Pathfinding

Age of Empires II: DE II - Discussion
41

@Avocet so I’m not trying to be pedantic, but I think it is worthwhile to make the distinction here. We can not solve a problem if we don’t understand the nature of them problem.

I don’t think the pathing is “bad” per se. If you look at collection rates for vills on a woodline, or trade carts or trade cogs, from HD to DE there are usually substantial increases. approximately 10%. So DE’s pathfinding when tested in these kinds of ways is better than HD. Definitely not perfect, always room for improvement. In fact right now there’s a bug where units sometimes get stuck you you move them in a group. So definitely not perfect.

But when you have a unit going from x to y, DE on average is better than HD.

What I think is more accurate to say is this video shows that when we’re talking about groups of units HD was approaching/prioritizing differently than DE does.

I hadn’t realized there were these differences, ### ##eing HD and DE compared in this video, the most obvious things i noticed were:

  1. homogeneous groups when commanded to a new waypoint, won’t turn around individually as units but rather u-turns as a group.
  2. heterogeneous groups when commanded to a new waypoint, all units travel at the same speed to get into position, instead of allowing faster units to go faster.

These two just make your units feel less responsive. However this third I think is most elucidating

  1. a group of units, when tasked to waypoint, if not in formation, more strongly prioritizes getting into a formation than moving towards the way point, which differs from DE which prioritized moving towards the waypoint more and being content with getting units in formation on the way if it could. This third one I what I think causes the most issues. when pathing works like intended it’s not too bad, but when pathing is sub par you get units going sometimes the opposite way you tried commanded them cause they really wanna get into position first.

This sets up the crux of my argument. This video really opened my eyes. I don’t think DE has “bad” pathing, at least not compared to “HD”. The execution of the intended priorities is better in DE than HD.

However I think HD conceptually had it right. AoE is not total war. 99% of the time, the precise arrangement of units isn’t as important as getting units where they are supposed to be. AND if you really really need them to be in that formation before going, you can give them a nearer waypoint for them to go toward, get into position, THEN go into battle. HD prioritized the going. HD says “You want your units there? I’ll get them there, may not be in a pretty formation when they’re there, but they’ll get there fast.” DE says, “You want your units there? Well I gotta get them in formation first. They won’t get there fast but they’ll look good while doing it.”

I look at some of these changes and honestly I can see where the thought process is. However, seeing the comparison, having played both versions, I think it can be said that HD pathing had the right priorities.

For a long time the fans have said “make pathing better” and diligently the devs tweak here, tweak there, optimize this, improve that, but fundamentally the units aren’t doing what we want them to, because the game isn’t TRYING to do we want it to do.

We want the units to GO. DE wants to make units get in formation. I know there were some exploits with unit speed in formations in HD, but if the devs took a good long hard look at the HD pathing from a conceptual level, and brought that back into DE, I think we’d think pathing had improved a good deal.

In summary, I think it is incorrect to say DE’s pathfinding is bad (it’s not great but it’s better than HD’s). However, DE uses it improved pathfinding to prioritize a fundamentally incorrect objective, namely getting units into formation instead of getting them towards the waypoint.

6 Likes
43

That’s one hell of a post.

While I think the formation thing certainly explains a lot about how units are, moving around. It does not explain how units are taking longer routes than needed, especially villagers. With the building tests at the end. A villager should not be walking twice as far to get to a location. I very much am sympathetic to the game being old and having it’s, quirks. Something is very much off when the pathing has come to such a horribly indirect path than even casual players will probably pick up on a villager walking a excessive distance.

I think reverting the way grouping works to HD instead of DE would be a quick fix for military units.

As for villagers walking double the distance to build something. That could be a quick fix or a long one, I do not know. It could be a calculation is off slightly or a series of bugs have cropped up in how the game finds a route.

1 Like
44

this issue has been going on for awhile most notably trebs. dont need to move one by one, next time it happens just press stop/hold and u can move them all again.

