Progress report as of July 12

If anyone is interested, here's a description of how the actual scripting is going. Also, if any scripters have any ideas on how to solve any of the problems encountered, I'd like to know!

My goal right now is to create a system that can depict realistic and complex military campaigns, taking place at the same time-scale as the historic campaigns. A large army moving through enemy territory might send out scouts as it goes. Small enemy bands might hover behind it picking off stragglers or ambushing foragers, trying to whittle down the force or prevent it from replenishing its supplies.

I think all of this is possible to script. However, it will probably make for a more slow-paced game than Native, when at times there is relatively little for player characters to do. Players might have to be prepared to spend some time watching the action unfold around them. Hopefully the action will at least be interesting to watch. Consider this all something of an experiment!

I received the Python files a little early, and spent a few weeks experimenting with hiding, ambushing, and camping.

In order to allow the AI to make complex decisions that are part of a larger strategic plan, AI parties have three different kinds of behavior -- tactical (the AI's reaction to immediate threats), operational (representing logistical, route-planning, or other somewhat less immediate decisions), and strategic (meaning the party's overall goal).

Most of my work so far has been done on the tactical AI. I set the spotting range for the game'e internal engine to zero. This means that spotting, and decisions to pursue and flee, are now handled by the scripts. In order to avoid too stiff a penalty to FPS, I spread out the spotting routines so that the engine only handles spot checks for a handful of parties at a time.

Both the player and the AI parties have the opportunity to hide if they see an enemy. The AI will make this decision if it has a good chance of hiding. Ideally, I'd like the parties to hide if they calculate that they have a better chance of hiding then escaping, but I can't get the speed of the relative parties that would allow me to do this.

Bandits and other parties will also go on ambush themselves -- scouting for a bit of good terrain on an elevated hill (elevation now affects spotting) and hiding there, waiting for enemies to come by, and then charging down on them.

If one side in a battle has not been spotted, then it gets to ambush the opponent. These work, sort of. The enemy marches in column past your position, without charging or attacking. (However, they do turn to look at you as they march).  If you attack, they will eventually realize your presence and react. Sometimes it doesn't feel quite right, but sometimes it yields fun results -- like when you charge a column of horsemen, and they start to react only a second or so before you slam into them.

In theory, the player should also be able to be ambushed -- although I have not tried out of a way to keep the player himself from running off and attacking the ambushing force by himself while the rest of his party plods on, unaware of the danger. I could presumably keep a one-way barrier between the player and the enemy until such time as the ambush is sprung, to block both missile fire and movement. Any thoughts on this would be much appreciated.

At some point, I would like to add in a version of Extended Gameplay's morale. It seems to work really well, and one big advantage of an ambush should be the morale impact on the defenders.

I would really like to be able to add baggage animals to forces on the battlefield, particularly the columns of troops who get ambushed. I was able to do this in an earlier version about a year or so ago, but unfortunately now I cannot get the spawn_horse operation to work. I also do not know if I can get riderless horses to move.

I would also like to do nighttime camp attacks, but I have not yet figured out a good way to implement them. I don't think spawn_agent works yet, or presumably I could have characters spawn near a tent or campfire as they wake up. However, this would take a lot of work, and it also should be very difficult for a player to pull off successfully. I intend to attach a lot of penalties to moving at night -- you might have troops will get lost, for example -- so that the default behavior should be to camp and fast forward through the hours of darkness.

I have also introduced forced marching and fatigue to the world map. Parties that flee from another party or pursue another party will move much more quickly, but will also begin to tire out. Fatigued parties will rest, and at night they will camp.

Anyway, I am fairly satisfied with these scripts, for the time being. I am now moving on to the basics of the strategic and operational systems, but I will be returning to the tactical stuff later as the game moves along.

I'd like to employ formations at some point. A good mod for checking out how they work would be Hegemony 268 B.C., no?

I haven't worked too much on the experience of larger battles, yet. I had imagined that the player might be involved in one part of a battle, while receiving occasional messages on how the rest is going ("friendly troops to your right appear to be giving way", etc). At certain points you might receive a rush of friendly or enemy reinforcements, or be able to intercept a melee in progress.

