Custom Hero Arena! (From WC3)

Level 7
Joined
May 15, 2009
Messages
168
Custom Hero Arena! (From WC3) Video & SS

======================================================
Custom Hero Arena
======================================================
This is a classic hero-arena style map; with a lot of advanced options. That is to say players are spread into 2 teams of heroes. To win the game, players must kill each other. It is not strictly a PvP style game though. Players must kill monsters who spawn in the arena to gain minerals and gain experience. If you try to focus entirely on PvP, you will get outleveled, which means you will have less stats, lower leveled abilities, and worse items. It is important to kill creeps. Killing players is how you win, and also offers bonus minerals.

What sets this game apart from normal hero-arena style maps (which do not exist on SC2) is the level of customization. In this, you select a hero skin, then you select 4 abilities and "innate" attributes. After that, you distribute stats however you want. The video below explains it all quite clearly.

Advanced stat systems:

[YOUTUBE]http://www.youtube.com/watch?v=XSi_QyJ1PnI
[/YOUTUBE]

When you hover over a stat, it tells you what the stat does:
Strength: Each point increases attack damage by 1.
Spell Power: Each point increases spell damage by 2.
Agility: Each point increases attack speed by 1%.
Vitality: Each point increases life by 50.
Intelligence: Each point increases energy by 15.
Constitution: Each point increases life regeneration by .3 and energy regeneration by .2.
Armor: Each point reduces all physical damage by 1 and spell damage by 2.
Speed: Each point increases movement speed by .03.


Creep/Terrain layout:
Terrain003.jpg


There are 6 major creep spawns on the map; 1 above each base, 1 below each base, and 2 in the center of the map. there are also 10 minor spawns, where 1 creep spawns at a time and wanders.

Creeps will be levelable. that is, there will only be 10 or so "base" creeps who will spawn at higher levels as the average player level increases. Creeps will be designed to have specific strengths and weaknesses, so that a single build cannot kill all creeps easily.
Examples:
Creep1: High life, low armor. weak to faster attacking heroes, strong against spell casters.
Creep2: Low life, very high armor. Strong against fast attackers, weak to high damage attacks and spells.
Creep3: Fast attacks, long range. Strong against ranged glass cannons with low armor.
Creep4: Slow attacks, high damage. weak to regen builds, strong against armored builds.


ITEMS:
Items in this game are based on ypgrades. You start with the base level of all items, and you cannot sell items or own more than one of a certain type of item. You can upgrade items through a dialog system, as shown in the above video. After the basic level of an item, you can pick 1 of 3 paths for each item to upgrade to. Once you make a decision, there is no turning back. The lower the rank of item, the less each stat point costs. While upgrading you need to choose between what might be a quick payoff but a less powerful later game item. If you watch the video and pause it frequently, you can see all of the current items.


PVP:
This game is pvp based, in that to achieve victory, you need to reach a set kill count.
Every 4-1/2 minutes there is an arena where an equal number of players from both teams are randomly selected to fight to the death. The winning team is rewarded a mild sum of minerals, and the losing team gets a bit fewer minerals. Player kills outside of the arena grant a nice bonus, they reduce the opposing teams points kills by 1, but cannot drop below zero.

Game Progression:
Early game goes pretty quickly. Levels 1-10 are fast. The difference in players level throughout the game doesn't have too potent of an affect on a player's ability to perform. The stats are a system of counters, if you pick stats to counter your opponent, you can be 4-5 levels behind and still kill them. This becomes a constant weight value throughout the game, as you try to get the best possible farming output while still maintaining strength in PvP. Stats from items do not have a strong impact on your build early on. Late game, half of your stats may come from items, making your early game item choices important.

Leveling to 30 (the max level) is intended to take about 40 minutes; and a complete game is intended to last 60-70 minutes. It would be nearly impossible for a player to max out all of his/her items.


