Heroes 2 - bug fixes and mod

You are right, sorry. I will see a bit.

***

Problem occur due to "- the restoration of movement points for heroes fleeing the battlefield and redeemed on the same day was removed". Last module H2_30_BugFixes_Misc probably is bug-free.

Module H2_70_AI_Imrovements updated (link also updated).
----------------------------------------------------------------
Additional (further) changes:
- improved algorithm for Anti-Magic spell weighting
- more reliable algorithm for assessing the usefulness of buying a hero when an enemy hero is located near this castle
- improved algorithm for assessing the need to escape from the battle

Module H2_80_GameplayImprovement updated (link also updated).
----------------------------------------------------------------
Additional (further) changes:
- improved code for feature "the restoration of movement points for heroes fleeing the battlefield and redeemed on the same day was removed (now as in Heroes 3)" - thanks to Unknown_Hero
- spell Haunt is disabled in Shrines on map
- cost of Set Guardian spells decreased to 12, prob. of such spells in GM decreased, cost of Town Gate spell is set to 12, cost of Town Portal spell is set to 30
- the starting number of creatures that join free of charge at map objects has been changed; Halflings 15-30 instead of 20-40, Archers 10-20 instead of 10-25, Goblins 15-30 instead of 15-40, Sprites 10-20 instead of 15-25, Dwarves 8-15 instead of 10-20
- the starting number of creatures that join for charge at map objects has been changed; Trolls (Troll bridge) 4-8 instead of 4-6, Power Liches (Dead City) 4-8 instead of 4-6
- the week growth of creatures that join for charge at map objects has been changed; Trolls (Troll bridge) 2-4 instead of 1-3, Power Liches (Dead City) 2-4 instead of 1-3, Nomads (Desert Tent) 2-4 instead of 1-3, Rogues (Wagon Camp) 4-8 instead of 3-6.
- artifact Statesman's Quill moved from level 1 to level 3

Hello @Ben80,

Well, I did a test with the latest "H2_80_GameplayImprovements.dll" plugin and it turns out that a Knight hero recruited after he fled from a sea battle can move two steps further than he should be able to move on land.
A Knight hero recruited with Peasants and Archers should have 1000 movement points on land whereas he has 1200 movement points on land.

Unknown_Hero wrote:

Hello @Ben80,

Well, I did a test with the latest "H2_80_GameplayImprovements.dll" plugin and it turns out that a Knight hero recruited after he fled from a sea battle can move two steps further than he should be able to move on land.
A Knight hero recruited with Peasants and Archers should have 1000 movement points on land whereas he has 1200 movement points on land.

Hello @Unknown_Hero,

Thanks, that's interesting. However, the situation can already be considered acceptable
(especially considering that at sea the Knight had a huge amount of MP, as I guess)

BTW, amount of MP of hero fled from a sea battle does not depend on his "new" army, it only depends on amount of MP which he had at the moment of sea battle. In original Heroes 3, there is no such fix, as I made recently (so, you can run original Heroes 3 and show us new video with huge amount of MP on land after sea battle )

A little about future plans or beyond classic Heroes 2 gameplay changes...


2 plugins will be developed, the purpose of which is to make the game even more interesting. 1 plugin is already ready:

H2_83_Anti_RedRush.
- The weekly increase in the players' castles occurs only after the corresponding player has been given a turn. That is, if a player on the 1st day of the week captures castle of some player who play later, then he will not get newborn creatures.

However, I would like to share my plans for writing another very interesting plugin. Its purpose will be to limit the number of hero-to-hero encounters
(meaning heroes of one player) for each hero to 1 per day, no more.
This will limit the length of the chains of heroes and the maximum distance
that troops can cover during one day (in fact, up to 2 distances that the heroes travel in a day).

We remember that in Heroes 4 this distance was in fact reduced to 1 distance traveled by the hero per day. The plugin I want to present will essentially be a trade-off between robust Heroes 4 system,  and the less robust ("instant" movement of troops from one end of the map to the other), but highly dynamic Heroes 2-3 system.
Still I hope the player will feel comfortable playing many of the classic Heroes 2 maps

P.S. I have the impression that I am sinking deeper and deeper into the philosophy of WOG This is a joke.

