Test games played

[Nelphine]

I think it's a bit worse than that Zitro. Typically, if the human picks Myrran, then there are 3 wizards in Myrror - so would they need 50% more land that Arcanus?

No I don't think so. I think Arcanus in that situation needs equal land (or more!) to Myrror. In that situation, the Arcanus wizards are going to be the late game opponents. BUT they'll be fighting each other. So they'll be slowed down as much as the wizards on Myrror are. But the assumption is that the human is going to dominate the plane they start on (otherwise they lose). That means when the human does get to Arcanus, they'll have the whole plane, while the 2 remaining opponents have half a plane each. Even at Myrror = Arcanus, this is no contest. Human wins.

So in this case, I would say Arcanus needs to be 50% larger than myrror (that way each of the 2 Arcanus wizards are 75% the size of the human after the human conquers Myrror).

Then we go back to standard human on Arcanus plus 3 AI. If Arcanus is 50% larger, that means each wizard on Arcanus is going to be a little more than 1/3 the size of the Myrran wizard. However, again, default assumption is that the human dominates the plane they start on. So, by the end game, the final boss wizard .. is only 66% the size of the human. In order to be equal, they need to grow 50% faster. No wars is definitely faster, but 50% faster? I'm not sure its THAT much difference.

My personal guess is that no wars lets someone grow closer to 33% faster.

And what happens if the human is Arcanus, but gets the rare case of 2 Myrran wizards? Now suddenly, they have wars, and aren't growing faster at all.


I think the solution is actually to break towers faster on purpose, so that the Myrran AI gets involved in wars, and the other AI have a chance to expand to Myrror, instead of trying to come up with a formula that covers all the possibilities of wizard distribution between the 2 planes.

[Nelphine]

In case anyone cares, my expert 'test casting skill shenanigans' game, I have 154 casting skill in Jan 1408 (including 2 amplifying towers), and I have all of Myrror to myself to develop further. I've only got 3 nodes so far (I'm super behind on nodes because of my choices); I've also got over 20,000 gold, which is kind of obscene and I really don't have any idea what to do with it (well, not true; as my beastmen outposts grow into hamlets, I'll be buying amp towers everywhere). And I didn't get any dark elves, which is sad, and will drastically reduce how much skill I'm going to grow into later (beastmen just don't produce anywhere near as much power.)

[Seravy]

In case anyone cares, my expert 'test casting skill shenanigans' game, I have 154 casting skill in Jan 1408 (including 2 amplifying towers), and I have all of Myrror to myself to develop further. I've only got 3 nodes so far (I'm super behind on nodes because of my choices); I've also got over 20,000 gold, which is kind of obscene and I really don't have any idea what to do with it (well, not true; as my beastmen outposts grow into hamlets, I'll be buying amp towers everywhere).

Wtf is all I can say to that. So, which retorts will need nerfing again...? (I'm more shocked by the gold than the casting skill, wish I remembered how much we assumed for this time. Was it 40 for 1406 and 100 for 1412? Then this is "only" about 2.2 times what it normally should be...)

I think the solution is actually to break towers faster on purpose, so that the Myrran AI gets involved in wars, and the other AI have a chance to expand to Myrror, instead of trying to come up with a formula that covers all the possibilities of wizard distribution between the 2 planes.

That defeats the purpose of the entire system. We are better off with only one plane if we want towers to break early, there is no real difference...

We'll need to look at each case individually, and have an appropriate modifier where needed. (Human included in totals)

2 wizards on the same plane : Good as is.
3 wizards on the same plane : Good as is.
4 wizards on the same plane : Good as is.
5 wizards on the same plane : Good as is.
1 wizard on each plane : Good as is.
2 wizards on one plane, 1 on the other : ???
3 wizards on one plane, 1 on the other : ???
2 wizards on both planes : Good as is.
4 wizards on one plane, 1 on the other : ??? (this is what we most often run into and have problem with)
3 wizards on one plane, 2 on the other : I think good as is, the difference isn't that great (and the human is usually in the "3" group)

These are all the possible cases for any enemy wizard setting and Myrran choice. We don't really need to care which plane is Myrror - obviously, the plane with fewer wizards needs to be the one that is smaller.
So, from a coding perspective, I'll have a 5x5 matrix and we use the amount of wizards on each plane as the indexes. So, element [2,1] would tell us the modifier for the case when Arcanus has 2 wizards, Myrran has 1. This would take up only 50 bytes. Perfectly doable, assuming we have any decent amount of space, but I think we will. The code to count the wizards might even be longer.