1 Like
45

You are so late to the party. It’s been there for at least half a year: Units don't move! When I gave them move commands 1 million times

Excuse me? What are you even trying to express?.. “DE pathing is good! Just in a wrong way.”

And show us an example of HD villager not walking straight line?

1 Like
46

Agreed. I think that specifically is a bug. Seems to happen mostly with vills constructing buildings and the result is just soooo bad when bad pathing happens.

2 Likes
47

@ Quasibrodo

I actually read your entire post, and firstly, let me say thanks for the thoughtfulness put into it.

There are things I will challenge your ideas on here. For example, the idea of

We can not solve a problem if we don’t understand the nature of them problem.

We not exactly solving any problems here. We are not devs. We are customers, consumers, giving direct negative feedback to the dev saying they have screwed up. Fix it.
It’s like a car owner experiencing issue with his car, say it’s not starting up. He’s not trying to solve the problem by taking it to the mechanic. He’s giving the mechanic feedbacks. The car is not working; it’s broken, please fix it.
So does the car owner have to have fully understand the nature of his car’s issue?

I’m not sure if you have watched the entire video of the link I provided, made by Survivalist, big credit to him. But at the end, you can see that in AOE DE version, there’s still some pathing issue with individual units like vils and choosing to take the long route to go build a building. This is just ONE examples of the nuances we are experiencing.

Although there are some improvement to pathing, like knights are better at targeting units in a group fight in DE version, pathing has predominantly not improved but made worse.
Yes, you hit the nail on the head thee when you said DE has their priority wrong with choosing to group in formation first before actually moving. It’s causing a lot of grief. This formation mechanic IS part of the pathing issue. It’s making units move backward, or do a U turn, or sideway, before actually going directly to where you are right clicking. And the idea is of good pathing is unit being able to be responsive and move directly to where you are trying to get them to go.

In short, pathing will probably never be perfect. I do not have such expectation for this game. But at the very least, the dev should really test out these new “improved” changes before they roll it out or at least listen to the customers. And overwhelmingly, many of us long time devoted AOE 2 players are saying the way our units are moving right now sucks. There’s just no defending the horrendous U turn, shuffling, and even vils choosing to take the long path to go do a task.

5 Likes
48

ya that was the voobly way, we just need that back. Look here in this video how nicely the units that are further apart group up without any unit moving backwards foolishly in voobly.

49

I would really hope that shuffling issue would be addressed. HD grouping pathing looks so nice compared to DE.

1 Like
50

I think it’s a bit of a stretch to say that ‘no one has mentioned pathing’.

All the pros, without exception, have complained on their livestreams.

Here is a gem also https://www.youtube.com/@AgeofPathing

3 Likes
51

True, however typically consumers are garbage at giving feedback and companies are terrible at interpreting it. It’s a bit allegorical, but if you’d polled consumers around 1905 about what they’d like different about their horse and buggy, I assure you extremely few would have said “actually I think the horse and buggy is fundamentally inferior mode of transportation, and we should re-design society around a cheaply produced internal combustion engine automobiles which overall will provide higher utility.” And yet by 1930 nearly everyone had switched to an internal combustion engine. Everyone wanted the utility provided by an internal combustion engine, but could only express it as some modest incremental improvement of the horse and buggy. Human brains don’t think in step-changes, they don’t think in first-principles.

I don’t want horse-and-buggy/group-then-go pathfinding, I don’t want modestly improved, iteratively changed horse-and-buggy/group-then-go pathfinding. We need step-change internal-combustion-engine/go-then-group pathfinding. If we don’t as a community of fans and developers, understand this then we will have horse-and-buggy pathfinding that is iterated until the end of time, and it will still not provide the step-change improvement of utility we all desire. We want internal-combustion-engine pathfinding and we need to stop describing it as fixes to the horse-and-buggy pathfinding.

In summary to this specific point, while yes I don’t believe it’s the “responsibility” of us as players to have to game-design conceptually optimal pathfinding, DE pathfinding has operated under this paradigm for four years and only now seems to be better understood by the fans thanks to this video, I see no good reason to wait and just hope and see if the devs get the right message from this video. I’d prefer to lay it out, clear as day. I want to be the horse-and-buggy owner in 1905 saying that I’d prefer an internal-combustion-engine.

