Hey folks, Let's put this to a vote! We've got some strong opinions and the Forum's blown up on this one, but let's gather the opinions of a larger pool. Some have suggested prioritize the Witch Doctor by number of troop (Option 1). Others have said by total damage of all troops in a group (Option 2). A few have said keep current AI, which prioritizes by total HP. (Option 3). This may not be a democracy, it is Midoki's choice to implement, but hopefully this is constructive. There may be even other options, feel free to reply below and keep this thread as the latest update and we should get a big response!
I've liked several of the ideas offered in various threads, but I think the simplest solution is to pretext the largest group of living pirates.
I like that, good idea. Or if that is somehow hard to code or whatever, they could offer two different WD's to recruit. 1 being the old way and the other the new goof ball way
I liked that option the most from other threads, but left it out as might be more complex to implement. That and if it is more complex than Midoki would prefer to go now, the survey would be devalued (that would likely be most favored option). I think player selected options have been suggested even for other troops. Who knows, if implemented maybe this will usher in a new era of fun, popularity, and accomodation of the widest variety of playing styles. Bombers who can be selected to favor walls or favor closest target after triggering Kaboom. Could be great!
+1 to Raptelan brilliant idea - just to add to it of you recruit more than one WD could we toggle different preferences for each of the four WD available, best of all world in one battle
Switch toggle would be great. But if I could choose only one, by largest total group number it is. But make him stick for some time not wander around like a bee. And thank you @Becker Redbeard for making this poll.
Just let us click on the troops we want him to protect. What ever troops within a radius of where we click on the island, the wd prioritizes. Simple. Then have the default ai for when no clicks or troops die.
I think the WD should protect the largest group that it is nearest too. I voted for option 1 but I just want to clarify how I envisioned it because I'm assuming most people deploy their WD near the troops they want it to protect. Just wanted to put that out there so the WD doesn't go running off halfway across the island once a gunner dies and another group is bigger.
WD is a WD, the toggle would affect how all WD's (including on Guild Ship) would behave in battle. Keep it simple.
If you target by any one variable, then how would you use two WD's if you want to protect two different groups of pirates, as both WD's will follow the logic and rush to protect the same group. For example if I want to protect a group of gunners on one side of a base with one WD, and have 3 HG protected by a WD on another side. This would be impossible if controlled by a single battle variable. I think the earlier suggestion of protecting an identified group has good merit. You select a deployed WD and you circle the group you want him to protect (or resurrect), and of he goes and stays with as many as stay within his boundary. You could then select another WD and repeat for a different group. Over the duration of a battle you could reassign a WD as often as you want, giving greater control over how you protect troops at different stages. True control...
Why da heck you include bombers in considering "total amount of damage", Becker? They move too fast and unpredictable. Honoured majority of voted, your basic "total amount of units" won't work good enough. Three gunners will be more preferable for WD than two HGs, all WDs will move to one point (thats why simple "total amount of damage" must be tuned also).
What really needs to be done is some sort of balance between hp and number of pirates, with a distance penalty that grows exponentially with how many tiles it has to walk, so that its much less likely to pick a different group far away. I personally prefer mine to protect my juggs, but I know a lot of people prefer them to protect their gunners, and there are those that want one wd protecting each group. Then if you deploy it with a high hp group, it will stay with that high hp group rather than running away, and if you deploy it with a high number of pirate group, it stays with that high number of pirate group, instead of running away. It could probably heuristically determine the individual preference based on what sort of group is nearby when you deploy it.
Another option is to have midoki run a million simulation fights, with the WD preferences tied up to a machine learning algorithm (NEAT is easy and would work well here), and then the WD would be able to genetically evolve their own optimal targeting.
Firstly thanks to @Becker Redbeard for putting this poll together. It's great to have all your thoughts laid out here in a constructive and positive manner. Not to get on the 'duplicate thread' high horse again, but please try and refrain from creating the same thread over and over again..... it's almost impossible to keep track of everyone's ideas. Anyway, back to the WD. You have to understand that we don't randomly go through the game trying to break things. Our goal is to only make things better, and that obviously means change. Not everyone likes change as you have pointed out on numerous occasions Anyway back to the WD We have an incredibly experienced guy here writing our AI code and any change is not under taken lightly. A lot of thought is put into how the AI works to gain the best result for the pirates as they make their way across the island. A key element is not only what am I going to gain by doing this, but also what is the cost of me doing this. For example it may be better to protect 3 guys right next to me than make my way right across the island to protect 4 guys. So, the largest group is not always the best choice to make. Also how is that group made up? Is it better to save 3 gunners than 4 bucs? Also how much health do they currently have? Triage is all about deciding the order of treatment. Is it better to save someone in close proximity, or travel a greater distance to potentially save 2? And it goes on. There are many many things that need to be weighed up when the WD makes a decision, and once you add more WDs this process gets even more complicated. And all that complication still needs to run on an iPhone 4. This isn't to say the values can't be tweaked and they are, constantly. We're are listening, don't worry Our AI guy also suggested I post the following. Our system is based heavily on John Nash's theories and so, if you have a bit of time on your hands, why not have a read. https://en.wikipedia.org/wiki/Nash_equilibrium#Coordination_game