Hello @Ben80,

Ben80 wrote:

BTW, amount of MP of hero fled from a sea battle does not depend on his "new" army, it only depends on amount of MP which he had at the moment of sea battle.

I'm trying to figure out how this should work.
This means that a hero (so with 1500 movement points in a boat, as the speed of the slowest troop is not taken into account in a boat) should have 1500 movement points when recruited if he had at least 1500 movement points or more left (thanks to the sea movement bonuses (Navigation, Artifacts, Lighthouses)) after fleeing from a sea battle.

And a hero who would have less than 1500 movement points after fleeing from a sea battle would keep that exact number of movement points (e.g. 900 movement points) when recruited, this regardless of whether he had any bonuses at sea or not.

This is the intended way of working, right?

Ben80 wrote:

A little about future plans or beyond classic Heroes 2 gameplay changes...


3 plugins will be developed, the purpose of which is to make the game even more interesting.

Nice.

Ben80 wrote:

One plugin is already ready, it is called H2_83_Anti_RedRush.
Brief functional description : The weekly increase in the players' castles
occurs only after the corresponding player has been given a turn. That is,
if a player on the 1st day of the week captures castle of some player who play later, then he will not get newborn creatures.
Download: link

It's good to give the user the choice.
Because it is nevertheless a change that can bring frustration to the player.
Personally, I won't use it.

Ben80 wrote:

However, I would like to share my plans for writing another very interesting plugin. Its purpose will be to limit the number of hero-to-hero encounters
(meaning heroes of one player) for each hero to 1 per day, no more.
This will limit the length of the chains of heroes and the maximum distance
that troops can cover during one day (in fact, up to 2 distances that the heroes travel in a day).

This is a change more related to and intended for multiplayer gaming in my opinion.
As this way of playing, the "chain of heroes", is more in line with competitive multiplayer.
Is this the goal?
Because the single player can already apply this if he wants.
In my case, I find it more interesting to play several "main" heroes than one "big" hero.

Ben80 wrote:

We remember that in Heroes 4 this distance was in fact reduced to 1 distance traveled by the hero per day. The plugin I want to present will essentially be a trade-off between robust Heroes 4 system,  and the less robust ("instant" movement of troops from one end of the map to the other), but highly dynamic Heroes 2-3 system.
Still I hope the player will feel comfortable playing many of the classic Heroes 2 maps

I never played Heroes 4, so I can't compare.

Ben80 wrote:

P.S. I have the impression that I am sinking deeper and deeper into the philosophy of WOG This is a joke.

That the user has a choice for each option is not a bad thing.
The point is that the options offered should be interesting and not just options for the sake of telling there are many options.
I had tested WoG at the very beginning, but not afterwards.

Users being different they may want one change and not another.
And if these changes are in the same plugin, it may reduce the desire to use the mod for potential users.
It is sometimes tricky to "impose" changes too much.
For me it should be possible to propose a certain vision of the game, but always let the user choose and adjust the changes made.

Hello @Unknown_Hero

Unknown_Hero wrote:

I'm trying to figure out how this should work.
This means that a hero (so with 1500 movement points in a boat, as the speed of the slowest troop is not taken into account in a boat) should have 1500 movement points when recruited if he had at least 1500 movement points or more left (thanks to the sea movement bonuses (Navigation, Artifacts, Lighthouses)) after fleeing from a sea battle.

And a hero who would have less than 1500 movement points after fleeing from a sea battle would keep that exact number of movement points (e.g. 900 movement points) when recruited, this regardless of whether he had any bonuses at sea or not.

This is the intended way of working, right?

It is intended to work as follows:
after fleeing remaining MP are recalculated by formula:
remaining_MP (in tavern) = remaining_MP (in boat) * MP_max (land) / MP_max (in_boat)

We see that relation MP_max (in_tavern) / MP_max (in_boat) is used. The same formula takes place in Heroes 3 when hero are using Admiral's Hat and goes from boat to shore.
However, when we are calculating MP_max (land) we use current army stats of hero fleeing from boat.

