|
Author
|
Comment
|
Jolly Joker
2002-01-05 10:15:15
|
Ok, I finished the Dungeon arena and I have a lot to say...
...I did the whole Dungeon tour (the fights AGAINST Dungeon), 8 rounds with 4 fights and 1 round with 5 fights.
Castle did 5 fights, the last one a siege against a 20 spell power Warlock with all spell books and at least two expert magic schools. In all the 5 fights I lost around 60 Halberdiers, a good Dozen Griffins, 5 Zealots, 3 Marksmen and, most notably, 2 Champions.
I may be a good player, but I'm not the best, that's for sure. However, there's not that much room for better results there, right?
I feel, there are some things that could be made better in those maps; since I made a series of tourney maps myself (the Battleground series for Ungo's site) I know my way around the editor and I know about good AI play. Here are some notes:
1) Why magic heroes only?
2) Why 8 pre-determined skills? (Not necessary to guarantee same situation for all players)
3) Why start with 0 experience? (Unreal; no hero with such an army ever had 1 experience)
4) Why hand out full spell books and not spell selections? (For instance, Conflux with Armageddon? Even against Dungeon I blasted myself through; just imagine how Conflux does against the others; furthermore, in battles consiting of parties both without access to spells like Sacrifice, Armageddon, Curse, Prayer and so on it makes no sense to hand those spells out)
5) Why make Heroes so equal ?(all have the same skills except Stronghold).
6) Why "alien" bonus fights? (Why not build a series of towns all defended more heavier than that before, for example?)
7) Why not at least one mandatory siege? (In a normal heroes game you usually can't win without a siege).
I want you to know that this is meant as a constructive post. I'm not just critisizing, I have suggestions also (but maybe no one is interested). One last thing to show that there are definitely some things wrong there: In the point list for creatures Familiars are listed with 17 points, while Hobgoblins are worth 44. A superficial comparison of a Familiar and a Hobgoblin showd that the main difference between both is: Familiar has 4 HPs, Hobgoblin 5 while the Familiar has the mana channelling special. Can't be right, can it?
JJ
|
Guinea
2002-01-06 00:57:13
|
Some answers
Hi and thanks for your comments.
I'll try to answer your questions.
1) Magic heroes only, because we already had a Might Tournament and wanted to have a Magic one.
2) Exactly, we wanted to guarantee the same situation for all players ;))
3) No idea why start with 0 exp.
4) We wanted heroes to have all available spells, not just a selection of them.
5) And why not?
6) Extra fights aren't obligatory, so you don't have to play them if you don't like them. Why are they like that? That was just an idea everyone agreed to.
7) Those maps weren't created to simulate a 'real' H3 game. They were created just to test battling skills with using magic.
Finally, the units' values were counted how they fight against each other. I didn't participate in that test myself, but I heard that 100 units of each kind fought against 100 of every other units and number of those that survived was counted if you understand what I'm trying to say ;) For example 100 Archangels fought against 100 Halberdiers, then 100 Archangels fought against 100 Marksmen and so on against all 63 upgraded troops. And then 100 Champions fought against all 63 troops and so on, and the number of units that survived was counted. Familiars are very easy to kill and their specialty isn't important in such test, and that's why they are worthless ;)
|
Marcus the Red
2002-01-09 09:39:19
|
The Creature Chart
For Lopez and the other designers:
I understand how you did it-- it was mentioned earlier-- but I think the process had some significant errors because it was done 1 on 1 (or rather, 100 on 100). The point is to try and get as accurate a representation of creature strength in "real life" action, because that's where you'll be using them-- in battles on the Arena maps. The method used does not fully accomplish this. It's a good foundation, but tweaking needs to be done, as it weights creatures with no retaliation too high, shooters generally too low, and does not always allow specials to be factored in. For example-- what good is the Chaos Hydras multi-hex attack against a single stack of X creatures? No good at all-- but let the CH attack 2 or 3 or more stacks at one time with no retaliation, and they should be worth significantly more points. Same thing with the 2-hex attack of the black and gold dragons.
