
The Southern Chains, a fanfic
“The difference between winners and champions is that champions are more consistent."
~Sierra
The Southern Chains, a fanfic
“The difference between winners and champions is that champions are more consistent."
~Sierra
Re: Player experience.
Involuntary matchmaking based on skill is an awful idea. Segregation based on skill leve is also quite bad. You only get better by getting your butt handed to you. Only be getting beaten badly and observing how the "pros" play, can you get better. Doesn't matter how good the documentation is.
- beetlenaut
- Developer
- Posts: 2825
- Joined: December 8th, 2007, 3:21 am
- Location: Washington State
- Contact:
Re: Player experience.
The current AI is very easy. Try playing with it in multiplayer when it has the same resources you do. It's boring because you don't have to think. (Still, it's better than in many other games.)Hulavuta wrote:I kind of agree that we should have AI on different levels, but that's about it. (Right now we kind of have a medium-easy)
The only way the AI can be made harder is by giving it much more gold than the player, and more high-level units. Adjusting the AI this way is a good idea, and we already do it: That is the difference between campaign difficulty levels. Then, we allow the user to choose the AI difficulty they want to try.
Campaigns: Dead Water,
The Founding of Borstep,
Secrets of the Ancients,
and WML Guide
The Founding of Borstep,
Secrets of the Ancients,
and WML Guide
Re: Player experience.
Gambit wrote:Involuntary matchmaking based on skill is an awful idea. Segregation based on skill leve is also quite bad. You only get better by getting your butt handed to you. Only be getting beaten badly and observing how the "pros" play, can you get better. Doesn't matter how good the documentation is.
73/100 and over 3 million users doesn't seem like a bad game. But the gameplay screenshots look terrible.
I guess I must suck or something, I don't think the AI is THAT bad. I mean, it's still pretty bad, but not that bad. (Although it's still pretty bad) Maybe it's because they always spam Gryphons against me? IDK.
F:tGJ, Saurian Campaign
The Southern Chains, a fanfic
“The difference between winners and champions is that champions are more consistent."
~Sierra
The Southern Chains, a fanfic
“The difference between winners and champions is that champions are more consistent."
~Sierra
-
- Posts: 637
- Joined: December 20th, 2009, 5:59 pm
- Location: Germany
Re: Player experience.
I agree that we need difficult levels for the ai like a changable error rate but at first, the ai have to get better and I think the ai do is that much worse. For example, while recruiting, the ai doesn't understand how to defeat DAs. The ai also don't knows how to use slow so the ai is really bad.
Re: Player experience.
I think, this is to complicated.
More easly is if the team makes many lvs of skill for the AI, maybe 5 can be sufficent.
Well, the default AI should be in the level 3, I know for newcomers its hard, but for officially veteran players its a fool. If you know well the AI you can do to AI moves its better unit in a shallow water to kill you an almost useless unit lol. I beated an AI with 125 starting gold and I with 75 starting gold in Hammlets, it starts with almost the double of the gold and it losses againist me, there should be an harder AI.
And AI dont knows anything of ambush, it always wants to move to an square, but always is surprised by a wose or by a shadown. And AI in many times dont think in ZoC, so it its fragible.
Other thing is when you pass in AI observing zone, and AI continues moving to the center of the battle... that is perfect to make an AI easly to surprise with many quickly units and this can mean its death.
More easly is if the team makes many lvs of skill for the AI, maybe 5 can be sufficent.
Well, the default AI should be in the level 3, I know for newcomers its hard, but for officially veteran players its a fool. If you know well the AI you can do to AI moves its better unit in a shallow water to kill you an almost useless unit lol. I beated an AI with 125 starting gold and I with 75 starting gold in Hammlets, it starts with almost the double of the gold and it losses againist me, there should be an harder AI.
And AI dont knows anything of ambush, it always wants to move to an square, but always is surprised by a wose or by a shadown. And AI in many times dont think in ZoC, so it its fragible.
Other thing is when you pass in AI observing zone, and AI continues moving to the center of the battle... that is perfect to make an AI easly to surprise with many quickly units and this can mean its death.