Unknown_Hero wrote:

This is a change more related to and intended for multiplayer gaming in my opinion.
As this way of playing, the "chain of heroes", is more in line with competitive multiplayer.
Is this the goal?

In general, I don't distinguish between single player and multiplayer (maybe I'm wrong).
Red rash and chains destroy not only the gameplay of a multiplayer game,
but also many author's maps (including the maps of the original game).

Unknown_Hero wrote:

Because the single player can already apply this if he wants.

I do not agree. Self-restraints usually do not work well. Much better when it is part of the game mechanics. Yes, using the plugin is voluntary ... but it works.

Unknown_Hero wrote:

In my case, I find it more interesting to play several "main" heroes than one "big" hero.

You have got a point ! Proposed mechanics will stimulate playing by several "main" heroes than one "big" hero.

***

One more reason to use future "anti-chain" plugin. In the H2_80 plugin, I slightly reduced the starting armies of heroes. Nevertheless, to be honest, the real value of heroes still remains much higher than 2500 gold (in most game situations).
By imposing restrictions on the number of hero-to-hero encounters, that is, in fact, on the hero's capabilities, we seem to additionally lower his real value.

Hello @Ben80,

Ben80 wrote:

It is intended to work as follows:
after fleeing remaining MP are recalculated by formula:
remaining_MP (in tavern) = remaining_MP (in boat) * MP_max (land) / MP_max (in_boat)

We see that relation MP_max (in_tavern) / MP_max (in_boat) is used. The same formula takes place in Heroes 3 when hero are using Admiral's Hat and goes from boat to shore.
However, when we are calculating MP_max (land) we use current army stats of hero fleeing from boat.

So, for remaining_MP (in boat), are sea-related bonuses (Navigation, Artifacts, Lighthouses) taken into account?

First example, the hero has 5000 remaining movement points in boat, which is the total of 1500 points for movement at sea (the speed of the slowest creature is not taken into account in a boat) + 3500 points of bonuses related to the sea (Navigation, Artifacts, Lighthouses).
Is that:
remaining_MP (in boat) = 5000 points
Or:
remaining_MP (in boat) = 1500 points

Second example, the hero has 2500 remaining movement points in boat, which is the total of 1500 points for movement at sea (the speed of the slowest creature is not taken into account in a boat) + 1000 points remaining from 3500 points of bonuses related to the sea (Navigation, Artifacts, Lighthouses)(the sea-related movement point bonuses are removed first before the standard 1500 points of the hero in a boat, right?).
Is that:
remaining_MP (in boat) = 2500 points
Or:
remaining_MP (in boat) = 1500 points

Third example, the hero has 1000 remaining movement points in boat, which is the total of 1000 points remaining from 1500 points for movement at sea (the speed of the slowest creature is not taken into account in a boat) + 0 point remaining from 3500 points of bonuses related to the sea (Navigation, Artifacts, Lighthouses)(the sea-related movement point bonuses are removed first before the standard 1500 points of the hero in a boat, right?).
Then, it's right that:
remaining_MP (in boat) = 1000 points

Same for MP_max (in_boat), are sea-related bonuses (Navigation, Artifacts, Lighthouses) taken into account?
For example, the hero has 5000 movement points in boat per day, which is the total of 1500 points for movement at sea (the speed of the slowest creature is not taken into account in a boat) + 3500 points of bonuses related to the sea (Navigation, Artifacts, Lighthouses).
Is that:
MP_max (in_boat) = 5000 points
Or:
MP_max (in_boat) = 1500 points

Ben80 wrote:

However, when we are calculating MP_max (land) we use current army stats of hero fleeing from boat.

So, for MP_max (land), with hero fleeing from a sea battle with only a troop of Peasants left in the army.
It means the speed of the slowest creature is taken into account, corresponding here to 1000 points with Peasants (speed 2), or it means 1500 points because the speed of the slowest creature is not taken into account in a boat?

Also, there is a bug with the Admiral's Hat in standard Heroes 3.

Ben80 wrote:

In general, I don't distinguish between single player and multiplayer (maybe I'm wrong).