This might not be possible to do as I had originally envisioned, as I am not sure that I can pick and choose which elements of a party spawn at the beginning of battle. This means that it might be tricky to have the player fight alongside Lords X and Y on the left wing of the battle, for example, or have the survivors of the center charge into the wings. I might still find a way, however -- possibly by relocating parts of a party group to other temporary parties -- but i will have to do a fair amount of experimentation.

Thanks to all for the kind words. Grailknighthero -- setting team relations is a good idea. However, I think that, like faction relations, team relations are always mutual -- ie, if I set it to neutral, then both sides think of the other as neutral. This would porevent the ambushers from attacking the defenders.

Right now, I'm preventing the defenders from fighting back by setting their weapons to blunt only, and then by setting all the game's blunt weapons to piercing damage. (This means that you cannot deliberately knock someone out with blunt weapons, but I prefer that anyway). The defenders keep their weapons sheathed until they are alerted. They are also set to march across the map in front of the attackers at a walking pace. It mostly works, except that the supposedly unaware defenders turn to look at the attackers as they walk, which spoils the visual effect somewhat.

Also, AI missile troops don't seem to lead, so even skilled archers tend to miss targets moving at an angle to their line of fire. This might be realistic, however, as I cannot think offhand of any pre-gunpowder battles in which the defenders were simply mowed down from hiding, although I could easily be wrong about this. I suspect that the main value of an ambush would be the psychological shock, and the ability of the attacker to engage a small part of the defender's force as the rest fails to react. Any historical examples would be highly welcome!

I look forward to seeing how the multiple formation script works. If one could have multiple infantry formations, that would be great -- particularly for the Byzantines, who seem to have maintained a greater sense of unit identity. I'd really appreciate seeing the scripts!

I mean setting the team relations to neutral for a limited time, until the enemies are a certain distance apart, then setting them to enemies.  Hopefully then when they walk past you, they wont look at you.  I might not be understanding what you want either.  Maybe have the teams start out as neutral, the player and his troops hiding in a forest and the enemy marching by.  When the player clicks a certain button, it is the signal for the ambush to happen, and the team relations are set to -1 and the player's troops charge. 

Actually that's quite a good idea. The teams could be neutral until the charge order is given, and then the defender would still be blunt-weapons-only until alerted. I might give that a try.

place an ambushing party on the map BEFORE the PC's party. and when PC and his party are spawned on the map they are immidiately attacked by ambushers (let's assume they spring out of the bushes) and have no time to form ranks or somesuch

I think that actually is possible, in a way. However, it would make an AI ambush different (and more effective) than a player-initiated ambush, which is an asymmetry I had wanted to avoid.

Another thing I was going to try was robberies, where the attacker forms in a line ahead of the defender party (the attacker is the top of a "T", the defender is the stem) at a relatively close range, giving them a slight advantage in combat. If I can do it, the "battle" would begin in a dialogue, and then the teams get set to neutral depending the player's responses.

On formations, I really haven't worked at all with them, so I don't know what is possible. I will try to keep an eye on what other mods (like Hegemony 268 B.C.) do, and probably work from there.

A way to make waiting less boring could be to implement activities that take time on the map. For example, sometimes I wait in the training fields while training my troops, this way I am doing someting and time passes on the world map.

You would still have to fast-forward the action on the world map to make time pass, unfortunately. One could however consider the time to be spent in training, which could serve as a multiplier to the experience won from a sparring match, or have time spent waiting accumulate to other long-term projects, like Native's book reading.

Adding complexity to fief management is part of the plan.

I haven't come to any conclusions about this, but ideas that came to mind would be...

1) Internal conflicts in the fief that have to be resolved
2) External conflicts with neighboring villages
3) Projects, including religious buildings, charitable endowments, defenses, economic additions (like mills), and probably most importantly, land reclamation and irrigation. If anyone knows much about medieval irrigation, or farming in mountainous areas, input would be very welcome! Also, if anyone can think of ways to make project building more interesting and complex than simply allocate money/see results, I'd be quite interested.