I don't have the solution to the problem, and I understand that other methods are even more time consuming, but I do think it's important for the designers to realize that their scoring system is skewed. Equally skewed for everyone, but skewed nonetheless. Personally, I don't see a significant difficulty with going through and adjusting the pt. values based on playing experience. Humans know the value of the specials, and how various creatures are best used-- so let a few humans (probably the designers, but players could be used to), go through and "tweak" the points values, then average all of those various values. So, maybe I do have a possible solution. =)
Again, as JJ noted-- the time and effort put into the maps is obvious. The quality is top notch, the concept is excellent, the end result is very good-- but it could be better.
|
Jolly Joker
2002-01-09 10:52:45
|
A point system of the designers already exists...
.. you can find it in the crstats.txt file that you will have access to with the Modhomm utility. That sytem, however isn't accurate either (!) as everyone with some playing experience will see after a closer look.
It is of course very difficult to find a system that does justice to every factor in the game, but one basic mistake of every rating system based only upon fights against other creatures WITHOUT the presence of heroes is obvious: in a real fight all specials influencing possible spells will play a role, in a fight against other creatures that is usually not so. For instance, the immune-against-mind-spell of the Titan is obviously a big plus in a real fight, but not against other creatures except against Unicorns (I think).
Furthermore the question is, of course, whether a human does the fights or the AI. A human will make full use out of the waiting possibilities higher speed offers, especially against or with shooters, while I'm not so sure about the AI here (in fact I'm pretty sure it is weak).
I made up a rating system based upon the numbers. Essentially the rating is the square root of the product of Health and damage, modified by defense and attack, respectively, and a factor for speed and all specials. Shooters are rated a bit different based on the fact that there are certain different situations for shooters (full distance, less then 10 hexes distance and siege). Anyway, while the system in itself is pretty accurate, the question is WHAT factors are "factored in" for speed and specials or what weight they should have. I'm pretty satisfied with the way I solved the speed problem, but my way of rating the specials is at least debatable (of course). Another question is, HOW to quantify certain creature specials. Anyway, the result of this can be seen since over a year or so at Ungo's site and I'm pretty surprised I never heard something from it.
JJ
|
Beltion
2002-01-09 11:29:57
|
Ungo's site...
is Portals of Might & Magic for those of you wondering.
|
TITAN TASE
2002-01-13 07:31:34
|
We should better play than talk so mutch...
Yes I agree that some things could be better especially in the way the creatures were ranked but there is a lot of magic around the battles and I guess that ranking a creature all by herself(with all his specialities) is not good in a M@M arena .
JJ says that 100 grand elves are better but if I cast a forgetfuless on them even a gargoyles is better for exemple.
The point is that some creatures have special abilities but they also have WEAK POINTS that I'm very corious how JJ or others will take them into considerations.
Someone said that hydras didn't used theys multi-
attak ability but neither do the archangels! If they were 2 staks of them (500 and 500)it would be a nightmare for the hydras.
So I say that we should concerned more on playng and have fun.
What does it matters if a hoglobin has 44 points? He would have the same score for all of us!
Like that I'm wonder why the Phoenics grow 4 per week? but I still like A LOT! Heroes3 .Maibe the differences makes it that nice.
I leave you now because I got an exam tomorrow and I still want to save some sime to complete some Battles.
Until again I salute al of you!
|
Jolly Joker
2002-01-06 02:36:37
|
Have to get into more detail here, it seems...
1) The thing is called The MIGHT&MAGIC ARENA. I only see the magic.
2) You don't know your game, then. A saved game will always offer the same skill picks. Example: You start a saved game with a normal hero who begins with, say, basic Wisdom and basic Mysticism. The first lvl gained will offer always the same pick, say, advanced Wisdom or basic Leadership. So now there are 2 ways to go for each and every player. All those who pick advanced Wisdom will get the same pick for lvl 2, say advanced Mysticism or basic Earth Magic. All those who pick basic Leadership will get another pick, but the still the same for all, say, advanced Wisdom or basic Eagle Eye. And so on. So no matter how many skills you predetermine, once all players start with the same saved game all those players will get the same skill picks (depending on the way they go).