Single player is for the fun of playing the game.
Multiplayer is about complicating your life with rules that take away all the fun of single player, sometimes with still a little fun, but often with a lot of rage, hatred and frustration.      

Ben80 wrote:

Red rash and chains destroy not only the gameplay of a multiplayer game,
but also many author's maps (including the maps of the original game).

Let the single player have the fun of playing the game without constraints, the AI will never blame him.
Multiplayer is about complicating your life with rules that take away all the fun of single player.      

Ben80 wrote:

Unknown_Hero wrote:

Because the single player can already apply this if he wants.

I do not agree. Self-restraints usually do not work well. Much better when it is part of the game mechanics. Yes, using the plugin is voluntary ... but it works.

Let the single player have the fun of playing the game without constraints, the AI will never blame him.
The game should be fun to play and not bring frustration.

Ben80 wrote:

Unknown_Hero wrote:

In my case, I find it more interesting to play several "main" heroes than one "big" hero.

You have got a point ! Proposed mechanics will stimulate playing by several "main" heroes than one "big" hero.

You see, I play the game in a way that I enjoy, without any constraints being imposed on me.

Ben80 wrote:

One more reason to use future "anti-chain" plugin. In the H2_80 plugin, I slightly reduced the starting armies of heroes. Nevertheless, to be honest, the real value of heroes still remains much higher than 2500 gold (in most game situations).
By imposing restrictions on the number of hero-to-hero encounters, that is, in fact, on the hero's capabilities, we seem to additionally lower his real value.

In my case, with my multi "main" hero style of play, I sometimes have more than one hero encounter per turn and that in very distant locations of the map.
I'm afraid that all these restrictions also "break" the different possibilities of the game, even for the player who doesn't use the "chain of heroes".

Hello @Unknown_Hero,

Quote :

So, for remaining_MP (in boat), are sea-related bonuses (Navigation, Artifacts, Lighthouses) taken into account?

Yes, of course.

Quote :

First example, the hero has 5000 remaining movement points in boat, which is the total of 1500 points for movement at sea (the speed of the slowest creature is not taken into account in a boat) + 3500 points of bonuses related to the sea (Navigation, Artifacts, Lighthouses).
Is that:
remaining_MP (in boat) = 5000 points
Or:
remaining_MP (in boat) = 1500 points

5000 points.

Quote :

Second example, the hero has 2500 remaining movement points in boat, which is the total of 1500 points for movement at sea (the speed of the slowest creature is not taken into account in a boat) + 1000 points remaining from 3500 points of bonuses related to the sea (Navigation, Artifacts, Lighthouses)(the sea-related movement point bonuses are removed first before the standard 1500 points of the hero in a boat, right?).
Is that:
remaining_MP (in boat) = 2500 points
Or:
remaining_MP (in boat) = 1500 points

2500 points.

Quote :

Third example, the hero has 1000 remaining movement points in boat, which is the total of 1000 points remaining from 1500 points for movement at sea (the speed of the slowest creature is not taken into account in a boat) + 0 point remaining from 3500 points of bonuses related to the sea (Navigation, Artifacts, Lighthouses)(the sea-related movement point bonuses are removed first before the standard 1500 points of the hero in a boat, right?).
Then, it's right that:
remaining_MP (in boat) = 1000 points

Yes, 1000 points.

Quote :

Same for MP_max (in_boat), are sea-related bonuses (Navigation, Artifacts, Lighthouses) taken into account?

Yes.

Quote :

For example, the hero has 5000 movement points in boat per day, which is the total of 1500 points for movement at sea (the speed of the slowest creature is not taken into account in a boat) + 3500 points of bonuses related to the sea (Navigation, Artifacts, Lighthouses).
Is that:
MP_max (in_boat) = 5000 points
Or:
MP_max (in_boat) = 1500 points

5000 points.

Quote :

So, for MP_max (land), with hero fleeing from a sea battle with only a troop of Peasants left in the army.
It means the speed of the slowest creature is taken into account, corresponding here to 1000 points with Peasants (speed 2), or it means 1500 points because the speed of the slowest creature is not taken into account in a boat?

1000 points. MP_max (land) is calculated for land conditions, not for boat.

Quote :

Also, there is a bug with the Admiral's Hat in standard Heroes 3.

I have fixed this bug several years ago together with RoseKavalier.
The bug is not related to the fact that the formula is incorrect or absurd.

Quote :

Single player is for the fun of playing the game.
Multiplayer is about complicating your life with rules that take away all the fun of single player, sometimes with still a little fun, but often with a lot of rage, hatred and frustration.      

Fortunately, I played almost all the games with my good friend, and not with strangers. It was wonderful and unforgettable. Haven't played multiplayer for a long time.

Quote :

Let the single player have the fun of playing the game without constraints, the AI will never blame him.

I'm a little surprised. I do not force anyone to play what he does not like. If you mean that I should not write such plugins and make them freely available, then this may deprive additional features of people who would like these plugins.

Quote :

Let the single player have the fun of playing the game without constraints, the AI will never blame him.
The game should be fun to play and not bring frustration.

I've read a couple of books on game design. People enjoy solving problems that are neither too hard nor too easy. So I'm just trying to create this optimal game design, removing absurd things from the game (like red-rush).
Depending on the level of the player, he himself can choose which plugins to use. The same is the case in the original game - you can choose the difficulty levels - from easy to impossible.

Quote :

I'm afraid that all these restrictions also "break" the different possibilities of the game, even for the player who doesn't use the "chain of heroes".

No, I see no reason to think so (although of course the last plugins we talked about are experimental in nature).

***

The latest plugins are actually more intended for new players, or for a few experienced players who don't like the shortcomings of the current gameplay.

It looks like a pyramid - bug fixes suit almost all seasoned players, AI improvements - less, gameplay improvements - even less.

The latest improvements might appeal to new players who haven't played (or have played little). But there is a problem with the presence of new players.

Hello @Ben80,

Thanks for all your answers about hero movement.
I did some tests, here is what I observed.

Hero recruited after combat at sea only moves vertically and horizontally on "Grass" terrain (penalty: 100%).
Attacking hero only attacks vertically or horizontally adjacent to the opponent (considered as one move) at start of the turn (so with full movement points).
Knight hero has a combination of Navigation Expert, Astrolabe and troops, and the player may control one or two Lighthouses.

Hero being attacked at sea and fleeing:

Champions: 15 moves

Peasants: 15 moves

Navigation Expert, Phoenixes: 30 moves

Navigation Expert, Peasants: 30 moves

Navigation Expert, Astrolabe, Phoenixes: 40 moves

Navigation Expert, 2 Lighthouses, Astrolabe, Peasants: 40 moves

Navigation Expert, 2 Lighthouses, Peasants: 30 moves

Navigation Expert, 1 Lighthouse, Astrolabe, Phoenixes: 40 moves

Navigation Expert, 1 Lighthouse, Astrolabe, Phoenixes and Skeletons: 40 moves

Hero being attacked at sea and surrendering:

Navigation Expert, 1 Lighthouse, Astrolabe, Phoenixes and Skeletons: 40 moves

Hero attacking at sea and fleeing:

Champions: 14 moves

Peasants: 9 moves

Navigation Expert, Phoenixes: 14 moves

Navigation Expert, Peasants: 9 moves

Navigation Expert, Astrolabe, Phoenixes: 14 moves

Navigation Expert, 2 Lighthouses, Astrolabe, Peasants: 9 moves

Navigation Expert, 2 Lighthouses, Peasants: 9 moves

Navigation Expert, 1 Lighthouse, Astrolabe, Phoenixes: 14 moves

Navigation Expert, 1 Lighthouse, Astrolabe, Rangers: 11 moves

Navigation Expert, 1 Lighthouse, Astrolabe, Crusaders: 13 moves

Hero attacking at sea and surrendering:

Navigation Expert, 1 Lighthouse, Astrolabe, Phoenixes and Rangers: 11 moves

I find it a bit excessive that an attacked hero keeps so many movement points compared to an attacking hero.
Shouldn't it work the same way as an attacking hero in both cases?
What do you think?