What I still need to add:
A lot of skills. As long as this game is being played, skills will be added. I plan to have ~30 when the public beta is released.
5 minibosses, they will give a secondary resource for end-game item upgrades, and unlock the final boss area.
The final boss.
Gloves and Rings (items)
Trinkets (unique, expensive items that greatly alter a build)
Dialog box that shows the score
Achievement system, rankings
 
Last edited:

Dr Super Good

Spell Reviewer
Level 59
Joined
Jan 18, 2005
Messages
26,664
Do be aware that such customizability of stats can lead to abuse where a player only skills 1 stat to stupidly high levels.

An example could be spell power, where a hero might only skill that at the cost of durability to become a rigged glass cannon, where he dies in a few hits but will OHKO most enemies. Armor is another good example since armor in SC2 reduces attack damage by a fixed amount instead of how in WC3 it reduced by a % so a player could potentially become immune to all physical attacks.

Even if you stop the inital stats going all into 1 attribute, the ones gained by level may be enough.

Terrain looks quite bland and flat for an arena. Consider adding spectators or a more natural feel to the place.
 
Level 7
Joined
May 15, 2009
Messages
168
That terrain is just a "paint job" layout so I know where everything will be going later. Terrain is usually the last thing I do; because I always end up adding something that needs a place, and I dont have a place for it, and a move everything around, ect. I am not a great terrainer, but, I am better than my pictures above ^_^.

The stats thing; I do see your point, but at the same time I am hoping to make it a non-issue; I am attempting to make as many counters to things as possible.

Why maxing a single stat is bad:

Maxing Damage:
If you max out any single stat (damage, spell damage) then you are entirely too weak to attacks. Keep in mind, your level is going to be based in large part to how many creeps you can kill; while winning the game is based in how many players you can kill. If you go alllll spell damage, a player can simply pump a level of vitality (300 HP, vs 30 spell damage) and survive that initial spell damage blast, then 1 shot the guy who is going pure spell damage. They could also go high constitution, so that while the "nuke" is on cooldown, they regenerate a lot of life.

Going high armor/life/constitution so you never take damage:
Sounds good on paper; but just the opposite of the above, you will never deal damage. If you cannot kill creeps, then you cannot level. Without leveling you dont get more stats and items.

"Balanced" build types:
If you wanted to be a glass cannon that was actually viable to play, you would need spell damage, intelligence, constitution, a hint of armor, and some good movespeed. Spell damage obviously makes your spells hit harder. You need intelligence so that you can cast more than 1 spell before needing to run back to base and heal. constitution will regenerate your life and mana while in the field, to minimalize a need to return to base. You need enough armor that ranged creeps arent going to kill you before you kill the spawn. Movement speed is very important to dodging creeps and players who try to kill you; and also, for killing other players in the field.

"Alright, so, casters seem pretty bad, if I need to do all that work just to use a spell, I think going melee meatbag is a great idea"
^^This is how many people like to play game. In this, it is just as difficult to place your attributes properly.
You obviously need a lot of strength to deal damage. Agility is a 1% increase to your DPS with every point. After you have 100 strength, agility increases your DPS by more than strength does. you need vitality and armor to survive all of those hits; and you really need move speed if you want to get the upper hand on any ranged enemy players. And you need these stats in pretty decent ratios, that allow you to creep to the best of your ability without dying too fast.

How the stats counter each other:
Strength: Countered by Armor+Constitution
Spell Damage: Countered by Vitality
Agility: Countered by Armor
Vitality: Countered by Strength
Intelligence: Every point needed here, is a point not put elsewhere.
Constitution: Countered by Agility
Armor: Countered by Spell Damage, high Strength
Speed: Counters Constitution, works with constitution, counters ranged, counters speed, works with ranged; but every point invested here, is one not invested elsewhere.

^^Hope that clears it up for you. I am sure there will be imbalances in the early editions of the map. The above stats in WC3 "Custom Hero Arena" all have similar roles to what they do in this. The major difference is that in this, you will need to pay attention to how your opponents are using their stats, and alter your plan to best counter them.
 
Level 7
Joined
May 15, 2009
Messages
168
Just a quick update, to show I am still here and working hard. Small gains despite the ~200 hours I have spent on the editor already with this map. It is a learning process...