3 and 4) Not such a good idea, I think (just think about all the spells a towns is not entitled to have. As I said, in a fight, for instance, Stronghold against Fortress where are all the high level spells from? And Conflux with Armageddon? Great.)
5) Why not (what an answer). Because they are different, that's why! The game is called HEROES of Might and Magic and not because all Heroes are equal.
6) Yeah, but my question wasn't about why there were bonus fights. My question is about why there are bonus fights with ALIEN creatures. I would have suggested a road with more and more difficult fights against the same town, for example.
7)Again that's no answer. Siege battles are battles, right? And Ballistics is an interesting skill in siege battles and more interesting for some towns and for others. This has something to do with points 2 and 5 as I see it. More freedom in forming the hero, getting Ballistics for the siege(s) or not would certainly make for more different battles.
 Sounds like a lot of work, but when the result is so silly, there must have been something wrong: As I said, the difference between a Hobgoblin and a Familiar is: they do the same damage. They have the same speed; Goblins have 1 more attack. Familiars have 1 more defense. Familiars have 4 HPs. Hobgoblins have 5. Familiars channel mana as a special. No one in his right mind will understand why a Familiar should have a worth of 17 and a Hobgoblin a worth of 44. That would mean that 1 Goblin is worth 2 and a half Familiar which is silly because it just isn't so. And if the method of calculating those points deliver those results common sense dictates that this method can't be right. Right?
JJ
|
ObywatelEracji
2002-01-06 07:03:29
|
Re: Have to get into more detail here, it seems...
Greetings Jolly Joker !
First, I would like to welcome a new competitor and, judging by appearences, very experienced warrior ; )
Since I was one of beta-testers and as far I can remember familiars vs. hobgoblins was my part, I'll try to answer your question. There is no doubt that hobgoblins are more powerful than familiars and the best way to make sure of it is... fighting 100 hobgoblins against 100 familiars. Try switching the sides, it won't make any significant difference - the hobgoblins win either way. Actually, in this case it's the higher HP that determines the winner. Hope I could help ; )
OE (Erathian citizen)
|
Beltion
2002-01-06 08:05:16
|
My take...
Sure, it's pretty clear that Hobgoblins are more powerfull that Familiars. The one hitpoint extra makes them better.
But JJ makes a good point that I agree with, they are not that much better. Sure, 100 versus 100 the outcome is clear... but if the Hobs are valued at 44 and Familiars 17 - wouldn't that mean that a Hobgoblin is worth as much as 2,5 Familiars? 2 Hobgoblins or 5 Familiars? If I had to choose between those two I'd take the Familiars in a heartbeat. Even with a multiplier of 2 I'd take Familiars every time.
Getting back to the two slugging it out, looking at the values you could think that a fight between 100 Hobs and 250 Familiars should be even (4400 points of Hobs, 4250 point worth of Familiars), with Hobgoblins coming out on top a bit more often than the Familiars do... and that isn't the case, so I too see a flaw here.
[edit] Hmmm... wrote this before seeing JJ's latest reply. Seems sort of redundant now. But hey, it's not the first time I'm redundant...[/edit]
|
Jolly Joker
2002-01-06 07:48:44
|
The question isn't that Hobs are better...
...the question is, how much better. The file with H3trio gives a rating of 44 for a Hobgoblin and a rating of 17 for aFamiliar. What that means is, that a Hobgoblin is worth more than 2 1/2 Familiars and that can't be true (why must I tell everything again?).
|
Jolly Joker
2002-01-06 12:20:14
|
There are some other flaws with the system as well...
... For example a fight 100 Archangels against 100 Mighty Gorgons will see the Mightys destroyed before they even had a chance to bring their special into play. They won't do better than 100 Familiars here. On the other hand it is clear that it is exactly the special coming into play that makes them so valuable. With this method the special will affect only lower level creatures and for them it is just an additional little bonus. So I'm not surprised that T-Birds are rated higher. However, in reality a fight 100 AAs against 100 MGs is pretty unfair. Normally you'd have at least 3 times as much MGs as AAs.
A fair method would probably be to determine, how many creatures would be needed to kill a certain number of every creature and how many of each creature would be needed to get a fixed number killed. That way the special of the Mightys would work against higher creatures, but not very good against lower level creatures. Probably this would be even the best method. But it would be a lot of arithmetic necessary.
