SORCERY Realm

[Nelphine]

First: yes, the triple realm and the mono realm have the same number of spells. But with the same amount of research spent, the mono realm has more spells finished being researched. Under the current system, that means the triple realm always looks stronger, assuming roughly equal skill.

It doesn't matter how much of the bar is research. What matters is using the bar for comparison purposes. If one player is super ahead, I don't care. That player is winning. The only time the bar matters, for comparison purposes, is when the two players are fairly equal.

So if I know that two players (myself and an AI, or two AI) have roughly the same skill (which, assuming equal size, and accounting for sorcery shenanigans, is the baseline assumption), then I look to the bar for comparative strength. If one person has the same or more spells in all tiers, that player will be stronger. If one person has more spells of a higher tier, and the other person had more spells of a lower tier, and the difference is equal, the one with more higher tier spells is stronger.

Therefore, the only case where the bar actually tells me something is when 2 players have differing amounts of different tiers - and one has more of one tier, and the other has more of a lower tier, and has noticeably more spells overall.

In this case, triple realm vs mono realm. The mono realm has researched 10 Commons, 8 uncommons, and 2 rares. The triple realm has (with the same research spent) researched 22 Commons, 9 uncommons, and 1 rare.

Now I grant you, there are certain combinations of Commons that will make those 12 Commons + 1 uncommon stronger than that 1 rare. But 80-95% of combinations, the rare is better. There are simply not that many Commons that actually add to your power rather than overlapping. The triple will already have started with the best 3 Commons of their realms, and they already have enough uncommons that a lot of their Commons are already obsolete. And the first few spells of a given rarity are usually very big jumps in your power.

But the triple realm will be shown as having 30% more research power than the mono realm.

[Seravy]

I still don't understand.

Let's say it's the case you mention and in my system it's a ratio of 75 to 150 in research but in yours it's 150 to 75 on the mono and triple realm wizard. (Numbers don't really matter, they are just examples, what matters is there is a significant difference.)
Both wizards have somewhere between 125-175 skill. (again what matters is not the numbers but that the gap is small)
Research weights one third of the skill. (25 vs 75)

Assuming 75% of the bar is made from skill, we'd get these bar lengths :

Bar 1 : (125 to 175) + (75/3) = 150 to 200
Bar 2 : (125 to 175) +(150/3) = 175 to 225

You can't know the casting skill of the wizards for sure. So if you see a bar length of 200, you won't be able to tell which wizard you're seeing, regardless of which system is implemented - both can be rated 200 in either, it just means a +/-25 difference in casting skill. That's about 3 Amplifying Towers. Basically, the minor uncertainty in skill far outweights the actual major gap in research so it doesn't matter which system is used.
You might see "wizard 1 has 10% higher bar than wizard 2" but that can either mean that wizard 1 has a 100% advantage in research but is 15% behind in skill or can mean they are 13.3% ahead in skill compared to wizard two. That's literally not meaningful information on research unless you open the save file in tweaker and look the exact casting skill for both wizards.
And god forbid the wizard to be an archmage perfectionist who grew casting skill to ungodly proportions and built all the amplifying towers, having a bar twice as long as anyone else while still providing no information on their research whatsoever.


...okay I think it's time to ask a silly question. If you knew what spells/tiers the enemy had to begin with (you did) and used that to try to deduce what skill the enemy had, then why the hell do you want the bar to contain information about research? I don't understand, you are the perfect example that players already know the research progress of the AI if they pay at least a little attention. It's kinda hard not to notice when a global is cast, or a huge dragon is flying across your continent. Combat spells tend to remain hidden but overland spells are painfully obvious. So why change the system to provide more information on something already known? Are there really players who don't know which tier the enemy is at in their research except those who play on Normal? Why is this information important enough to appear in that bar? If it's not then why care how accurate the 25% of the bar is, when the 75% overwhelms it anyway?

PS : I agree the current system overrates commons and underrates very rares. What I disagree with is that the amount of bar this result in is relevant enough to care. If we did, we would need something completely different anyway, like each spell being worth 3 times as much as the previous tier but each spell in the same tier being worth 0.9 times as much as the previous spell in that tier.
...then again the time we spent on talking about it is probably more than it would have taken to implement that system. But I still don't want to do it if the difference is so small no one would even notice.

[Nelphine]

Honestly, it's probably the perfectionist in me at this point. My original issue was actually that I was trying to use the bar to estimate skill, instead of the other way around. I wasn't at war with my opponent, and I was trying to use the bar to correlate whether he should be casting spell binding instantly, or if I should look into his actual skill and see if there was a bug going on.

In this case, I was trying to use the knowns of my own spell casting, and my own research, with the mostly knowns of his research, to try to work out his spell skill. (Since I wasn't at war, and he's been casting numerous unit buffs, I had no other way to guess his casting skill, except by checking maximum number of summons per turn - he got 4 djinn one turn - but that is a fairly inexact estimate due to unit buffs not being shown when he casts them, and not knowing how much of the first one was from the previous turn.)

Since he was ahead in very rares, I mistakenly thought the research portion of his bar would be larger; since he was around 90% of my bar (maybe as high as 93%, but that bar is really hard to tell that specifically), I assumed he would have to have a skill no higher than mine, and probably slightly lower, in order to balance him being slightly ahead in research.

However, the current formula for the bar weights quantity of spells much more highly than quantity of spells, which I didn't expect, and led to me thinking there was the bug with the Spell Binding spell cost modifier.

[Seravy]

While working on making Spell Binding not disappear from the spell list, I found there have been some problems with the turn counter, not only did it use a memory address that overlapped with Time Stop's counter but also one of the addressing was wrong and used 5FD6 instead of 5DF6 so whatever was on that address got corrupted (unless it was a 0). I'm surprised this wasn't found earlier and seemed to be working during the initial tests.
Long story short, the current version's Spell Binding seems to have a broken turn counter. You might want to wait for the next release if you (or the AI) will be using the spell.

[Nelphine]

Crap. Ok thanks. I think I've got.. 3 or 4 turns left till I can recast it.

[Seravy]

Well, considering the counter decrements the wrong address, idk about that 4 turns.
Meanwhile I realized the "fixed" version isn't the same address as the broken one so my guess is you'll end up with 0 turns left but 12 uses and will need to use the tweaker to fix it. (In fact you definitely should unless you like having a broken Time Stop counter.)

[Seravy]

btw we seem to be lucky, I looked it up and -5FD6 is an unimportant temporal diplomacy variable - overwriting it doesn't really do anything. So at least there was no data corruption from this bug.

[Nelphine]

Can we change the order of what is applied first, shadow attack weapons, or focus magic - currently, focus magic takes priority, which is.. ungood.

[Seravy]

That...shouldn't be possible. The "apply item power" procedure is literally called before the "apply unit enchantments" one.
But! We changed Focus Magic to look at the base attack type on the unit for some reason (I think it was for Focus Magic boosting itself since if it already added the attack type, next turn the unit would be eligible for the +3 attack boost as well). So the order is correct but Focus Magic will override Shadow anyway. However if I read this right that shouldn't affect the outcome : attack strength is set to 3 ONLY if it's below. So while the thrown should change to ranged, the strength should not be lowered which is exactly what happened before the change, anyway.

...Unless, that's only true for turn 1. On turn 2 the unit already has a magical ranged attack so Shadow cannot be applied to it, thus the attack strength stays zero and raises to 3. 

Which means we also need to change Shadow to look at the base attack type instead of current to fix it? I'll add it to my todo list...

[Nelphine]

Interestingly, it also results in a strength 2 magic attack, not strength 3.