Creator of: Deathmatch new in 1.12 server.
Co-creator of: Era of Magic in 1.16 server
Developer of: Empires in 1.12 server, Ageless Era in 1.10 to 1.16 servers (but innactive recently)
Try My winning Orocia Guide
Co-creator of: Era of Magic in 1.16 server
Developer of: Empires in 1.12 server, Ageless Era in 1.10 to 1.16 servers (but innactive recently)
Try My winning Orocia Guide
Re: Player experience.
Writing a Wesnoth AI is hard.IPS wrote:I beated an AI with 125 starting gold and I with 75 starting gold in Hammlets, it starts with almost the double of the gold and it losses againist me, there should be an harder AI.
The Wesnoth AI is much worse than losing 125 gold to 75. It can lose with more than double your starting gold, extra income, extra village income, etc.
-
- Posts: 637
- Joined: December 20th, 2009, 5:59 pm
- Location: Germany
Re: Player experience.
No, really not. You just need a random regenerator for the error level and if it worked, just make a thing being in the eyes of the ai not the best thing and since the ai calculate at least for attacking units every power of every possiblity (I don't know really good because I just took a look very uncarefully since I understood, that ai is to complicated for me). So the random regenerator just choose a random number depending from the set errorlevel. The action with the power comparing to the "best" action being as near as possible to the number will be done. But jsut like I said, the ai have to get better first, which is, if you take a look at Caphriel's post, not very easy.IPS wrote:I think, this is to complicated.[...]
Re: Player experience.
Actually, additional randomness to the AI's decisions might make it harder, by making it less predictable.
Currently AIs greatest difficulties seem to be, moving units together/using zoc to protect weaker units and how it almost never ever attacks with low HP, even when it can't retreat.
Currently AIs greatest difficulties seem to be, moving units together/using zoc to protect weaker units and how it almost never ever attacks with low HP, even when it can't retreat.
Re: Player experience.
That is unless you set agression to max, causing it to frenzy on everything and expend it's forces on hopeless assaults. (Sisal's AI in Northern Rebirth FTL)
Project Battlescar: An rpg engine of my own design.
http://battlescar.wikispaces.com/
http://battlescar.wikispaces.com/
Re: Player experience.
We're talking about Mp though...unless you are suggesting implementing that idea into Mp...
F:tGJ, Saurian Campaign
The Southern Chains, a fanfic
“The difference between winners and champions is that champions are more consistent."
~Sierra
The Southern Chains, a fanfic
“The difference between winners and champions is that champions are more consistent."
~Sierra
Re: Player experience.
Definitely not. (On standard MP anyway. Tower defense or team survival on the other hand, not so bad an idea.)
Project Battlescar: An rpg engine of my own design.
http://battlescar.wikispaces.com/
http://battlescar.wikispaces.com/
Re: Player experience.
Hmm... I had see in the survivals maps AI frequently is more careful than in default games.
In some maps like colosseum, it risk a lot.
I think, there should be many types of AI, and many or all of them should be able to pick to default games. Doing some team games more interesting. And more if each AI haves it grades of difficulty. If you acept the levels of each different AI, them should be 3: Easy, Average or Medium and Hard.
Example:
Medium careful AI and Hard Spycho AI vs 2 human players.
In some maps like colosseum, it risk a lot.
I think, there should be many types of AI, and many or all of them should be able to pick to default games. Doing some team games more interesting. And more if each AI haves it grades of difficulty. If you acept the levels of each different AI, them should be 3: Easy, Average or Medium and Hard.
Example:
Medium careful AI and Hard Spycho AI vs 2 human players.
Creator of: Deathmatch new in 1.12 server.
Co-creator of: Era of Magic in 1.16 server
Developer of: Empires in 1.12 server, Ageless Era in 1.10 to 1.16 servers (but innactive recently)
Try My winning Orocia Guide
Co-creator of: Era of Magic in 1.16 server
Developer of: Empires in 1.12 server, Ageless Era in 1.10 to 1.16 servers (but innactive recently)
Try My winning Orocia Guide