Master of Magic unofficial patch (latest version)

[merlinp]

While I expand aggressively in 1.40n, I was spamming settlers like never before in 1.50, which is not balanced, in my view. It also favors cheap races, who can make settlers very quickly. 

I think it depends. I played a game with 1.50 starting on an island that fits 4 cities with a lot of empty islands and continents nearby. First, I went all settlers which ended up in a disaster as I was unable to defend neither the expansions or the starting island as everyone declared war on me at some turn. Then I loaded early save and didn't expand beyond the starting island until strong units, which worked out.

Also, you said that AI declares war on turn 1 with 1.50 unlike 1.31, I think this is untrue as Seravy explained somewhere else the behavior is identical with 1.40n and 1.50 where the AI waits until some turns until it can declare war and this was my experience with all games played. Although I think this is AI flaw, the AI should attack immediately if beneficial, like in MOO.

The AI naval play is by the way great in 1.50, can't access the seas unless having local naval supremacy.

As a general feedback on these patches, I played first several games on impossible/extreme with 1.40n and got beat badly until finally managed to win one at extreme with 11 death books and early wraith exploit. I noticed quite a lot of bugs still in 1.40n which all were fixed in 1.50 when I found that patch exists. With 1.50, all of the game's features seem to work and the game doesn't crash which is a huge imprevement over 1.31. 1.40n didn't crash for me either. Only one bug I encountered in 1.50 was where over 9 units somehow ended up in a tower which bugged the display when any of the units there were selected but that was fixable by disbanding enough units at the tower. Another maybe a bug I encountered was that quite often especially hero units would seem to miss a turn for no apparent reason, being unable to attack or move.
With 1.50, for me, at normal, the game is pretty easy if one survives the early game, at hard a challange but still beatable by stinging enemy with strong heroes/fantastic creatures/armies after surviving early game.

One thing I would change with 1.50 is the random movement of enemy units when they are unable to attack in tactical combat, mostly what this feature contributes is just waste of time and annoyance chasing the enemy units around in combat.

I tried the 1.51 RC5 but for me it was bugged, some of the features from caster of magic ended up in the game like different looking starting cities and no food required for some starting units.

The game is a lot of fun but incredible time sink, I think mods should try to speed it up.

[Seravy]

As it was explicitly mentioned I decided to take a look instead of trying to rely on my blurry memory.

First turn for war declaration :
1.31 : 100
1.40n : 100
1.51 : 100

First turn for hostility (being able to attack without war declaration if player fails a roll based on current relation) :
1.31 : 100
1.40n : 100
1.51 : 100

So...as far as I see in the disassembled code, there were no changes regarding this.

What did change is, the bug of burning down an outpost having no diplomatic effect was fixed so you can't go around destroying them with magic spirits without getting into wars.

Also, if the wizard is Peaceful or Lawful (personality), they skip the hostility roll entirely, and can't attack you until you attack them or they declare war. This is also new. However, neither Aura of Majesty nor Charismatic do this, the former simply raises relation by such an obscene amount you will almost never lose your treaties - and treaties do prevent the AI from attacking without declaring war, that's their primary function.

As the game is not modded, Subversion is still a spell in it - if an AI casts this on the player a few times, they'll probably find themselves at war with all 4 wizards almost immediately. Dropping relations by 25 each time is no joke. However, being a curse, it'll only be cast if the player does not have a treaty with the wizard casting it, which might be another reason why treaty = all peace, no treaty = all war.


War declaration was significantly changed because the original had so much bugs it effectively was no better than "declare war at random" and it was based on very silly things - such as the relative strength of armies on the wizard's home continent instead of the entire world. It's quite harsh to be honest but it had to be - neither the lack of game balance, nor the AI itself supports leniency : if the player is not kept under pressure, they can and will win the game easily using any of the things the AI cannot beat. (which does not mean it's harsher than before, quite the opposite - some of the bugs in the original caused the AI to declare war at the wrong (random) player so pretty much anyone who wanted to go to war with anyone else through those buggy parts of the code ended up declaring war to everyone eventually, including the human player)

Different looking starting cities is the correct behavior. That's the original game's sprites for cities without a City Wall built, but they were never displayed because the check for the wall was incorrectly implemented. So for 20 years everyone only ever saw how cities were supposed to look like after building the wall.