Ben80 wrote:

Quote :

Also, there is a bug with the Admiral's Hat in standard Heroes 3.

I have fixed this bug several years ago together with RoseKavalier.
The bug is not related to the fact that the formula is incorrect or absurd.

Nice.

Ben80 wrote:

Quote :

Single player is for the fun of playing the game.
Multiplayer is about complicating your life with rules that take away all the fun of single player, sometimes with still a little fun, but often with a lot of rage, hatred and frustration.      

Fortunately, I played almost all the games with my good friend, and not with strangers. It was wonderful and unforgettable. Haven't played multiplayer for a long time.

Yes, it's always better to play with people you know.
I've never played multiplayer, but I've watched multiplayer games and read user experiences on various game forums and it's not always a wonderful experience.

Ben80 wrote:

Quote :

Let the single player have the fun of playing the game without constraints, the AI will never blame him.

I'm a little surprised. I do not force anyone to play what he does not like. If you mean that I should not write such plugins and make them freely available, then this may deprive additional features of people who would like these plugins.

No, this was not intended for you in particular, but rather to say that the single player can play the game the way he wants, the goal being that he has fun playing the game, even if he uses cheats code for example.
The AI is just there to entertain him, and the game should not be boring or constraining.

Ben80 wrote:

Quote :

Let the single player have the fun of playing the game without constraints, the AI will never blame him.
The game should be fun to play and not bring frustration.

I've read a couple of books on game design. People enjoy solving problems that are neither too hard nor too easy. So I'm just trying to create this optimal game design, removing absurd things from the game (like red-rush).
Depending on the level of the player, he himself can choose which plugins to use. The same is the case in the original game - you can choose the difficulty levels - from easy to impossible.

Yes, that's the point.
It is necessary to give the player as much flexibility as possible for possible adjustments in the game.
So that he can play the game the way he likes without anything being imposed on him.

Ben80 wrote:

Quote :

I'm afraid that all these restrictions also "break" the different possibilities of the game, even for the player who doesn't use the "chain of heroes".

No, I see no reason to think so (although of course the last plugins we talked about are experimental in nature).

***

The latest plugins are actually more intended for new players, or for a few experienced players who don't like the shortcomings of the current gameplay.

It looks like a pyramid - bug fixes suit almost all seasoned players, AI improvements - less, gameplay improvements - even less.

The latest improvements might appeal to new players who haven't played (or have played little). But there is a problem with the presence of new players.

As long as the player has the choice, I think it's okay to offer him interesting possibilities.

Hello @Unknown_Hero,

Quote :

I find it a bit excessive that an attacked hero keeps so many movement points compared to an attacking hero.
Shouldn't it work the same way as an attacking hero in both cases?
What do you think?

Game mechanics (concerning MP after fleeing) is intended to work absoultely in the same way both for attacking and attacked hero (I even did not thought about it). The difference can only be in which day they are recruited in tavern - if it is new day after fleeing, then MP are completely restored as usually (and like Heroes 3).
If you see any bugs, I would ask to send me save files via PM.

Unknown_Hero wrote:

The AI is just there to entertain him, and the game should not be boring or constraining.

I know there are few people out there who view AI as a virtual adversary, but I see AI as part of the game mechanics - to create interesting gameplay.  Therefore, if the AI ​​was too strong, and it spoiled the gameplay,  then I would probably write plugins that make the AI ​​more dumb.

"Boring" and "constraining" are very different things. Any good game ("not boring" game) is an optimal combination of possibilities and limitations.

The point is, however, that different people assess the interestingness of the gameplay differently. Therefore, different people play different games. In the case of my mod, it is possible to use different plugins of your choice.

Hello @Ben80,

Ben80 wrote:

Unknown_Hero wrote:

I find it a bit excessive that an attacked hero keeps so many movement points compared to an attacking hero.
Shouldn't it work the same way as an attacking hero in both cases?
What do you think?