So all we need is to decide on numbers where there are ???, based on what we see in our next 5-10 test games. Those that need no modifier will have a 0.
The other thing to decide is whether we want to achieve this my raising the size of one plane, or by lowering the other, or by doing both. Remember that more territory does shift the game balance towards "early game expansion wins".

[Nelphine]

The game I sent you is my casting skill test game.

2 life, 1 chaos, 4 sorcery, omniscient, cult leader, astrologer, myrran, archmage. Dwarf. Put a dwarven city on every crystal on the plane, put dark elf (or beastmen if no dark elves *pout*) city everywhere else. After dwarf cities are done, maximize number of cities to get more amp towers, do NOT maximize quality of cities (so even if you lose 1-6 max pop, and a bunch of production or gold, it doesn't matter, as long as you get an extra amp tower).


On topic of myrran wizards, I want to note, that I do NOT think 3 wizards on one plane and 2 wizards on the other plane is balanced (human is Myrran, so 2 Myrran opponents, and 2 Arcanus opponents being the most common occurence of this). This game is so much unbelievably easier than 1 Arcanus human, 3 Arcanus AI, 1 Myrran AI.

I think the CONTRAST between these 2 gameplays is actually the biggest problem. I believe NOT choosing Myrran increases the difficulty by about 50% (so if the difficulty is 6 for Arcanus, then the difficulty for the same game but with the human choosing Myrran is 4). In reality both games should have a difficulty of 5.

So especially if you test a game where you play Myrran, and then your next game you test is where you do not play Myrran, then the game seems far more difficult. In reality, if you do NOT choose Myrran, then you should be playing the game differently. Having one opponent on the other plane, necessitates different play choices - and i think this is a GOOD thing. It means choosing Myrran has an extremely important element in terms of replayability - it is in some ways a different game.

So yes, I do believe that there needs to be some modifications. But they are NOT all because the single Myrran AI is too hard. It is ALSO because the other game styles currently available are relatively too easy.

Specifically in the topic of what we are doing now, if we do go with a modified land size, I do FIRMLY believe that if the human chooses Myrran, then ARCANUS should be the bigger plane, in order to make the wizards on the other plane more dangerous when they eventually are met.

[zitro1987]

I do believe that if you play myrran, the arcanus world will be parallel in size. the difference is that in myrran you are normally battling 2 wizards (total of 3 wizards) while in arcanus there's just 2 wizards in whole plane. So if playing Myrran, arcanus IS BIGGER on a land to player ratio without changing anything, which fits the arcanus wizards being more dangerous. This becomes more pronounced if playing the 3-AI option instead of 4-AI.

And if you play arcanus, the wizards can be packed a little too closely while the myrran AI (almost always alone) has an extreme land advantage. This makes the game a lot more difficult as the land to player ratio is extremely low in arcanus and extremely high on myrran. This needs to change.

My attempt at percentages for starting plane (and opposite plane)

2 wizards on one plane, 1 on the other : ??? -> 55% (45%) (one is 22% larger than the other)
3 wizards on one plane, 1 on the other : ??? -> 65% (35%) (one is 86% larger than the other but the land per player is still lower)
4 wizards on one plane, 1 on the other : ??? -> 70% (30%) (one is 133% larger than the other but the land per player is still lower)
2-3 wizards on one plane, 0 on the other: probably also 70% / 30%? or leave as is? Depends on whether we want gameplay to shift to an empty mirror.


If these advantages and disadvantages are toned down, we will have much more more balanced gameplay and difficulty among arcanus games and myrran games.

[Nelphine]

I think i miscommunicated my point.

