Explorations at Agincourt

After playing Agincourt three times I was ready for a change, and wanted to see how pathfinding worked in more difficult areas. So, my army went wandering.

1 My first don't-just-stand-there battle was to pull back to the hill behind the English initial starting position and defend it. I ended up with VERY few casualties except for two units of archers which I had to sacrifice to keep the French off my back while setting up.

Conclusion: the combination of steep hill and archer fire made most of the French units rout within two minutes of initial contact. Fighting downhill did seem to offer very significant hit & kill rate benefits, while fighting uphill (which I did just to try it) was very hard and without flanking I would have lost a unit. Good news so far.

2 The second was to pull back and go a bit to the left of the hill, to the small farm. There I set up amongst the walls, with my archers in the middle and infantry plugging the wall holes.

Conclusion: pathfinding in the fairly complex walls environment was excellent. The AI actually tried attacking from two angles, quite the surprise because of the heavy scripting. That may simply have been luck though with the way I arranged my troops. Finally, once I got the French stopped just outside the walls of the farm and they bunched up into a horde, the archers cut them down more effectively than when the French were on the run.

3 Third, I pushed my troops to run all the way up to the cliffs by the windmills, then lined my archers up along the cliff face and put infantry at both ends to protect them. The AI tried again attacking from both sides of the cliff at once with all three waves of infantry.

Conclusion: longbows, though they have good range from flat ground, have truly phenomenal range from a height like that. I'll have to estimate distance difference at some point but it was quite impressive. Also, pathfinding around the cliff sides was again very good, no weird lines against which it looked like troops were squished. Fighting downhill again seemed to give a distinct advantage. And, finally, French units fall very nicely off of clifftops ~:)

4 Fourth, I pulled my troops up to the clifftop and then ran them all the way over to Agincourt town where they set up inside the town itself. I lost a unit of knights and one of bowmen during the run because they were flanked by the charging French. When I got set up in the town, my men were "Tired", but the French were as well so we still won.

Conclusion: pathfinding in the town was flawless, which was a huge relief. Fatigue does not seem to accumulate quite as quickly as in MTW. It is difficult to tell what effect it has because both armies were equally fatiged at engagement.

All in all, it's been great fun to go "off script" with Agincourt. These maps may be only slightly larger than RTW's, but they feel huge with the variety of objects and terrain in them. I have yet to explore behind the French starting position, the church on the hill north of Agincourt town, and quite a bit of territory inbetween.

I imagine one could do the same thing at Pavia, but the enemy are so much closer it would probably be quite difficult to pull of well.

Thats some good news. Hope to hear about more of your findings.

Aye good job there Tamur, appreciate the info

Im gonna go exploring myself now :)

Good news indeed!

I appreciate the info. glad someone had the imagination to explore in such manner.

Wishazu, let us know what you find! There are a lot of scenarios that could be played out with such an aggressive enemy.

Nice man, I dont know how you got to Ag and had time to set up b4 they caught you! :P

He sacraficed two archer units to keep them at bay whilst setting up, as he said in his post :P

Well done Tamur,
Some good work

Yah, great stuff, if the final game shows that kind of AI/pathfinding, I'll be pretty excited.

Though, I have to say, when it considers itself sufficiently strong in RTW (ie similar to the situation of Agincourt), the AI will come after you with great force as described & can cause some really good battles on the RTW engine so I'm still cautious about this till I can see a more balanced battle with no scripting.

Good info.

Quote:

---

Originally Posted by hoom

...so I'm still cautious about this till I can see a more balanced battle with no scripting.

---

Exactly. My runs with different scenarios did not prove (or disprove) anything about the AI. The more I played it, the more I was sure that the French units were scripted to attack specific English units, and therefore they tried multiple flanks at once, charged into the stakes, &c.

Still, all the other things (such as pathfinding and terrain) that I was able to test without too much bias seemed very good, so I'm cautious but hopeful as well.

Great work Tamur, some potentially very interesting insights there. It'll be interesting to see just to what extent the historical battles are scripted.

(Also I'd point out to anyone who missed it that the English have extremely experience compared to the French in this battle, which may be a large part of why it's so difficult to lose.)

Antagonist

Ahhh good nice to know PF/Ai is good:juggle2:

At Agincourt i have tryed retreating to the big hill after the initial cavelry charge and then i just sat there. That was a very easy battle, another time i attempted to retreat to the cliff but the flanking french cavelry attacked and the rest of the french caught up resulting in a bit off a massacer. I still took out a resonble ammount of french though

In Pavia i retreated up the wooded hill, but that was kinda a unconcious dissicion. Still won me the battle. I want to try retreating to the hill to the side off the encampment out of range of the artillary so i can play my game with out having to waste my cavelry on the artilallry after the first cavelry wave, still got to try that though.

Pavia is a harder battle so I could see why you may not want to sacrifice to units just to reposition yourself, espically with the cannons :laugh4: