|
Post by Fleck on Aug 6, 2011 21:31:18 GMT -8
Ask, and ye shall receive, my son.
|
|
|
Post by Fleck on Aug 6, 2011 21:55:13 GMT -8
ERROR MESSAGE "The program can't start because MSVCP100D.dll is missing from your computer. Try reinstalling the program to fix this problem." Eh?
|
|
|
Post by countlieberkuhn on Aug 7, 2011 2:25:14 GMT -8
That's odd. Never had that error message before :/
Aha. Internet says its because I have visual studio 2010 and you don't. Hmm. Need to fix this. Gimme a bit.
|
|
|
Post by countlieberkuhn on Aug 7, 2011 2:39:02 GMT -8
Right, hopefully this will fix it. I've downloaded dev-C++ and compiled it using that instead. The file size is larger, which hopefully means that it's adding in extra bits for it to work on computers without visual studio 2010. It's in the link below under SimpleArena.exe. Let me know if it still doesn't work, Fleck! blitzball.goatpen.org/miniarena/files/
|
|
|
Post by Fleck on Aug 7, 2011 10:14:35 GMT -8
It works now! Hooray for Count!
*canonizes you.
I'll do some tests and let you know the results, Saint Count.
|
|
|
Post by countlieberkuhn on Aug 7, 2011 12:19:51 GMT -8
Exxxcellent!
|
|
|
Post by Fleck on Aug 7, 2011 16:55:32 GMT -8
Okay, I've just got done simming 239 matches, SO YOU BETTER APPRECIATE ME. Right, now I'm trying to see if there's one straight job class that is overpowered. So I created thirteen characters, each one with three of the same job classes, and then I gave them stats that I thought would best suit that job. I then pit them head to head in three matches each to see who's the best.
I didn't notice any glitches or oddities, but I didn't read the results all that carefully. Mostly I was looking for imbalances. More good news concerning that: almost all the job classes won between 40 and 60 percent of their battles. There were a couple (gambler, summoner, and gunner) that didn't make that cut, but I think I can improve their builds and pick better stats. There's an interesting rock-paper-scissors thing going on, but it's not a guaranteed win if you have the favorable job class.
My results are showing a bias towards whoever's name is written first, but that might just be an oddity of the way I did it (the jobs at the top of your list tend to be better than the ones at the bottom, and the ones and the top were generally favored with more first billings), so don't take that to heart just yet. Further testing is required.
Anyway, I'm going to play around with straight job classes for awhile, and then I'm going to experiment with mixed classes, see if I can assemble an Unbeatable Build. Things are looking pretty good so far, though.
|
|
|
Post by countlieberkuhn on Aug 7, 2011 17:26:59 GMT -8
You are truly a God... of testing. GlaDOS would like you. The bias to whoever is written first is likely pure coincidence. I randomised the first turn to avoid that being the case Good to hear that no one class is dominating though. I'm surprised that the gunner didn't perform too well though, they can dish out some really heavy damage with a bit of luck. They're best suited to people with a little more magic than strength though, since enchanted ammo is a major source of burst damage for them. Gamblers are very hit and miss, but I hope you saw a lot of variety with them. Gamblers have about as many potential moves as the other classes combined
|
|
|
Post by Fleck on Aug 7, 2011 20:05:36 GMT -8
Ah, that would explain why my gunner sucks. I saw that gunners only have one skill that uses magic, so I decided to focus more on strength and left the magic stat to be relatively weaker. If the enchanted ammo is really as big as you say it is, then no wonder my gunner isn't winning.
Gambler is weird. Just when I thought it was a useless class, it would suddenly dominate one of the better job classes and make it look easy. I'm trying to figure out the best way to build them. Right now I'm focused very heavily on HP, so that they can stick around in a fight long enough for some good luck to pass their way. That might not be the best approach though. It's also possible that Gamblers are better as support to other job classes, as I notice that they have a couple of abilities that carry on until the end of battle. So maybe they aren't the best straight-class job available.
Summoners might need a little tweaking. I think you gave them too many drawbacks, as they seem to be almost as luck-based as Gamblers half the time. Occasionally a summoner would pull out a miracle win, but they were overall one of the weaker classes I tried. More experimenting is necessary.
And I tell you what! *spits into spittoon* Dem dere Dragoons are mage-rapers. If you're thinking to yourself, "man, I think I'm going to have a straight mage and just rock the magic damage" then a dragoon is going to do obscene things to your body. Dragoons don't perform as well as I thought they would when the balance is shifted the other way though. A heavily physical fighter with low magic isn't as affected by a dragoon as I thought they would be. Maybe it's just that physical fighters have more HP in general.
Saboteurs might be my favorite class so far though. It's funny when the status effects start piling up and the Sab just keeps adding more and more to their countdowns. When a Sab gets on a roll, they can take down pretty much any job class.