I did see that part of the video. Yes there are bugs in DE’s pathfinding, however I think that’s a distraction to understanding the nature of the pathfinding discontent. obviously bugs are bad and should be fixed. Saying bugs should be fixed requires no insight and if every bug were to be fixed, we’d still have these fundamental design choices that leave units feeling less responsive. Also why should we spend time fixing bugs if we’d all prefer they re-design pathing anyway? Re-design, implement, THEN fix any bugs with THAT. STOP tweaking the horse-and-buggy and work on the internal-combustion-engine.

I hadn’t run into it until after the mountain royals came out. My invitation to the bug party must have gotten lost over the internet somewhere lol. I’ve run into it a couple times since tho.

More or less. I wouldn’t go so far as to say DE’s pathfinding is “good”. I acknowledge there is the units getting stuck bug and the vills pathing to building bug. Those seem to me are two specific bugs. Bugs aside tho, I do believe that DE has generally superior pathfinding than HD, though not to the level of “good”. I believe comparative collection rate tests on woodlines and trade routes prove that DE is an improvement. So then I have to ask myself, why does it feel worse?

I am a developer and I know there have been times where I have some feature I need to add, I mentally come up with a solution, I then code that solution, and even w/o bugs I’m not as happy with that finished solution as I’d anticipated when i’d conceived that solution. My code executed my intended solution, but I then realize my intended solution was sub-optimal and I needed to come up with a better solution. My code was good in a wrong way.

Similarly it seems to me, that when DE’s pathing is working (which I admit it isn’t always there are definitely bugs that need to be addressed) but when it is, it is properly executing the intention of the DE developers which seems to be a group, get into formation, then go solution which having seen HD and DE compared seems to be a sub-optimal solution when compared to the go and get into formation on the way solution devised by Ensemble.

1 Like
52

I think you’re fundamentally overstating the role of customer feedback by implying that customers should be the ones on the cutting edge of innovation and somehow predict or request disruptive technologies in their feedback. Some visionary people will always do this anyway, but it’s definitely not a good standard for average feedback. I see what you’re getting at in the sense that “customers didn’t know what they wanted yet” when retroactively referring to life before the invention or adoption of a disruptive technology. But AoE2:DE customers asking for improved pathing know what they want. It’s really not much more complicated than GO, as you’ve said.

Overall customer feedback is extremely simple: judge the product at how well it performs its intended function and relative to other (known) options. When something new comes on the market, it joins the pool of known options. The mass adoption of automobiles over horse and buggy is an example of customers being good at giving feedback by responding powerfully to an incentive. But before that, any given horse and buggy was judged on how well it performed against an ideal horse and buggy, and against other existing options. Not whether it was the most theoretically optimal possible transportation device of all time.

Similarly, pathing in DE is judged by how well it stacks up against the pathing known to exist in other games and in earlier versions of AoE2. The vast majority of things people refer to when they talk about bad pathing in DE are things that were better at some point in the past. The rest, better in other games. If DE would just fix the regressions and obvious bugs that have accumulated while maintaining its gains (e.g. lumberjack efficiency), pathing would be very good and I think almost all of the complaints would dry up. I really don’t see a need to fundamentally reinvent pathing in a way that’s at all comparable to the ICE/Automobile Revolution, and I have no idea what you mean when you say that, other than perhaps just getting carried away with an analogy.

1 Like
53

Perhaps what I said wasn’t sufficiently clear but what I was saying was the opposite of what you seem to have understood I was saying.

It’s not my “job” to conceptualize better DE pathing, but I also feel there is no advantage to anyone if I were to instead of fully articulating these insights such that there is no ambiguity as to the desired remediation, I just said “make pathing better” cause that is what is expected of me as a consumer.

I think you are conflating the simplicity of conceptualizing a given solution with the requisite insight to ponder a specific solution to a given problem in the first place.