I don't think there are any "no food required" starting units in either CoM or 1.51 - the only such units in CoM are ships and catapults and you can't start with those - but if you find and report any, I can try to fix it.

[Aureustgo]

As it was explicitly mentioned I decided to take a look instead of trying to rely on my blurry memory.

First turn for war declaration : 
1.31 : 100
1.40n : 100
1.51 : 100

First turn for hostility (being able to attack without war declaration if player fails a roll based on current relation) :
1.31 : 100
1.40n : 100
1.51 : 100

So...as far as I see in the disassembled code, there were no changes regarding this.

I'm confident this is true but, operationally, this is not how the game actually plays. Wizards can attack your cities as soon as they meet you. Full stop. They don't "declare war" they just attack you non stop (so I end up attacking them anyways b/c it has the same net effect.) If you like I can give you 1.40n and 1.51 save files to prove it. The reason is, I think, Kyrub added some extra line of code that allowed the AI to know where our fortresses were, which was absent in 1.31. Or perhaps he just enabled the AI to attack it earlier, not sure. He made a comment about it in one of these threads.

What did change is, the bug of burning down an outpost having no diplomatic effect was fixed so you can't go around destroying them with magic spirits without getting into wars.

With due respect, unless there is something in the code I'm not aware of, this is not a bug but a feature of Master of Magic. Thus changing it is only a preference for how diplomatic interaction should work - a change that, as I've laid out, greatly impacts (hurts) game play, both in length, strategic planning and entertainment value. I understand how one might feel that way and indeed, at first blush it may seem odd, but there's nothing in the manual that suggests taking over outposts should result in war. In fact, the manual says just the opposite, highlighting attacks on cities and troops as grounds for negative diplomacy effects. For instance: "If you attack an opposing wizard’s city, that is viewed as an automatic declaration of war." Outposts are consistently not mentioned as affecting AI in any way. 

Also, if the wizard is Peaceful or Lawful (personality), they skip the hostility roll entirely, and can't attack you until you attack them or they declare war. This is also new. However, neither Aura of Majesty nor Charismatic do this, the former simply raises relation by such an obscene amount you will almost never lose your treaties - and treaties do prevent the AI from attacking without declaring war, that's their primary function.

My concern is that this fundamentally changes the game to make it exceptionally easy in certain cases, betraying the setting of "impossible," while exceptionally difficult in other cases (c.f., wizard attacks turn 1). This removes much of the nuance (and skill) in game play.

War declaration was significantly changed because the original had so much bugs it effectively was no better than "declare war at random" and it was based on very silly things - such as the relative strength of armies on the wizard's home continent instead of the entire world. It's quite harsh to be honest but it had to be - neither the lack of game balance, nor the AI itself supports leniency : if the player is not kept under pressure, they can and will win the game easily using any of the things the AI cannot beat. (which does not mean it's harsher than before, quite the opposite - some of the bugs in the original caused the AI to declare war at the wrong (random) player so pretty much anyone who wanted to go to war with anyone else through those buggy parts of the code ended up declaring war to everyone eventually, including the human player)

I think we agree but, operationally, in 1.31, the AI typically did not attack before the first 100 turns/1408 (unless one stupidly took out a neutral city the AI had its eye on), and on impossible, almost always attacked right after the 100 turns (unless you had tremendous magic realm similarity, etc.). Now it is easy to end up never being attacked which, again, betrays the "impossible" setting, or being attacked immediately. My feeling is the cure (as well intentioned as it was) is worse than the original "problem."

Please don't misunderstand - my feeling that the work done on 1.40n and 1.51 is impressive, detailed, and clever. The work put into these changes took lots of hours and frustration, I'm sure. But some of the changes have changed the game so dramatically from 1.31, i.e., contrary to what was intended. In particular, in 1.51, despite many great aspects, as a whole is borderline unplayable - unplayability shouldn't be a feeling experienced from any "patch," official or otherwise. 