The custom skill system has finally been made, but I need to make a handful of skills for each category before it can really be tested actively; for potential flaws. Right now, dynamic tooltips are beating me down and spitting on me. Who knew it would be so difficult to update the information on a tooltip as an ability levels. After playing a handful of games though, I realized a lot of games do not use dynamic tooltips; they simply have like "deals (40*ability level) in damage over 3 seconds. I want to stear away from that... despite how many dozens of hours it is going to take to make accurate tooltips...

Within a week, if I don't have a video up of the picking system, someone can slap me.
 
Level 7
Joined
May 15, 2009
Messages
168
Good suggestion. This is something I have thought about, and wanted to add. From what I read though, it is very easy for people to change the values of their information banks. to my knowledge there is no way to encode them in a way that would make them tamperproof. That is the only reason I don't plan to add them. If people can crack into their save and set their starting stats to 5000, it would be problematic.

I heard a few people are attempting to make an encoded system, like the WC3 save systems. When someone else creates one, I plan to borrow it and allow people to save a few builds with it.

the only argument, which isn't a terrible one, is that most people would always think of something minor to tweak in their build, so a lot of people wouldn't take advantage of the feature.
 

Dr Super Good

Spell Reviewer
Level 59
Joined
Jan 18, 2005
Messages
26,664
Nice, although you should avoid placing buttons over the portrait or if you must then consider moving the portrait.

Also it would be nice if you could change the tileset from a dirt or sand lookalike. I find grass or other colours much more cheirful than sand due to its overuse as it is the default tile all maps are set to.
 
Level 7
Joined
May 15, 2009
Messages
168
In a fullscreen game, the dialogs line up better. Something about running the game in a window makes the dimensions of dialogs scale in an awkward way.

The tileset is still completely up in the air. I have some lines and circles to represent regions later, but I have spent maybe 10 minutes on the terrain. The theme I plan to go with is a "Concrete Jungle" Along the lines of a ruined city that has become overgrown by a jungle. there will most likely be no sand at all.

Also thanks for the idea of doing items through a dialog. Despite the workload of it, I think is it worth it for how player friendly the system turned out.
 
Level 8
Joined
Jul 20, 2004
Messages
2,760
Looks amazing. I'm just curious how you made the 3d "unit preview". Is that like the actual unit rotating on the map placed at the right coordinates, or some sort of special control? (I'm not sure I made the movie control work, though it's possible last time I checked it I still had version 1.0)
 
Level 7
Joined
May 15, 2009
Messages
168
lol, I really wish I could claim to make something fancy.

The unit buttons on the selection dialog are all set by a record; which i set in the editor for each unit. One of the variables I set in it is a game link -model. Simply enough, based on the hero picked, I set the model for that player to whatever the model of the hero picked is. If you repick your hero, you will get a new unit rotating.

It is shown through a portrait - create model catwalk. You can set various "acts" for the model there, along with a camera angle to view the model, and the spin rate.

^^It is one of the many, many, added on features with the GaX3 mod. http://www.sc2mapster.com/assets/gax3-starcraft-ii-editor-enchant/

I have never seen this used before, and stumbled upon it by accident. If you wanted to be fancy, you could create backgrounds for it, and make it play a string of animation. When I get to the "graphical makeup" phase, I might make all of the actors "playfight" for a while, as you pick skills.
 
Level 7
Joined
May 15, 2009
Messages
168
11 Day bump!

Been grinding out all of the items in the object editor. Making 200+ items is... slow.. painful.. very painful. Well, I guess triggering all of them to work with my dialog system of upgrades is what is really killer. If I get some alone time this weekend, I should finish up all of the items. I need them finished before I can make abilities, creeps, or bosses. The game needs to be balanced around stat progression so it is never too difficult or too easy at a given time; and so that abilities mana cost and affects are still more important than stats at most stages of the game.

In terms of game progression, I have decided (and this may change before any form of an/a alpha/ beta is out) that stats will outlast level. By this I mean, the level cap will not be too difficult to hit. Say, 30 minutes of play time to reach the max level; but games last for 50 minutes.