JJ
|
jOjO dwarf
2002-01-06 14:18:49
|
Hi to everyone
especially to honourable Jolly Joker
It's my fault ;)
2 years ago I started to making tournament (after this great moment, when I saw Ghwen's one at H3Trio)
I wanted to find out, units of which race are best
It was planned to be tested by real players, what race is the best.
My idea was simply: n units of XXX vs n units of YYY
4 figths (basic vs basic // basic vs up // and so on)
I thought, that it's clear. Then Behemoth published it on his site.
That was ßtest (I was thinking about magic, but we planned it for later)
ßtests was successful, so after some improvements it started as well-known Might Tournament ST
After one year, with group of enthusiasts, we tried to create second part of this tests - New Tournament (NT)
it was after all named M&M; ;)
It was nearly 3 months discussion on concept.
The idea is clear (I wrote an extra documentation on it!) -
* primary skills - appropriate to table of probability
* secondary skills - only good skills (rest was removed from list) and ranked appropriate to table of probability
* magic heroes - because we don't want fighters ;)
* tome of magic - it was interesting conception (too much discussion on what to choose, if e.g. leaving magic scrolls)
* number of units - it was easy ;)
* ranking of units - toughest topic of our discussions - tests!!!
* saves - the truth is out there ;) it's because of cheating!
So, a short resume:
our ideas have been clearly defined and realised!
I know, the name of this tournament is not accurate ;)
but we want to have some specific fun with our top game
ST is almost 2 years old (counted from the first idea!) and it's fans are still sending saves. We like it ;)))
NT is only 1 month old, and more and more fans want to play
I think, that it's similar to situation with chess and war
Our tournaments are not really like playing HoMM, just like chess are not real war
but many people find much pleasure with playing chess, upss ;) our tournaments, I mean
after all
Do you think, that our team can invite you to cooperation on making H4 tournament?
I think, that we have a good practice, very good testers, and many opinions - so it should be more realistic, better tournament.
So, how about you?
|
Jolly Joker
2002-01-07 03:51:26
|
First of all...
I already made my own "tourney"; the maps are available at Ungo's Portal site under the "Battleground" icons. I have a lot of experience with the SoD editor and know that you can do pretty interesting things with it. I think, there were not many participants for our tourney maps for a certain reason (aside of the fact being not known): these maps are real maps and the opponents are real opponents; I took much care to make sure it really makes a difference what difficulty level you pick (map score was result) and of course people want to play on impossible difficulty. On impossible difficulty, however, you must be a really good fighter to make it through all the 14 battles of one map. The reason for that is mainly, that the opponents in many cases are intelligently casting their spells (they have the right schools and the right spells for their town, and that is true for the Might heroes, too).
Furthermore I made my own unit ranking system based on mathematics (I think, it's pretty accurate because I spend a lot of time with specials, testing and so on). In this ranking system, when an Archangel would be rated 5111 points as in yours, a Familiar is worth 28 points and a Hobgoblin 31 (with a Halberdier having the same 52 points. With your system many specials simply won't have any effect, but wil have as soon as heroes fight (and they do in your tourney).
Then I spent a lot of time with the way Heroes develop (human AND AI). Here, and this is what I really don't like with the new tourney maps of yours, you blundered because you probably didn't know what is happening there. What you need to do is simply removing all skill useless or not wanted, Learning, Pathfinding and so on. Then, when you have a saved game, the development path is always the same (I explained that in my previous posts). What that means is, that all heroes will be alike at the end - provided the players all have the same preferences (and they won't have). But a saved game is enough to guarantee the same chances for every player without having to pre-design heroes (I think that is the major flaws of the maps).
I can't say whether I'll participate on a H4 tourney because I don't know yet, whether H4 will be worth it. I would rather make sure that this tourney is a challenge; I mean, the opposing heroes are simply too weak with their limited spell selection. For example, only the water hero is not helpless against Blinding; that's just silly, if you ask me. As I see it, the major difficulty of the "course" is finding the "right" path: for every town.