As a consequence, and the motive of my postings, is that it might be wise to issue a patch that maintains the game play of the original game on central functions, while keeping specific mods to caster of magic or the mod section. Perhaps that would be a fruitful line of discussion - i.e., what elements are central/foundation, and which are mere nuance.

[Nelphine]

I believe one of the big problems is that diplomacy is all or nothing. With possible exception for things like burning outposts, you either get the full 1.5 diplomacy changes, or you get the 1.41 diplomacy, including any bugs it still has. From discussions with Seravy, there are too many bugs that affect too many things so the fixes interact too much.

[Seravy]

. The reason is, I think, Kyrub added some extra line of code that allowed the AI to know where our fortresses were, which was absent in 1.31. Or perhaps he just enabled the AI to attack it earlier, not sure.

Neither. He reduced the amount of forces the AI considers "beatable" to one fifth of the original. That certainly made the AI more aggressive as they thought "hey I can beat that sky drake with this swordsmen!". I suspect this wasn't intended and he wanted to do a 125%->100% reduction which ended up 125->25 instead but idk.
He also made it so that when the AI has the option to walk "through" another player's stack instead of not doing that, they prefer to do it and attack more than before - but these weren't intentional attacks that went through AI decisions.
But ultimately, the AI needs to have that "hostile" flag on to even look at your things and consider them targets.
btw neither the original nor insecticide nor any of my works have AI scouting for attack targets as a feature and it's not even viable to have that due to how simple AIs like this work - the AI just knows where everything is for all purposes except picking targets for city curses.

If the AI is indeed attacking you without the hostility flag enabled, that means a bug. Do note however if you attack as much as one of their spearmen, even if not killing it and fleeing, they are automatically hostile because they treat you as the aggressor.

Outposts are consistently not mentioned as affecting AI in any way.

This is a quite subjective matter as one can also say outposts are population 0 cities, thus everything written about cities applies to them as well. However the fact that the code was pretty much "if pop<>0 then do the diplomatic penalty" does seem to imply you are right and it was intentional. I wonder. The problem is, it's a horrible thing for gameplay. You can literally destroy all outposts so enemy wizards never have a second city without any repercussions - the AI is incapable of escorting settlers. Not that this looks like an attractive strategy in the vanilla game where you can just summon a Wraith on turn 1 and win with that but that doesn't make it right.
(which btw also raises the question if you get an elevated penalty for killing a Settler unit in battle - and you definitely do, same for engineers  - then how come an outpost which is effectively a used settler unit, has no penalty at all? If both are intentional, that means the design is contradicting itself.)

Difficulty has a massive negative modifier on diplomacy for the player, so I find it unlikely that you never get into wars - maybe with some of the more peaceful wizards, but to have everyone like that takes gigantic luck.

I seriously doubt the original intention of the developers was to make every wizard act like a maniacal paranoid militarist who declares war immediately but that's how the original played like due to the bugs.

If there is a significant demand I can allow the peaceful and lawful wizards to get hostile again but so far you are the first to ask for that. Most people just complained like "how is that peaceful he just sneak attacked me at my capital omg they are all crazy!".

[Aureustgo]

Neither.... If the AI is indeed attacking you without the hostility flag enabled, that means a bug. Do note however if you attack as much as one of their spearmen, even if not killing it and fleeing, they are automatically hostile because they treat you as the aggressor.

Kyrub writes in his pre- 1.40h patch: 

"AI should be actually able to target your fortress (bug removed)." 

I don't know if this is the source of the change in game play but something has changed. It don't think it's limited to your fortress, either. I take your point, but even in cases where you have done nothing to incite aggression, in 1.40n and 1.51, without declaration of war, the AI will, sooner rather than later, attack your fortress if there is a large spell book mismatch (death vs. life, for instance). Maybe personality plays a role as well. 

Laughably, I once had to fend off Tauron sending streams of hell hounds at my fortress, and Sharee sending just about everything else, at about turn 40 (i.e. immediately). With multiple AI capable of attacking you so early makes the game exceptionally difficult. I certainly didn't last. 

