1 unit per hex: failed experiment

So let me ask this. How do a million different wargames do no stacking just fine with a competent AI? How is it that it's been done a million times before with no problem and suddenly now it's impossible to program an AI without a big stack? Seems like it's a giant crock to me.

If you have bad AI, then it's because it was programmed poorly, not because of a system that's been around forever.

They must not have any game designers who played tactical wargames. If I was in charge of creating an AI for this game. I would take some time to play hex style wargames, and put what I learned into the game. Tactical wargames were huge during the 80' but fell off the map so to speak. Also you could look into allied and panzer general. The AI on those games gets tough, real tough on hard levels. I am very surprised they did not take any of these things into account.

I do have one thing to say. In the game I am playing there is a city-state near me named Venice. They have their army in a ring around the city. 3 archers in the back, flanked by two warriors, and in front of the archers, are two units of spearmen. Now this is a good AI defensive scheme. So, evidently offense is the problem. This can be worked out. This is after all the Vanilla version. Civ 4 Vanilla also had a terrible AI, and the AI did get better with expansion packs and patches.
 
I think the most realistic solution is the relaxation of 1UPT rules for
a) either AI only so they can move the required units around
b) for both AI/human for certain units (ie: ranged can be stacked with 1 escort unit).

Not reading much else, my eyes caught this and I agree with this solution, moreso with a than b.
 
The hexagonal 1 unit per square is exactly what civ needed.

It makes war that much more complicated and requires serious strategy to overcome obstacles.

For example, try the true earth mod and try and invade south america. The mountains and jungles present serious logistical problems for invading armies. You actually have to think about how you can get siege units in the best possible locations and protect them from the defending army.

A monkey could put a stack of doom together and trounce through a civ with enough units. That is not the case in Civ 5.
 
They must not have any game designers who played tactical wargames. If I was in charge of creating an AI for this game. I would take some time to play hex style wargames, and put what I learned into the game. Tactical wargames were huge during the 80' but fell off the map so to speak. Also you could look into allied and panzer general. The AI on those games gets tough, real tough on hard levels. I am very surprised they did not take any of these things into account.

I do have one thing to say. In the game I am playing there is a city-state near me named Venice. They have their army in a ring around the city. 3 archers in the back, flanked by two warriors, and in front of the archers, are two units of spearmen. Now this is a good AI defensive scheme. So, evidently offense is the problem. This can be worked out. This is after all the Vanilla version. Civ 4 Vanilla also had a terrible AI, and the AI did get better with expansion packs and patches.

Having spent years playing the Battlefront game, as well as the other John Tiller games, I can assure you that the AI was not good in those game. They only worked when PBEM.
 
Tactical combat refers to low-level engagements between squads, platoons or companies. What Civ needs is a proper strategic warfare engine, something that can handle army-level combat (in multiple eras).
 
1 upt isn't bad on its own; it's bad in combination with the maps of CIV5, where there are simply not enough hexes to go around for a good tactical& strategical warfare. Panzer General (or steel panthers) works, but that's only because it is on a different scale and also: you jumpstart with a mixed bag of units, combined you can call this your "armygroup".

In Civ5, these parameters are not present:
1) At the start till many,many years later you have some units, which hardly can be considered a "armygroup", let alone a Army. Just a pity group of warriors to begin with, many years later upped to 5 units or so. Gamebalance ? NONE.

2) When thing DO get interesting, and you can muster something you can call "armygroup", the map is one big carpet with units. No room to manouvre, outflank, whatever. You have to BOMB you way into that land, HEX by HEX.

Just played my second game, this time on "Empire-lvl , huge and marathon with islands map. It's 200AC now, Rome was on my island but we could not make direct contact at first, since one City-state sets us apart. He and i futher where surrounded by another 4 city-states. So, battleplan was as easy and boring. Stratical one big joke, i took his three cities with a total of three units, two archers and a hoplite unit; most likely most work have been done for me by those three city-states surrounding him, which afcource, with nothing else to do; i made allied.
Montezama island i find later, five cities total with NO city-states for him. Took my three units, three trimeres and three or four new units to "ínvade" his island. After i settled my troops on some free spots he left on some bad ground, it's just one piece of cake to take him out.

I bet by the time i find the others, i am techwise way back and then the "fun" really start; right ?

Ow yes, and there is a good reason why hex-based wargames aren't so populair anymore. There are other, new game types that show Hex-based gaming is becoming obsolete.
RTS is more poplulair, but i like the Stratigal/tactical gameplay of the TW series the most; it combines best of both two types. Hexed based/turnbases (more or less) for the great overall map, realtime strategy for the battles.