My point is that by having 2 wizards on Arcanus when the human picks Myrran, those 2 wizards have to split that plane. There is NO end game boss. And neither of those 2 wizards is strong enough to fight the human player after he conquers Myrror.

Compare this to when the human does not pick Myrror. Even after the human conquers all of Arcanus, the Myrran AI is still as big as the human. He IS strong enough to fight the human player. (And because the Myrran AI was in no wars, he was free to focus on economic expansion and is actually ahead of the human.)

I think the 'end game boss' design of Myrran AI is a GOOD one. But by choosing Myrran, you just take that out entirely. There are no strong end game AI at all. Which means if the 2 Arcanus wizards are to pose any threat to the human who picks Myrran, then ARCANUS must be bigger than Myrror, despite having fewer wizards.

Edit: Remember, the default assumption is that the human fights through the wizards on his own plane before he fights the wizard(s) on the other plane. This means once the human starts fighting those other wizards the default assumption is that the human CONTROLS THE ENTIRE PLANE, regardless of how big it is, or how many wizards started there.

Edit2: This default assumption is why I'm thinking we're looking at this completely wrong. We either have to change that assumption, or we have to ensure that no matter what, the plane that the human starts on is NOT bigger than the other plane. (This default assumption is also why peaceful strategies have very mixed results. The only way to really make them work is to have a much bigger economy, which, since you're peaceful, you can't get through having more cities. So, you really have to work to make it successful. Which, you can do via nodes, so they do work, they're just a lot of work.)

[zitro1987]

I did understand and noticed the error of my earlier suggestion.

So my current suggestion has it at 50%/50% as opposed to starting plane being the bigger one.

As far as keeping the 'end boss' feature to myrran games with 1 arcanus (even with 4 AIs), that'd bring back the earlier debates whether we (including Seravy) wants this to happen. Otherwise, a clause of 3-2 where player is myrran may justify a larger arcanus.

[Seravy]

I do NOT think 3 wizards on one plane and 2 wizards on the other plane is balanced (human is Myrran,

Or not. It's uncommon but the game can roll 2 Myrran AI wizards in a "normal" game.

But they are NOT all because the single Myrran AI is too hard. It is ALSO because the other game styles currently available are relatively too easy.

Well, those two statements are the same, so of course. If something is relatively too easy to something else, that something else is relatively too hard compared to the first thing.

In my recent Myrran games, I found the difficulty to be fairly appropriate for the setting. (all Dark Elves though)
In the Arcanus games, the difficulty felt too high (all Halfling however).

Since we can't conclude something based on only a single race, I'll have to play more games.

Specifically in the topic of what we are doing now, if we do go with a modified land size, I do FIRMLY believe that if the human chooses Myrran, then ARCANUS should be the bigger plane, in order to make the wizards on the other plane more dangerous when they eventually are met.

I'm not so sure about that, the difficulty in my Myrran games felt appropriate. (All Dark Elves is hardly conclusive though)


By the way.
Not a fan of it, but if we want to cut down on testing time, we can opt to implement this now, without more testing. If Myrran wizards are so out of control that players feel it's unbeatable and quit, that's a very bad thing, and needs immediate solutions without a delay. I have no idea how bad it is without testing though, so that kinda puts me back to square one.
I guess after watching that video I'll know slightly more but I still think I should at the very least see games where the Myrran isn't Dark Elf to be able to judge the situation.

[Nelphine]

One of the problems is that, I would say.. 70-85% of games will have either a dark elf AI, or a dark elf neutral, which has priority. So in almost all games, no matter what the AI starts as, they end up as dark elf.

Edit: if unrest affected power production of population that might help?

[Nelphine]

I do NOT think 3 wizards on one plane and 2 wizards on the other plane is balanced (human is Myrran,

Or not. It's uncommon but the game can roll 2 Myrran AI wizards in a "normal" game.

Not that it matters much, but you seem to have missed the rest of the bracketed sentence:

human is Myrran, so 2 Myrran opponents, and 2 Arcanus opponents being the most common occurence of this

I absolutely acknowledge that it can happen in other situations. The case where human is myrran which generates 2 myrran opponents and 2 arcanus opponents, is simply the most common variety.