With certain setups - 11 books, conjurer + hell hounds, etc. - maybe this can be resisted, maybe, but anything else and you're done. At least give me 100 turns.

I'll look into how to change this.

does seem to imply you are right and it was intentional. I wonder. The problem is, it's a horrible thing for gameplay. You can literally destroy all outposts so enemy wizards never have a second city without any repercussions - the AI is incapable of escorting settlers... 

It's a trade off. 


The original rules and design allow raiding of outposts, yes, but the game play strikes me as interesting - the player must balance it against development of cities/armies, scouting other regions, etc. Also, at some point, your scouts have better things to do. It also impacts diplomacy- raiding prevents the signing of a wizard pact, which would otherwise be very beneficial to the player (especially in 1.40n and 1.51 because it guarantees never any attacks, even beyond 1408). 

The 1.51 rule, all else equal, simply encourages spamming of settlers (which some races do much better than others), lest the AI take up all the land, especially in large or huge landmass games. Perhaps there is a better way constraining AI development (when you are in a weak position) but I can't think of one.

I don't believe this is a good trade off in terms of game play. But I think my argument is broader: this is not a bug so it is not suitable for a "patch."

(which btw also raises the question if you get an elevated penalty for killing a Settler unit in battle - and you definitely do, same for engineers  - then how come an outpost which is effectively a used settler unit, has no penalty at all? If both are intentional, that means the design is contradicting itself.)

I agree that there is some discontinuity, but it's not necessarily a contradiction: settlers are "innocent" travelers, yes, but when they form outposts, historically, the effect is pushing borders and challenging land/resource rights. Arguments and disputes naturally erupt. Skirmishes "on the frontier" have always been part of human history, and not typically been grounds for all out war. 

Difficulty has a massive negative modifier on diplomacy for the player, so I find it unlikely that you never get into wars - maybe with some of the more peaceful wizards, but to have everyone like that takes gigantic luck.

I would say it's the norm, rather than the exception. What makes it even easier is that, when one can wizard pact nearby wizards, they often end up fighting each other (AI to AI diplomacy seems almost random), while the player gets stronger.

I think the intention, on impossible, was to give the player 100 turns, and then have wizards declare war soon thereafter (diplomatic standing still mattered in some cases- peaceful, lawful wizards with good magic matching would delay war for another 10-20 turns iirc). Yes, it was all or nothing, but the cutoff point was fairly clear (those complaining about paranoid AI likely did not understand this, and perceived AI actions to be random). Under 1.40n and 1.51 it is now still all or nothing, but in a much more radical way: the player has peace forever with magic matching (until you start knocking off wizards), or with contrary magic, fortress raiding on "turn 1." Talk about paranoid!

I think this takes a lot of the skill element out of the game on Impossible.

I think the goal of any patch should be to honor the original design. Players do like the idea that their actions matter, but only to a degree on the hardest levels. I'm not sure what the right trade off is, or if there is a solution, but that's perhaps something to discuss.

[Nelphine]

If you're having attacks that early I highly suggest posting the save file (and probably as many save files from earlier as possible to show that no hostile actions were taken on your part) as Seravy has indicated that would be a bug.

For the outpost thing, I'm going to explain my experience. Its one of the things I very much like about CoM, but I will also say that when we implemented it in CoM part of the discussion was that the consensus was not only that it was better gameplay to add a diplomatic penalty for attacking outposts, but as you've stated, the broader and more important thing is that we concluded it was actually a bug in the original, not a feature. While there are no indications that it should work that way, there are also no indications that it should NOT work that way, and given the number of other bugs in the original (especially regarding diplomacy), I don't think its definitive either way.

However that was the CoM community, so I wouldn't suggest we enforce our own decision on the 1.51 decision. However, similarly I don't think one person should make the decision for that community. I know Seravy receives feedback from people on all the versions he works on, so I would definitely try to get more of the community involved to see what the consensus is regarding diplomacy on outposts.

