I had plenty of games where we were up against a mediocre team and had one player in our team who quit. In many cases they weren't even rage quitting. Those games were ruined by the AI giving gold to the enemy. What exact problems do you have with a simple change as the AI simply not giving gold to the enemy team?

Player leaving mid game will always happen, regardless of any implemented system to preven that. And when that occurs, the AI will feed the enemy.

I don't understand what this has to do with the proposed change. This change will certainly not use as much time and effort as replays, clan options, pantheon fixes etc.

If a player quits often but doesn't rage quit then he has connection issues and you should make a note not to play with him since he may arbitrarily dc... will be easier when player ban is implemented... does not require further support intervention.

I consider it unnecessary in each of the three cases I've described... clearly.

Replays are already basically implemented; in fact there's already a third party utility for viewing them.  They just need some polish.

And yeah clan options, pantheon fixes, etc. they are higher priority IMO at this time.

BTW, I've already described an enhancement that would be nice to see down the road that more directly addresses the problem what I consider to be more correctly and robustly.

Obviously if everybody is quitting sporadically, then there is a deeper problem than fixing the AI that we are left with and we should fix that instead.

Obviously if it is just one player quitting sporadically, then we should avoid that player.

Are there any other cases I have to debunk?

Unnecessary in that

- if they happen at the start we remake
- if they happen at the mid (rage) you were gonna lose anyway [and the rage is being dealt with anyway]
- if they happen at the mid (not rage) see above debunk

none require the kind of developer intervention proposed in this thread. ergo, it is unnecessary.

I don't see you mentioning anything about "fixing" the problem of a player leaving for any reason. You just think it all comes down to pure rage quitting where your team would lose anyway. But a player leaving for no apparent reason whatsoever can always happen and there is no reason to screw up such a game further by having the AI feeding gold to the enemy.

It's here:

Basically, what he's saying is that it's better to fix the actual problems instead of swatting at symptoms like this idea is doing.

I agree.

I think the OP's idea.  The ai should ether give a small amount of gold and xp or nothing depending on the results of playtesting.

I'm shocked that there isn't universal agreement on this.  Such is the nature of opinion, right? But nonetheless, I'm sure that anyone disagreeing with the suggestion at hand must realize how this will eliminate the AI feeding problems (and eliminating the feeding problem would make any game which was unfortunately made unfair due to incoming AI all the sudden a bit more reasonable to deal with).  Therefore, your disagreement confounds me.  And though any team stuck with AI will essentially be short-handed for the rest of a match and thus at a disadvantage, they won't be at any further disadvantage because of AI feeding issues.  I cannot imagine this fix being substantially time or resource consuming (maybe I am wrong, though; nevertheless, I still feel it important enough).  If anything, it will nicely complement all of the rules being set to deal with rage-quitting.  Must it really be one or the other?!  No, it need not be.  To ensure our utmost gaming experience, I am confident that Stardock will push this through.  This is necessary. 

 - andpancakes

My suggest is a change of the killing gold alltogether.
Each DG has a gold pot that fills slowly over time. If he dies the killer and assisting DGs get the gold, and the pot is reset. This would solve the AI farming and the farming of weak DGs, and also lead the focus to killing strong DGs (whose pot is still full). The pot also should have an upper limit. Not that you get like 10k if you kill one DG after like 20 mins of gameplay.

signed

please reduce the AI farming

I like your idea but I would put the reward for killing a comp down to 50% of reg xp and gold.

I think in a perfect world that the programmers would like to think that the AI plays at about the level of an average human, so that would be an obstacle to this sort of idea...but I don't really think it would be that bad of an idea to have the AI only reward a fraction of what a human death would. Its so easy to trick the ai into thinking it doesn't need to run until its too late...

The obvious solution is smarter AI  though . Probly not gonna see it till mod support is available though.

Disconnect penalties will not "ensure" that no player will ever leave for whatever reason. Quits will still happen.

But it is a problem at this time and as already pointed out, there is no certainty about whether or not the upcoming changes will reduce the amount of quitters so significantly, that it won't be a problem anymore. All you are saying right now is, that AI farming is fine.

And it's not like Stardock or GPG never intended for the AI to be in. The AIs are deliberately put in.