There’s a clip out there somewhere, Neil DeGrasse Tyson is speaking with…Joe Rogan I think, doesn’t matter. When the space shuttle was being tested, specifically it’s landing, it’s coming in pretty fast compared to normal ariplanes. It was laterally gliding all over the run way. They decided to try adding grooves to the runway so the rubber tires would mold into the grooves so the wheels wouldn’t glide laterally. It worked so well, it’s now used on highways on tight off ramps.

In my estimation, the concept of adding grooves to pavement to increase traction and reduce sliding is something so simple, I could explain it to a kindergartener and they’d understand.

We’d been driving on highways for decades at that point. We had all the time in the world to come up with this super simple to understand solution.

And yet it took a national program, full of literal rocket scientists and billions of dollars, before this simple solution was even pondered.

Yes, make units go first and group up second, is so simple that you can explain it to anyone and they’d understand immediately. And yet before the release of this video, I haven’t once seen ANYONE express the issues with pathfinding with 1/10th the insight as this video.

I know this insight wasn’t rattling around in my head before this video. I know I wasn’t thinking “DE group pathing is operating on fundamentally incorrect priorities, and if they priortized going it’d be so much better”. And I don’t believe myself to be fundamentally smarter or more insightful than the FE devs.

My guess is 4-5 years ago when FE was working on DE they made some fundamental changes to pathing, which made some things better (like vills on a woodline, trade routes (which are probably the easiest to test empirically), knights chasing down x-bows) and made somethings worse (dispersed units going to a waypoint, homogeneous and heterogeneous unit comps performing a U turn) when we now, years later, say “make pathing better” it is not readily apparently what in the vast possibility space of path finding situations is needed to be improved when the problem isn’t some obvious bug (units getting stuck, lone vill going the long way to construct a building).

54

I meant specifically this forum. Before I made this topic any related to pathing were many months or years old. So not current.

55

Firstly, thank you for the well thought through reply. I have this very same opinion in mind, but could not articulate it as well as you did. So you basically have put two in for us both in replying to @Quasibrodo previous comment about customers are garbage at giving feedbacks.
Well said SirWiedreich.

Similarly, pathing in DE is judged by how well it stacks up against the pathing known to exist in other games and in earlier versions of AoE2. The vast majority of things people refer to when they talk about bad pathing in DE are things that were better at some point in the past. The rest, better in other games.

That’s exactly it, spot on. The only way people compare anything in life is to have a different experience to compare it with, be it from previous experience from the same game or a different game.
Many of us have played A2D since the AOK and knows things were better back then with archer micro and group formation. We have also played many other RTS games like SC1, WC3, SC2. Pathing are better in those games, and it is especially true with how responsive the unit control is. So yea, that’s something for @ Quasibrodo to consider.

1 Like
56

Okay, so there’s a lot to be said here.
First about the “however typically consumers are garbage at giving feedback and companies are terrible at interpreting it.” That’s sort of answered from @SirWiedreich 's reply. What I like to reemphasize on is that that statement isn’t always true. It is definitely not true with A2DE here. Consumers/customers are giving very direct, specific feedbacks. They are saying pathing sucks with statements such as:
“units walk the longer path.”
“if you attack move, your units will shuffle.”
“If you focus fire, your army will do this little jiggle and die.”
“Group formation makes your unit move backward instead of going forward to where you are rallying, it’s terrible.”
“My units are doing this U turn when I try to get them out of the way of a mangonel shot and they end up getting hit.”

Those are very specific feedbacks. If this game, or any game, wants to be successful, they need to learn how to listen to their consumer’s feedback. Usually companies do survey and find ways to figure out what people want. But in this case, it’s served on a silver platter for them. It’s ONLY complaint about on every forum. “PATHING SUCKS!!” backed up with videos, examples, and streamer’s complaints.

Perhaps what I said wasn’t sufficiently clear but what I was saying was the opposite of what you seem to have understood I was saying.

It’s not my “job” to conceptualize better DE pathing, but I also feel there is no advantage to anyone if I were to instead of fully articulating these insights such that there is no ambiguity as to the desired remediation, I just said “make pathing better” cause that is what is expected of me as a consumer.