And yeah, I thought that my first name bias might be a statistical anomaly. But I felt it was worth noting, since the person listed first won 60% of the time in my games. After 239 games, that's a fairly noticeable trend. But yeah, it's probably just the way I arranged the battles. I'm sure if I flipped things around, I can get that closer to 50%, like it should be.
|
|
|
Post by countlieberkuhn on Aug 8, 2011 5:52:45 GMT -8
Yeah, some classes like the gambler are more support-based, although they CAN wreck you, especially if they get 3 identical slot reels. It's a small chance, but they do absurdly good things if it happens. I couldn't be bothered to list the slot outcomes though, as there's so many. The main thing about gamblers is that they operate using a combination of strength, magic, and non-stat based effects, meaning that they can make a pretty nice third class for anyone. Of course, the idea of a gambler is all about luck, so its good that they are erratic. Enchanted Ammo deals 100% magic damage on top of the regular attack 20% of the time. Potshot is based on enemy HP, and trigger happy deals 18 damage * the number of bullets fired, which means that only attack is based on strength (and that can be enchanted too). As for summoners, I did kinda rush them before I had to go, so they're the ones most-likely to be imbalanced. I'm glad that you like saboteurs though, especially as you suggested them I've never tried 3 sabs, but I can imagine it can be pretty damn scary. I balance most things around the idea that you can only have one of each paradigm on a character though. 3 Thieves could potentially get very strong very quickly. The only shame is that dragoons aren't doing as well against strength-heavy fighters as magic-heavy fighters. The amount of damage dealt to a physical spammer should be the same, but as you say, physical people get more HP in general. They should probably still win though, and to be fair, magic-heavy users are capable of much more damage than strength-heavy people on the whole. Black Mages seem to perform very well, and a high-powered flare or ultimate rapes most things. As long as the 3 dragoon person is losing to more well-balanced classes like ninjas though, its ok.
|
|
|
Post by Fleck on Aug 8, 2011 9:41:16 GMT -8
For gunners, I figured that they'd be using their regular physical attack the most often, so that's where I put my focus. I'm gonna do more testing, but with a more magically-inclined gunner than a physically, and see if that makes the class better.
Another thing about dragoons: I just noticed that the disparity between strength and magic is almost always wider for mages. If you want to overload on strength, the best option is Assassin, which gives 14 strength and 5 magic. But if you want to overload on magic, the best option is mage, which gives 1 strength and 19 magic. The widest disparity physically is for the monk option, but that's only 8 to 1, versus the mage's 19 to 1. I think that's probably the biggest reasons dragoons don't do as well against physical fighters.
|
|
|
Post by countlieberkuhn on Aug 8, 2011 11:03:28 GMT -8
That's a good point. The question is, does it feel balanced? Is the tradeoff for getting epic magic power worth the risk of lower HP and a particularly high vulnerability to dragoons?
|
|
|
Post by Fleck on Aug 8, 2011 17:20:27 GMT -8
MORE ANALYSIS:
This time around, I messed with stats, to see what arrangement of stats was generally best for each class. For the most part, I was right with my initial guesses, but there were a couple of surprises.
Warrior: Warriors do well with a balanced approach or by overloading HP or STR. A warrior with maxed out strength can do some terrific damage, and the warriors automatic damage reduction allows them to stay relatively safe with lower HP. The high HP plus high defense option results in a lot of long battles, but can do fairly well, but is probably the least effective use of a warrior. Overall, they tend to do best with a balanced amount of HP and STR. Their passive ability to reduce damage might be a little unfair, but I need to test further to prove that.
Monk: High HP is the only way to go, really. More strength just reduces the effectiveness of the monk's passive regen skill and makes them more vulnerable to their own suplex attack. Monks are best when they outlast their opponents, not when they overpower them. Overall, they're one of the better classes, albeit one of the duller ones.
Ninja: Probably the most versatile class available, ninjas do fairly well no matter how you stack them. High HP will make them damn hard to kill, since they dodge so many attacks. High strength can be devastating against enemies with lower HP. High magic makes the Mirror of Equity attack into a thing to be feared. And balanced combines all those traits. Because of their absurdly high evasion, this class runs dangerously close to being overpowered, especially when you straight-class them.
Gunner: I'm still not a fan of this class. I experimented with about every single arrangement, and it always comes down to how many times they can get Enchanted Bullet. No matter how you arrange their stats, if you don't get a few lucky enchanted shots, you probably aren't going to win. Maxing out their Magic or staying balanced are probably the two best options, although both have heavy drawbacks. Also, the Backfire penalty is a little unfair. This job is already held back by low damaging attacks and little defense, adding a chance that they will damage themselves as well seems to be stacking the deck against them.
Black Mage: High magic is really the only way to go. Kill your enemy before they kill you. You can balance out their HP some more so that they are more likely to survive a lucky shot, but that really negates the best trait of BMs: their damage output potential. If you're going to cut into a BM's magic, you might as well not even use one.
White Mage: I initially designed them a lot like Black Mages, with high magic and low everything else, but a balance between HP and Mag makes this class a threat to every single other job available. Higher HP increases the effectiveness of the regen skill and keeps the white mage out of danger, while also making curing more effective (by having more HP to cure). A WM with balanced HP and magic is quite difficult to stop. Overall, White Mages might be the singularly best straight-class option available. This is probably the only class that I would suggest might need to be toned-down.