Look, in the games I play the absolute overwhelming percentage of the time quits are because of rage or connection issues.  Rage is being dealt with; throw that out the window.  Connection issues are also being dealt with (global fixes, ban players with known problems in your list).  So all that's left are "miscellaneous" walk-the-dog type of quits that happen so insignificantly they aren't worth attention at this time IMO.

I am not saying AI farming is fine; I'm saying this is hackish band-aid solution to a problem that, as you've stated, we're not certain will be nearly as pressing an issue once existing solutions are deployed.  We can reevaluate at that time.  When they do reevaluate, most likely they will implement the non-hackish intuitive more direct and customizeable solution I provided.

This is another great solution to a big problem. Either a. remove the AI from taking over, which has been suggested many times, or b. this, which is a reall good idea. No feeding on silly AI to get easy wins. Like someone else said, its a big enough advantage as it is. Fed on a stupid human player, who quit, leaving 3 v 2 (+ AI). Thats already a huge advantage so no, no reward at all for killing the AI.

Please, GPG, do something about this, for the love of god.

Heh, more like for the love of GOLD/EXP! lol 

Your solution would encourage kamikaze behavior from players though. It might not be too bad for most demigods, but this would be a terrible solution once Oak is in the game. Oak is already a problem on some maps and game modes with Last Stand upgraded.

I do like the overall concept of reducing gold from AI kills.

May I remind you that I have had numerous games that people quits due to them lagging. As well as some people that quitted for no apparent reason, e.g. shortly before 1st kill. Any anti-rage quit measures are unlikely to be 100% effective either. AI will play a part in the game. What you have proposed is not 100% fail safe.

A handicap option is not bad as you suggested. I actually would like to see that implemented. Having options is always good. However I would like the default value to not give any god nor exp to players.

This is true.  It is a very basic, realistic and practical fact.  Thus, our argument stands as long as AI exist in this game, and no one can deny that they will always exist in this game.  Despite any future upgrades/improvements to the artificial intelligence in Demigod - whether cutting edge upgrades or not - AI will always be predictable and easily exploitable for the purposes of men (okay, whoa, maybe that was too deep for the purposes of this thread).  And in this case, the purposes of men (and women :)) are GOLD/EXP! 

This idea has been around almost since the 3rd week the game was out.  RQ, DCs, walk the dog, (screaming kids/wife), you will always lose a player.  AI feeding is the reason people dominoe-RQ after an AI takes over.  Reducing or elimiating the gold/xp an AI replacement gives is the simplest way to fix this.  It changes the AI from a cash-cow for the other team to an acutal ally that may even help in some situations.

You may, but I don't need the reminder because I've already addressed this.

If it's going to lag, it's probably going to lag at the start of the match and persist throughout.  That means people quit (at the start of the match), which as I said means we can remake the game.  Devs say in the next patch if people quit at the start (I believe first 5 minutes) of the game, the game will not be counted in their stats so this addresses the lagging and player dropping at start.

In addition, I believe there is a feature being implemented like concede but more of a "global" concede where people can just call off the match.  Another solution to games lagging.

There are two main problems:

1) Sometimes people don't quit after they 'die a lot'. Some people quit after maybe their second death in game, and bitching about it to all their friends. This may be 5 or more minutes into it, and turn a reasonable game into something unwinnable.

2) Sometimes people have client side connection problems.  Their network kicks out for 10 seconds, and they get disconnected.  It'd be nice if this always happened in the first minute or two of the game, but it doesn't. It would be nice if this didn't severely penalize the team who had the disconnect.

3) Lastly, the argument of developer time is pretty moot.  We aren't suggesting some incredibly huge system change, but the injection of likely two or three lines of code:

What will take longer is to determine whether such a change is important enough to make. And that is the discussion we are having here.

Sly_Squash, I invite you to quote me and comment on my thoughts and opinions expressed in this thread.  I feel as if you are avoiding my arguments.  I feel that I have sound reason backing my ideas.  So, please, what do you think?  Read my posts in this thread.

EDIT: Zechnophobe, thanks for the insight about coding.  That's interesting.  And, if anything, it should help people realize that the changes being discussed in this thread (and, in fact, the purpose of this thread) are very practical and could be conveniently implemented.

- andpancakes

What unique arguments have you made that I've not already debunked elsewhere?