[Seravy]

I take your point, but even in cases where you have done nothing to incite aggression, in 1.40n and 1.51, without declaration of war, the AI will, sooner rather than later, attack your fortress if there is a large spell book mismatch (death vs. life, for instance). Maybe personality plays a role as well.

I think we are talking about the same thing. Hostility is the mechanic you talk about. The AI rolls the random generator and based on relation (spell book difference affects that most) and the result of the roll, they either set the "hostile" flag or not. However in all versions this can't happen before turn 100.
If it happens earlier that either means the AI attacks without the flag set which it shouldn't, or you set the flag by attacking them first

By the way, just checked, while attacking an empty outpost had no diplomatic repercussion and had no war as a result, it STILL set the "hostility" flag. So if you burned down the outposts, while it didn't have any VISIBLE repercussions, it allowed the AI to ignore the 100 turn limit and attack you.

This further questions what the intention of designers were - even if there was no WAR, setting the "hostile" flag to allow the AI to attack is almost the same thing except it's not formal. If they honestly wanted outposts to have no consequences, this should have also been skipped.

So there are four possibilities now instead of two :

1. There was no bug, the developers wanted the players to think there is no repercussion for burning outposts when in fact there is a one, or they assumed the players will be aware of the hidden hostility mechanic and know how it works (the strategy guide might talk about it, it's fairly detailed in diplomacy albeit 90% of the information was actually wrong and didn't match the game at all...)
2. They intended for no repercussions in which case the hostility flag being set is the bug
3. They had conflicting ideas and the result was a mix of both, maybe even done by two different people
4. They intended for war to start but a last-minute change or a rouge coder added an exception against it on outposts and no one noticed (we are talking about software were a bug of displaying city walls where there is none went unnoticed for the entire development of the game from 0.9x to 1.31!!!)

I seriously doubt anyone can answer this, probably not even the original developers remember...


Also, the strategy guide was released way after the game (at least I think so, otherwise it couldn't have up to date 1.31 information) and aside from being full of incorrect information on at least diplomacy, I feel it's more like explaining how the game ended up, pretending those were the original intentions regardless of the truth, as they had to live with the bugs at that point - the game no longer received patches. You can't really print a book that says "yeah actually 75% of the game does not work the way we wanted to, but you know how this is, deadlines, bosses, profit, now that it's sold we have no reason to do anything about it, it's not earning money."

[Aureustgo]

Good points raised by all. Just to add a few points:

The AI would never attack you in 1.31 even if you burnt outposts down. In fact, I just played a couple games to verify this (burning a few white wizard outposts c. 1404 loaded with black magic and no hostility at all till 1408). I don't disagree with your points and reasoning - perhaps there is an interaction with some of the other diplomatic changes made in 1.40n and 1.51 which has caused the change. Or something else still. If it helps I can send you saved game files (but I think you could probably do these all yourself pretty quickly). Also, I have one from 1.40n where I don't think I attacked a single outpost, and I was still attacked - though maybe I was attacked out in the open first (not sure if that counts as hostility triggering). 

As to the internal logic of the design - I agree it is tough to tell. But nothing in the game is really consistent. For instance, the AI will use the presence of your nearby forces as grounds to end a wizard pact, but you can't really ask their forces to leave yourself; Neutral armies don't disappear after a battle, but rampaging monsters do; etc. 

For better or worse, when in doubt, I suspect the end product was the defacto intention... as haphazard as some of it is. Some of the inconsistency likely came out of playtesting - perhaps the thinking in some spots was: "doing x makes the game too hard, so let's let the players do y, even if it violates our principle z."

The point of my comments are, effectively, the "patches" are changing the game, perhaps much more than either author has appreciated. And this can be most clearly seen in diplomacy and difficulty... and in a manner not envisioned, hurting game play for the previously stated reasons.

[Nelphine]

Given that seravy has fixed literally hundreds of bugs (that are most definitely bugs, often game crashing or game freezing bugs), let all the ons fixed for 1.41, I unfortunately think you can't say something is not a bug simply because it ended up in 1.31.

However I am liking the walk down memory lane!