Because it's so stupid to see that enemy trimere, getting hammered by my three trimeres, before he even get a chance to shoot back (if he servives the first volley).
For example.

Spears, getting hammered by enemy ranged archers and finished of by the enemy cav or warrior. Really, you are a great commander ! (not)

In TW, that would never happen, i would withdraw that spear, counterattack with cavalry and counter arty fire, whatever.

See my point ?
 
Exactly, thinking of others hex strategy games look at one of the best, The Operational Art of War...

toaw3_agshun.jpg



It has a tactical stack due to the scale and the size of armies.

The one using the 1UPT was Panzer General a game from 199x that used a battallion size army and scale for his scenarios.... You understand what means battallion scale?

A world scale as Civilization can't use tactical 1UPT, only ignorant people can approve it, and only because they lack experience in hex games.....

Hex is OK, but need the rules for the appropriate scale: more movement points and stack of units per tile in a tactical way. Maybe not complex as TOAW, but an halfway solution, with armies under Great Generals like...
 
please stop posting crap. Seriously, your idea is ********. You are comparing a tactical game on battalion level to a grand strategy one as Civ 5. People who use "realism" as an argument are simply fail.

Moderator Action: Attacking other members in this manner is not allowed, refute points or don't post at all, thanks.
Please read the forum rules: http://forums.civfanatics.com/showthread.php?t=422889
 
To start with, I want to avoid the logical fallacy called "exclusion of the middle" which is very common here. Even if you disliked the combat in earlier Civ game there would have been other solutions than an arbitrary 1 unit per hex limit. And for this game and scale I'd contend it is an extremely poor match, and the problems with it make it hard to enjoy the other innovations in the game. To wit:

1) It makes the mechanics of moving clunky and the game run very slow. The CPU is spending all of it's time doing complex pathing calculations to shuffle units around, and the more units (higher difficulty) they have, the longer it takes. Tasks which were never a problem in earlier games (assigning workers to build a road) are tedious and glitchy. Moving large armies is a buggy pain, and units in battle are frequently sent to their death (both human and AI) because it can't figure out how to get from A to B.

2) Roads are not only rare by design (fine) but almost useless in practice because of the stacking limit. Single NPC units can perma-block roads in neutral territory, and frequently do.

3) It distorts the rest of the game. Civ was designed around a different paradigm, and the changes needed to avoid unit overpopulation made the peaceful game imbalanced and boring (e.g. weak production and high costs for large empires, both driven largely by the need to avoid massive unit production.)

4) It is inappropriate for the scale of the game. If you wanted to have fights resolved on a tactical map with no stacking: great idea! But when the British Isles are 4 hexes, for example, it does violence to the feel of the game. And it scales poorly with size: the feel is best when you have a lot of room to maneuver, but the game design harshly penalizes large empires and maps, favoring smaller ones where the stacking limit performs the worst.

4) It's prone to artificial tactics. Once these are widely known the claim that combat is now more "strategic" will be falsified - because it's false. There is a reason why wargames abandoned the "I move and attack, then you move and attack" mode. It's because it rewarded unrealistic tactics, like soldiers darting from building to building and never getting attacked when they cross the street. Modern wargames have things like opportunity fire (e.g. when you move in range of my city, or artillery, then I attack you *first* as you charge at me.) Civ 5 has taken the worst aspects of the alternating turn approach and amplified them - for example, with cavalry which not only attacks first but which can retreat, or with insta-heal combat promotions. To eliminate the extreme distortion of "all my units attack, then you go" it's important instead to give both sides a chance - in other words, if you can damage someone else then you can be damaged yourself when the other guy gets a move. It's basic wargame design, and it was ignored. Civ 5 fails as a compelling wargame because it didn't pay any attention to decades of lessons from the tabletop world (I'd bet the Civ 5 team is utterly unaware of the principles behind the boardgaming renaissance led by German designers like Reiner Knizia, for example.)

5) The new problems created with 1 unit are worse than the big stack problem they solved. No stacking favors big units over little ones, replacing "stack of doom" with "unit of doom". Large armies create gridlock, and the absurd consequence that you can't even use most of your units because you can't even reach the field (in a battle on the size of a continent). This is especially a problem for the AI, which gets clogged and paralyzed with gigantic numbers of units at the highest levels.

6) The above problems cripple the AI as an opponent. It isn't that they didn't try, it's that the problem they're giving the AI isn't soluble. A bigger army can actually be worse than a smaller one because it can't move: that's very hard to program.