Okay, so I am noticing something here. And I don’t know how to put this in the best words, but I will say this respectfully, without anything personal directed towards you. Take it as a constructive feedback. So hear me out.

From what I can see through a few of your replies, it seems to me you like to talk in parallel, abstract, using analogy(though the analogy there weren’t the greatest to be honestly.) and “sophisticated words.”
You are saying a lot without saying much. Now I was guilty of this as well, way back in college. I used to write papers this way. I wrote a very long paper, lots of fancy words, lots of sentences, sounding very intelligent, but not saying a whole lot. My professor called me out on it. Yikes.
What she said to me was quite profound. She said, “if you want to want to demonstrate that you have fully understood a concept, you should be able to condense it, make it simple. You should be able to explain it to a 7 y/o child. If he doesn’t understand you, you haven’t really understand what you are saying.”
That piece of advice was shocking and powerful. It shook my world but has changed me for the better.

I took her advice from that day on and made it my way of talking and writing. I try my best to break down complex subjects into something simple.

There’s a clip out there somewhere, Neil DeGrasse Tyson is speaking with…Joe Rogan I think, doesn’t matter. When the space shuttle was being tested, specifically it’s landing, it’s coming in pretty fast compared to normal ariplanes. It was laterally gliding all over the run way. They decided to try adding grooves to the runway so the rubber tires would mold into the grooves so the wheels wouldn’t glide laterally. It worked so well, it’s now used on highways on tight off ramps.

In my estimation, the concept of adding grooves to pavement to increase traction and reduce sliding is something so simple, I could explain it to a kindergartener and they’d understand.

We’d been driving on highways for decades at that point. We had all the time in the world to come up with this super simple to understand solution.

And yet it took a national program, full of literal rocket scientists and billions of dollars, before this simple solution was even pondered.

There’s something very important you forgot here: bureaucracy. Our government is good at taking our money with taxes, but are very inefficient and quite frankly ineffective on many levels. I do not wish to get into politics, but yea hopefully you get the point.

1 Like
57

Not conflating, this is just an extension of the earlier conversation of it not being the customers job, which we seem to agree on. Customers can, and do, articulate the undesirable behaviors that constitute “bad pathing” in as much detail as they want (as is done in the Bug Report forum, for example), but then the ball is entirely in the devs’ court to find the root causes and provide solutions. Hell, the devs are in a far better position to be aware of the undesirable behaviors in the first place, but unfortunately they outsource almost all bugfinding to the players themselves. Whether any particular forumer at any given point in time has a comprehensive understanding of all pathing issues is less relevant than the fact that the devs, whose job it is to track these issues, have been given more than enough cumulative feedback and reports to have a pretty solid understanding of what players perceive to be the problems.

I’m glad the video helped make the pathing issues more clear to more people. It’s certainly very insightful and fairly comprehensive and goes above and beyond the level of feedback provided by the average player, and shines a brighter spotlight on the formation issue than I’ve ever seen elsewhere. That said, everything in it has been noted in the past, or at least pointed to, while not described so clearly. Regrouping being such a major focus now is understandable as it has reached new lows since the last patch, as noted by others. But the core issue has been pointed out before, as well as suggestions to include a free-form “formation” choice that prioritizes movement over formation.

Even if the superior explanation and exhaustive examples of the video were particularly revelatory, and therefore tempt you to make comparisons with innovation and new technologies, I can’t understate the significance of those good examples coming from past a version of AoE2, as I’ve said. Which is why the devs don’t have to reinvent the wheel or otherwise take some revolutionary new approach in order to address the pathing issues that players have noted. They just have to assimilate the past behaviors that were superior while maintaining the DE-era improvements.

2 Likes
58

I honestly don’t believe this directly contradicts what I was saying. Feedback can be very specific w/o being insightful. Also I already admitted there were specific areas in which DE pathfinding was buggy, addressed those, and set them aside to get down to the more fundamental issues. If you feel this response addresses what I was saying then I have to politefully disagree. Perhaps it is beyond my capacity to explain myself.