Dark Knight: Since Dark Knights use a lot of HP, Str, and Mag, I initially built my DK to be very balanced all around. However, I just discovered that a Dark Knight with high HP is devastatingly powerful. Even with low strength and magic, the extra HP balances out the equation to still dish out a lot of hurt. Using this method, DK's will usually do more damage to themselves than they will to their enemies, so if your DK is up against a monk or some other enemy with a lot of HP, then it's likely to lose. But opponents with low to moderate HP can't stand this onslaught for long. Alternatively, a DK with low HP and outrageous strength can be amusing, but results in a LOT of draws, as it usually kills itself after two or three turns, often taking its enemy with it.
Gambler: I've found that it really doesn't matter what you do with a gambler, they're luck-based all the way. I guess that makes them a decent support class if you don't know what to put in your third slot, but otherwise I haven't really seen much of an upside to gamblers. If you're going to go with a straight gambler job class, then your best bet it so max out the HP so that the gambler gets as many turns as possible and hopefully gets something good. Maxing out stats other than HP just makes this class more luck-based, as now they have a narrower time limit to get lucky before they get slaughtered.
Thief: These guys can go a couple of different ways and still be effective. Because of the ever-accumulating Thievery stat, Thieves only get more powerful as the fight goes on. Getting High HP makes the fight go even longer, thus making the thief more and more likely to win. However, thieves are kind of weak until they build up this power, so high strength is also an option, to kind of bridge the gap between the start of the battle and when the thievery stat starts kicking in. But thieves are probably best when given maxed out HP.
Dragoon: Kind of a weird class, because they depend so much on how your opponent is built. Like I've mentioned before, Dragoons utterly dominate mages, but struggle a bit with physical fighters. Dragoons can be improved by increasing their strength... but this makes them more vulnerable to other Dragoons. A balanced approach is the safest method, but the least fun. I'll need to test these guys some more before I decide if they're balanced or not.
Berserker: Unquestionably, they are best when you max out their HP. Because so many of their best abilities don't kick in until they are at low health, you want "low health" to be as high as possible. Berserkers do best against consistent job classes, meaning ones that don't often get lucky shots. Classes that generally do moderate damage are the most susceptible to Berserkers, as they will have trouble finishing off B's before they themselves are torn apart. Classes that occasionally score high damage, like black mages, can usually handle themselves pretty well against Berserkers, so long as they get lucky at the right stage of the fight.
Saboteur: This class probably requires a balanced approach more than any other mage class (except perhaps white mages). A sab's strength doesn't come from how hard they hit, but how many statuses they inflict. So it's in your best interest to keep the sab alive as long as possible so that they can rack up a ton of statuses against their enemy. Maxing out their magic is probably just going to get them killed, unless you get lucky with a Death spell or the right status at the right time. Probably best suited as a support class, but they do quite well as a straight job class.
Summoner: Still needs to be tweaked. They take too long to build momentum and take too much damage in the interim. I tried about every possible combination of summoner stats (the best being a balanced approach. 20 Red Mage packages was probably the best option) and they still consistently lost to most job classes. This is probably going to be the trickiest class to balance, as they can easily become overpowered.
|
|
|
Post by countlieberkuhn on Aug 8, 2011 18:24:00 GMT -8
Wow, that's a lot of testing and info, I feel quite honoured you spent this much time on it As it's looking from your info, it seems like most of the classes are in a good place. White Mages can be toned down a little, but I'm not sure in what regard just yet. Summoners could probably be improved by making them take a little less bonus damage during the summoning phase, right now it's quite high. As for dragoons, they're intended to be a reactive class, but maybe I could tone down the damage on a couple of moves slightly, and instead add a component that's based on max HP into the equation. That should hopefully make them a little less overpowered against straight mage classes due to their lower HP, and a bit stronger against straight strength classes. It's good to know the best way to build things though, especially when it comes time for people to make characters when the time comes. Hopefully when everyone's back and mallboro picks up again. I won't be able to edit the program until sometime next week though. Right now I'm visiting family, but on friday I'm off to Spain for a few days, so I'll only be spending a small amount of time at home in the interim. Question: If Squall participated (he probably wouldn't), what would his 3 paradigms be, assuming they all had to be different?
|
|
|
Post by Fleck on Aug 8, 2011 20:29:25 GMT -8
Eh, it's fun to play with, so it's hardly a chore for me to test it. Plus I'm getting a bit of an advantage over everyone else with all this testing, so WOO CHEATING.
Most of the classes are pretty well balanced. Despite all my tweaking, I have yet to make an Unbeatable Build (or even a Mostly Unbeatable Build), so that's definitely good news.
And I wouldn't want you to edit the program yet anyway. I'm still experimenting with straight-job classes and my results are still preliminary at this point.
If you do a tournament or something, I would certainly participate. As for what Squall would do... well, Warrior-Black Mage-Summoner would certainly fit his in-game character, but that would probably result in a horrible mess. I have a couple ideas for what might be some rocking combinations, but I'll test those later.
|
|