Game mechanics (concerning MP after fleeing) is intended to work absoultely in the same way both for attacking and attacked hero (I even did not thought about it). The difference can only be in which day they are recruited in tavern - if it is new day after fleeing, then MP are completely restored as usually (and like Heroes 3).
If you see any bugs, I would ask to send me save files via PM.

There is certainly a bug then.
As the attacked hero (thus attacked during the opponent's turn) can only be recruited the next day, his movement should not be calculated in relation to the sea since he is already out of the sea at the end of the previous turn.
A specific save is not necessary.
A scenario played in multiplayer allows to check it easily.

Ben80 wrote:

Unknown_Hero wrote:

The AI is just there to entertain him, and the game should not be boring or constraining.

I know there are few people out there who view AI as a virtual adversary, but I see AI as part of the game mechanics - to create interesting gameplay.  Therefore, if the AI ​​was too strong, and it spoiled the gameplay,  then I would probably write plugins that make the AI ​​more dumb.

"Boring" and "constraining" are very different things. Any good game ("not boring" game) is an optimal combination of possibilities and limitations.

The point is, however, that different people assess the interestingness of the gameplay differently. Therefore, different people play different games. In the case of my mod, it is possible to use different plugins of your choice.

Yes, and it's a plus to be able to select the plugins in your mod, thanks for this.

*

So, I completed the Dominion scenario (PoL), impossible difficulty, Blue player as Knight, with your mod (plugins 30, 48, 49, 70, 71 and 80).
I have not encountered any AI heroes using the Teleport spell in this scenario.
(Remember, I added the plugin during the game so the increase to apparition percentage of the spell was not taken into account at the beginning of the scenario.)
Maybe I'll see that in another scenario.

I finished the scenario in 81 days, score 243, Black Dragon.

I found the AI players quite passive in this game.
I was able to execute my plans without difficulty and without having to adjust them during the game.

Usually, in the original game, this scenario is quite challenging and even more so when the difficulty is set to impossible.
From the beginning of the game the human player is constantly harassed by the green AI player, by the yellow AI player and often by the orange AI player.
A little later in the game, the red AI player and the purple AI player also come to disturb the blue player, and this can sometimes mean the end of the scenario if the blue player does not react or cannot react seriously.

Here everything went quietly, as if the AI players were playing a defensive game but without much result.

Unknown_Hero wrote:

There is certainly a bug then.
As the attacked hero (thus attacked during the opponent's turn) can only be recruited the next day, his movement should not be calculated in relation to the sea since he is already out of the sea at the end of the previous turn.

Thanks for bug reporting, I have fixed it, and updated Dll and link. Though problem is not calculation of relation to the sea (calculation is done at the very moment when hero disappear from map). Problem was in presence of flag "at sea" in hero structure when he was already in tavern.

Unknown_Hero wrote:

So, I completed the Dominion scenario (PoL), impossible difficulty, Blue player as Knight, with your mod (plugins 30, 48, 49, 70, 71 and 80).
I have not encountered any AI heroes using the Teleport spell in this scenario.
(Remember, I added the plugin during the game so the increase to apparition percentage of the spell was not taken into account at the beginning of the scenario.)
Maybe I'll see that in another scenario.

I finished the scenario in 81 days, score 243, Black Dragon.

I found the AI players quite passive in this game.
I was able to execute my plans without difficulty and without having to adjust them during the game.

Usually, in the original game, this scenario is quite challenging and even more so when the difficulty is set to impossible.
From the beginning of the game the human player is constantly harassed by the green AI player, by the yellow AI player and often by the orange AI player.
A little later in the game, the red AI player and the purple AI player also come to disturb the blue player, and this can sometimes mean the end of the scenario if the blue player does not react or cannot react seriously.

Here everything went quietly, as if the AI players were playing a defensive game but without much result.

I did not add some special defensive algorithm. From my playing experience  new AI is far superior to the old one, so I would only advice play more with the new AI and after that draw conclusions.
May be it is worth to playtest on some hard original maps from SW ?

***

Actually, yesterday I put some defensive AI things, but I think you did not use last H2_70 (in some sense, experimental version).

@Ben80.