Solutions? A modest stacking limit, either with overstacking allowed (but only the "limit" worth of forces permitted to engage in military action) or the AI coded to keep itself 1 or more below the limit at all times (to allow movement.) Unlimited stacking for civilians. Combining units to create armies (and attaching generals) would be a cool idea that would work well. Ranged units can still attack more safely (but don't need to be able to do so from 2-3 hexes), and weak units can be guarded by strong ones in the same hex (a godsend for the AI.) There are plenty of answers to the Civ 4 problem, and unfortunately the Civ 5 model isn't the right one.


I think you have some valid points there and are well thought out. Although I may not agree with everything you stated, I applaud you for your well thought out feedback.
 
Tactical combat refers to low-level engagements between squads, platoons or companies. What Civ needs is a proper strategic warfare engine, something that can handle army-level combat (in multiple eras).

Civ V doesn't make that distinction. The question I'm asking is, can we mod the way AI makes decisions about how to move individual units, what kinds of formations to use them in and how to keep them that way, and so on?
 
Because it's so stupid to see that enemy trimere, getting hammered by my three trimeres, before he even get a chance to shoot back (if he servives the first volley).
For example.



See my point ?

Triremes physically tried to ram each other, or board each other. Of course there were catapults and archers. But ramming was the fundamental strategy for triremes. And what happened to the Galley? They should have added these elements into the game.

The other point I wanted to make about your reference to RTS and hex based games, is that this is a hex game, and they have to install some kind of decent hex based strategy that the AI can use and understand. The best way to get this is from multiplayer data. Take human experience and learning from actual game play. Use the data collected for the AI, to make it smarter. Program it to react like a real human player, and to use some form of tactical common sense. If they can teach an AI to play Chess and master it, why not Civ 5? I don't understand why so many PC games have stupid AI's.
 
Because chess has limited pieces and limited moves. Civ has a randomly generated map, more tiles, and a ton more pieces. It also has players who aren't willing to wait a minute between decisions.
And despite all those limitations, the calculations required to do just chess are huge and processing sinks.

I'm no AI expert, but I think the usual approach for "smarter" ai is to take some rules of thumb used by human players and make the ai use those rather than actually calculating what is the better play. So it looks smarter than just mindlessly tossing units forwards, but you can still learn and abuse its rules of thumb.
 
Love me some TOAW, THAT is stack and hex yummy goodness. And oh yeah, you can stack...
 
I still don't get why people are so resistant to the idea of having Total War style RTT battles for Civilization games as an idea. It seems to be perhaps the only solution to this combat problem.
 
In Total War, the battles ARE the primary draw of the game. In Civ, it's just not. War is of primary importance, but micromanaging combat is not. Civ is already a time vampire, imagining zooming in on EVERY battle like in Total War...

Plus, that's ALOT of combat modeling, and it frankly only works for melee era units. Once you hit the US Civil War you start to run into problems modeling combat...
 
It's too late for scale to be properly adjusted in Civ 5; because computers wouldn't be able to handle the graphics engine with scale increased to a proper size. So scale is stuck at pretty much what it is.

There is no easy re-working without Major changes being made, and with 2K being 2K, we all can rest assured that 2K won't be doing anything that spends much money to actually fix problems; instead they will do their best to 'cover them up'.

This is what happens when they 'don't fix problems of the previous civ' and instead delete working things and implement new broken ideas.
 
For a strategic-level game like Civ, the more I think about it, the more I feel that indeed the OP is right and that 1UPT is a big fail. As others have noted, it just doesn't fit the scale of the game map. 1UPT makes sense only when you have lots of tiles with lots of room to maneuver. But when the map is essentially zoomed out so much that you have AIs and humans with nearly every single tile containing 1 military unit and forcing you to play that "moving tile puzzle" game over and over agsin, 1UPT is just pointless and tedious!

What Civ5 should have done is instead of going to 1UPT is to develop a much better high-level strategic-level of combat more appropriate to the scale of Civ. For example, many people have suggested SMAC style combat or CTP style combat as good starting points!
 
The problem is cost. The more styles you stick into 1 game, then either the less time you can devote to either one or the longer and more costly it becomes to make the game. It's a fair guess that a lot of civ players have no desire to rts and a lot of rts players have no desire to civ, so it's unlikely to be cost effective. Total War does successfully fill the hybrid role, but their economy model is more simplistic than Civ's and their combat ai is worse than say SC's, so IMO it survives more on being the sole hybrid provider and 3d graphics rather than strength of gameplay. Civ's prior appeal lied in an engaging economy model and combat that didn't get much in the way of the better part of the game.

Plus, really, while a real time combat map would solve a lot of unit congestion issues at the strategic layer, you know that even Total War's combat ai was not very good. You could still roll armies far larger than your own with limited losses. If you're going to bother to add an entire extra layer to an existing franchise, it needs to be good.
 
Back
Top Bottom