I would have liked to participate here, had I known about it, but now it seems too late. The impression is, however, that this is a tourney designed to find out, how to exploit the weaknesses of the AI. I played the whole Dungeon arena and no matter what, Dungeon should be a town that inflicts losses! Now, when you have a look at my saved games you will see that the losses weren't so grave at all. With Castle I completed the whole round with all bonus fights and these are the troops I have left: 11 Archangels, 20 Champions, 28 Zealots, 43 Crusaders, 70 Royal Griffins, 96 Marksmen, 80 Halberdiers. I think, this shows that the maps are not as they should be because it shouldn't be possible to keep that many troops. I'm not the best player in the world, but there isn't much room to do better there and there should be plenty.
All fights with Castle will be easy.
All fights with the Conflux will even be a farce. I mean, even against the Dungeon I Armageddoned my way through needing only a few troops. But what will happen against the others?
Clearly the hardest fight against Dungeon was with Stronghold, but that was only because the Stronghold hero has Air Magic only (and I still managed to keep enough troops) which is pretty dumb, if you ask me: I've never in my Heroes life seen a Battle Mage with Air Magic as the only magic school.
So, great idea, but I think the current way is too much tricking the (weak) AI heroes into something and exploit that (or having all the weaponry and the AI has none) instead of real hard fighting that only comes when the AI heroes have expert magic schools (which they don't develop often when left alone) and the appropriate spells.
You may visit the Portals site and have a look at some of the Battleground maps I made there just to see what I mean. Start a map, save the game right after the start and begin with developing a hero. Start the saved game anew and develop the hero again. Try a first fight, for example with Castle against Dungeon (will be hell even though you have troops and your hero is developed. Or try to attack the Tower (you will meet Neela in the field). This can easily be desastrous, even though you have all you need. However, 9 times out of 10 the AI will be DAMN good and that's how it should be.
JJ
|
jOjO dwarf
2002-01-07 04:26:24
|
small notice
the main problem is, that our targets were not same with yours
(I see, now, that - maybe - it's not tournament, but kind of fun ;) )
It's realy too late, to change this ;(
maybe the next tournament will be better
but what about fights ForXXX or StrXXX?
they are also so simply?
|
Jolly Joker
2002-01-07 04:58:42
|
Fortress is. Stronghold isn't because of the rather...
... wild limitation of Battle Mages to Air Magic only (I did say that already).
It seems we have different conceptions of fun. For me fun is NOT Armageddoning your way through a bunch of helpless suckers.
What I find disappointing is the way my posts are answered here. I took great care to play the maps and check things out before I came with questions and criticisms. I bet you didn't try one of the maps out I made, but it would have made sense because I mentioned some points you could have easily made better had you THEN known more about some of the mechanics of the editor and the game. This is of course hero development. It simply is not necessary to pre-determine those heroes as I mentioned often enough, but no one seems to be able to realize that it would be a whole lot of more fun when a player was able to start out with only the normal 2 pre-determined skills, then breezeover an event giving limited experience and then being able to pick a few skills prior to the first battle. As I said, the chances were still alike and we would probably see more different heroes. So this is clearly a point where the maps could easily be more FUN. Of course they could be more of a CHALLENGE, but that is something else completely and here indeed it is matter of what anybody prefers (and a majority seems to prefer the easy ways.
JJ
|
Ingham
2002-01-07 08:42:07
|
Re: Fortress is. Stronghold isn't because of the rather...
Critisism is always good JJ, but you shouldn't do so hard. We managed to create this tounament after many months of hard work. It is it's possible best form and we really couldn't make it any better.
We know that it isn't perfect, but what is in my opinion most important - EVERY PARTICIPANT HAS THE SAME CHANCES.
If it is so simple, it will be simple for everyody.
Castle fights are truly easy, but the rest is not (that's why, in Castle fights, you have an obligatory fifth fight).
There is no point in showing us the mistakes now, because it is impossible to change anything. The machine is moving...
I'm sure our next tournament for Homm4 will be unspeakably better. With your help, of course.
|
Jolly Joker
2002-01-07 10:58:37
|
You can always change things - of course you must want to...
... as I said a half dozen times: it is not necessary to pre-determine the hero skills to guarantee the same chances for every participant. Without thinking too hard I see a handful of things you could easily correct:
1) Combat order: 2 siege combats and 2 field combats would be okay.
2) The Tome idea may be a good thing for the player, but not for the opoonent AI heroes; they are simply too weak. This always accepts the fact that a lot of player heroes would have a lot of spells they are not entitled to have.
3) Hero development. Leave it to the players. Starting with a saved game everyone has the same chances no matter what (I explained it several times). It would be nice, however, to be able to pick Ballistics or not or Resitance or not and so on instead of having to fight with basically the same heroes: Leadership, Luck, some magic schools, First Aid (great!) and so on.
4) The point system could be fine-tuned; you are not the only ones with a rating system and your system has some obvious, umm, unbelievable scores (I mentioned a few).
Maybe I'm simply disappointed because of the limitation to magic heroes (which certainly sounded different in the H3trio announcements; I just didn't know you had a might tourney already).
As I said, you can always change things - you just have to want to.
JJ
|
Marcus the Red
2002-01-08 11:48:46
|
JJ's criticisms
Having read through all the posts in this thread, I will say this: I mostly agree with JJ. First off, if you haven't tried his maps over at Ungo's place, you should-- very challenging. The one downside to them, they take quite a while to complete (but, then again, there are only 9 of them).
The hero skills d/n have to be preset. JJ is essentially correct, on this. The one downside to doing it the "normal" way (i.e., starting with two skills and developing from a SAVED game so that the skill tree is the same for all), is that it would make refighting the first battle several times to test that skill tree almost a necessity.
And, yes, that table of creature points needs some work. Who beats who and how badly 1 on 1 (or 100 on 100) is really an awfully limited way of trying to determine the relative worths of the troops. It tends to way things like non-retalitory strikes heavier than should probably be the case, and tends to downplay the power of shooters (since there can be no turtling or the like). Thus Cerebri are rated higher than Ice Elementals and Grand Elves-- but taken in the context of a larger army, would you really rather have 100 cerebri than 100 grand elves? I wouldn't.
That's all I have to say for now, as I've only tried forFOR so far, and I've realized that the magic path I took wasn't the ideal one. Another point JJ made-- there really probably is one best magic path for each scenario, which is kind of limiting, I think.
As JJ noted way back in his first post-- this is meant as constructive criticism. I have no doubt that much, much, much, much time was put into this project, and I think it is very cool as it is-- but that d/n mean it can't be better.
|
lopez
2002-01-09 04:04:04
|
To Marcus the Red
I helped to prepare the table of creature points and I think I have to explain some things.
To prepare this table we have done battles between all HoMM3 creatures. Always we have taken 100 vs. 100. And nothing more, no war mashines, no artifacts, neutral ground.
In such a battle 100 Cerberi always beat 100 Grand Elves or 100 Ice Elemental.
In real game (probably) you choose 100 Grand Elves instead of 100 Cerberi, because you have another creatures to support Elves. But in test 100 vs 100 it looks otherwise. Therefore Cerberi are higher than Elves in the table of creature points.
Another way we couldn't manage to do this table or it would be much difficulty.
|
Jolly Joker
2002-01-09 02:34:23
|
So here is a suggestion...
I have to say that I've considered a long time whether I should hang my ass this far out the window, but in the end, why not? Seeing that Marcus the Red (one player I also remember doing quite well in the motw contest and therefore knows his way around) agrees with me reassures me that my criticisms at least have a true core.
So here is the suggestion then: Send me one original map (I don't care which one) and I will make some changes. I'm pretty sure I can make the map more fun AND more challenging and I can do that even with using magic heroes only.
Let me tell you again, that I'm awfully aware of how much work was put into those maps. I don't want to devaluate this work: the maps are well done, no question. But they just could be better (and what couldn't be?) THIS IS NO CHALLENGE! It's just an offer to try and help getting a really great "Arena" on the way, okay?
Should you agree, send the map to:
Ch.Jentzsch@t-online.de
JJ
|
|
1
visitor
in the last 15 minutes:
0
Members
-
1
Guest - 0 Anonymous
|
|
|
|
|