Ben80 wrote:

Thanks for bug reporting, I have fixed it, and updated Dll and link. Though problem is not calculation of relation to the sea (calculation is done at the very moment when hero disappear from map). Problem was in presence of flag "at sea" in hero structure when he was already in tavern.

Nice, okay.

Ben80 wrote:

I did not add some special defensive algorithm. From my playing experience  new AI is far superior to the old one, so I would only advice play more with the new AI and after that draw conclusions.
May be it is worth to playtest on some hard original maps from SW ?

It was more an observation than a conclusion.
What scenarios have you played with that you have noticed an AI improvement?
(Scenarios where the player starts without a big army and without a castle already built.)

Ben80 wrote:

Actually, yesterday I put some defensive AI things, but I think you did not use last H2_70 (in some sense, experimental version).

No, I did not use yesterday's update of this plugin.

In fact, in every test I see significant improvements - AI behaviour on map and in battle.
However, your observations make me a little alarming, since I already realized that you are good at testing.

The fact is that I did not play all the maps, even from SW. Therefore, after improving the AI, I wanted to play those maps that I hadn’t played yet.  For example Shipwrecked !. It was not possible to pass it the first time. But there were some maps that I went through both before the AI ​​changes and after, for example, Seven Lakes.

@Ben80.

Okay, I'll give it a try with Seven Lakes, random castle, impossible difficulty, with all the recent plugins and see how it goes.

Unknown_Hero wrote:

@Ben80.

Okay, I'll give it a try with Seven Lakes, random castle, impossible difficulty, with all the recent plugins and see how it goes.

Thanks

The old H2_70 is more reliable. Besides, I have already received positive feedback on it from one good player .

@Ben80.

Ben80 wrote:

The old H2_70 is more reliable. Besides, I have already received positive feedback on it from one good player .

You mean the one in "H2_fixes_and_mod v6" (12.5 ko)?

Also your file is named "H2_70_AI_Imrovements.dll" with "Improvements" without "p".

Unknown_Hero wrote:

@Ben80.

You mean the one in "H2_fixes_and_mod v6" (12.5 ko)?

Yes.

Unknown_Hero wrote:

Also your file is named "H2_70_AI_Imrovements.dll" with "Improvements" without "p".

I was in a hurry to give the players the best )))

@Ben80.

Okay, thanks.

Hello @Ben80,

So, I completed the Seven Lakes scenario, impossible difficulty, Blue random player as Wizard, with your mod (plugins 30, 42, 48, 49, 70 (previous version), 71 and 80).
I finished the scenario in 92 days, score 212, Giant.
It was more difficult with this scenario.

The AI uses more different spells that are relevant to the combat context (though not always optimally, as you already know, such as casting the "Mirror Image" spell on a troop of one Lich).
With the increased cost of the "Dimension Door" and "Town Portal" spells, I found it more tedious to complete the scenario once you know the scenario is won and all that's left is to eliminate the AI player(s) still in play.

I also had the bug of the AI hero casting the "Dispel Magic" spell on the Titans image from the "Mirror Image" spell:
https://youtu.be/aLdknN3qXjA

**Edit.**

Here is the save file.

Hello @Unknown_Hero,

Thank you very much ! Bug report also very important.

Hello @Ben80,

I tried once again and completed the Broken Alliance scenario, hard difficulty, Blue random player as Necromancer, with your mod (plugins 30, 42, 48, 49, 70 (previous version), 71 and 80).
I finished the scenario in 72 days, score 160, Troll.

I find that Necromancy no longer has the same "savour" with the modification made.

I had a bug where the AI tries to cast a spell on the Liches troop on which the "Anti-Magic" spell is active:
https://drive.google.com/file/d/1ChbSHveQPDzASn4hXNeaweFNSLtOi4V7/view?usp=sharing

Here is the save file.

Unknown_Hero wrote:

I find that Necromancy no longer has the same "savour" with the modification made.

It is very sadly...

Unknown_Hero wrote:

I had a bug where the AI tries to cast a spell on the Liches troop on which the "Anti-Magic" spell is active:

I will see save file, however, may be you tell, does this bug occur with original game or only with some of my DLLs ?