I again see what you’re saying.

image
image496×664 64.9 KB

I’m a high functioning autistic person and I’ve always been very self conscious about how I explain myself. For whatever reason there seems to often be this communication barrier between myself and most people. Much less so with those who’ve known me for a while, but still. Therefore I typically find myself over explaining, trying many different methods, use of analogy, rephrasing the point, using simple terms in summary, using precise words in other places, and often I find that instead the specific method that is best understood by the recipient being being acknowledged, there is issue taken with the least compatible or most objectionable method in the eyes of the recipient and somehow it ends up confusing the recipient and the message is less clear because of it. It’s like the message is only as clear as the least clear method in which it’s conveyed, but because no method seems reliable I try many, but invariably one method is confusing to the recipient, and the point of the message is lost. I don’t say this to attack you. This has been common in my life. Occam’s razor would suggest the issue lies with me, the world isn’t out to get me. But I’m not trying to be difficult to understand.

All I can assure you is I put great thought into what I wanted to say, how to say it, how it could best be understood, but I as the communicator have evidently failed for I clearly haven’t been understood.

As a random aside, I thought the horse-and-buggy to internal-combustion-car analogy was quite good, highlighting the difference between iterative improvement of something that is near it’s ceiling and a step-change over to something with far more potential that just needs insight to recognize it’s potential. However perhaps the parallels were too esoteric, or only meaningful to me.

2 Likes
59

Ahh yes that meme. I’ve seen it before; it’s hilarious.
The way you write and conduct yourself, I suspect you are an INTP in the Myer’s brigg type. Just my observation, could be totally off but I usually get it right. And the thing about the INTP type is that they value clear communication that is efficient and as concise as possible. I can imagine the frustration when you have a great idea/concept in your mind, but when you present it, people don’t understand what you are trying to say. I personally had to deal with this.
So what I did was try to hang out with people that don’t understand me, have a different way of communicating and talking, Eventually I get their own lingo and use it with them. But still, people will never understand one another completely, just enough to carry on with our lives.

Regarding the Occam’s razor, I have my doubts about that theory. For those who don’t know, Occam’s razor, in short, suggest that the simplest explanation is usually the correct one compared to something more complex.

So this is totally off topic from AOE 2, ### ####### this.
Imagine a group of remote tribal, indigenous humans that have been living on an unknown island that’s soo remote that they have NEVER come into contact with any other human beings besides their own kind on that island.
Now imagine one day, a lost cruise ship w/o people on board, suddenly floated to their island. These people must be in awe and wonder the origin of this ship.
Explanation one: The ship just suddenly materialized and is just is. It’s be there without an origin.
Second explanation: There’s a more advance civilization of beings that has mastered math and science to a degree who were able to construct such a marvel. These beings have a mind, intelligence, thus they were able to create something that shows intelligent design.

By Occam’s razor, we suppose to pick explanation 1 cause it’s more simple. But the truth is, we all know cruise ships don’t just make themselves and materialize out of thin air. It takes intelligent beings to create.

I’ve used this explain to make a case about creationism vs evolution. Anyway, just thought it’ fun to share. Now I’ll go back to AOE 2 topics 1111111

1 Like
61

110% spot on about the dev having the ball on their court. Yet they are not taking the appropriate action to remedy the pathing situation. Thus, a bunch of upset players are outraged and giving them flame.
It would great if the dev were just transparent about it.

1 Like
62

The current pathfinding is awful. Sometimes vills take the long route instead of moving a couple of tiles in front; even when there is nothing blocking their path.
Other thing that I noticed is that they sometimes circle around mule carts when they are ordered to build something else next to it. Like they get in a loop going back and forth and never really finish the task.
It’s very similar to the problem scouts used to have when you placed a shift queue flag in the middle of a forest and scouts didn’t know what to do and just stood there doing nothing, or moved around in a space of 2 tiles forever. It’s the same behavior except in this case it involves vills orbiting mule carts.

2 Likes