Items cannot be sold/swapped out once you have decided on an item path. For those who haven't read above, every item, of 8 starting items, will have 3 upgrade paths. Once you start down an upgrade path, there is no way to change the path of that item. As the game goes on, you pay more minerals per stat point on items. To push hard early game, it is in your best interest to upgrade all of your items evenly; for maximum stats. End game though, you are stuck in those decisions, which may not be the best later game choices. If I get all intelligence and spell damage items, but neglect armor and life, then late game when I can no longer level, I will be stuck with an armorless build. An opponent who still has a lot of item choices, can gear to counter me; despite the fact that I may have been winning a lot early, there is a good chance I wont win late game.
 
Level 7
Joined
May 15, 2009
Messages
168
Bump for a fully playable alpha map.

I still haven't added the last 2 item types, there is no victory condition, and some small areas of terrain aren't done. Some other things, which wont change gameplay are also undone.

the good news is, you can play games without encountering any of those moments where it is "ohh, that isn't made yet, you can look at it, but it doesn't do anything."
My plan is to advance to a beta stage by the end of the month. By that I mean, have everything I plan to add, added, and focus more on bugs and balancing.

Usually people in "looking for custom games" channel are willing to play. The map name is CHA-Alpha I am Flurgenburg #920 on NA server. If anyone would like to test it out. I am always looking for ways to improve first impressions.
 
Level 7
Joined
May 15, 2009
Messages
168
Bumpity Bump for BANKS!

Finally added a banking system. Along with it has come an achievements list.

There are many achievements, and many more to come; so I am not gonna bore people who dont care with the full list; but, in general"

Creep kills
Boss kills
Ragnarok kills
player kills
kills per game
deaths
game leaves
game wins
total games
experience
experience per game
level (based on experience)

these things at various levels are saved and achievements are based on them. Just something fun for a player to look forward to, as they score their X number kill.

It physically pains me to grind items in the data editor; so I keep putting off my last item set... which keeps putting off a switch to beta. ahhhhhhh; data editor.. I curse you!
 
Level 7
Joined
May 15, 2009
Messages
168
everything on my to-do list has been finished, with exception to trinkets.; though I have decided on 1 last major change before I take the map out of alpha testing; as it is... quite large.

Rather than all abilities being available to all heroes, there will be 3 classes. Auto-Attackers, Spell Casters, and Hybrids. Each of the 3 classes will have their own list of abilities to pick from; with some abilities being available to multiple classes.

The reasoning is that, the way abilities are balanced so as not to be overpowered when used by the right build, is that they are underpowered when used by the wrong build. Obvious statement is obvious. If you are an auto-attacker, picking a high-energy spell damage skill is not going to get you very far; since it is balanced more for people who are building heroes to cast high energy spells. There are many skills an auto-attacker would never pick; and many a spell caster would never pick.

To allow for more "viable" customizations between build classes, this seems like the best solution.

The only thing on my to-do list now, is add 25 more abilities and update the selection system. Give me... 2 weeks!
 
Level 7
Joined
May 15, 2009
Messages
168
I dont update this frequently enough; and I am sure I will keep up that pattern in the future.

More skills have been added, a good handful of them. The game has undergone a "solid" balancing stage where nothing is bugged, haxxed, op, worthless, ect. Me and a lot of the more active players pick "random" every time, and end up with usable builds; occasionally terrible ones; but, sometimes better than you could pick; so it evens out.

If anyone pays much attention to SC2 triggering; I had triggered this map like anyone from WC3 would. I found out recently, that SC2 has a a much better method of triggering using things call "Action definition" which runs triggers; err;actions, across multiple threads; opposed to just a single one. Long story short "games have better response time". I have been remaking the game using action definitions; and have only seen minor improvement; but I guess it would scale much nicer on a weaker system.

Anywho; anyone interested in playing this, we have a mildly active group that hangs around channel Clan CHA on the NA server. The map floats between page 4 and page 10. I have recently talked to a data editor who might be able to take some time and spice up the graphics on abilities, and help me pump out more skills. I will get trinkets and "class items" added when I finish converting all of the triggers over.
 

Dr Super Good

Spell Reviewer
Level 59
Joined
Jan 18, 2005
Messages
26,664
If anything starting new threads decreases performance because of the extra overhead of creating the thread. Additionally they are not true threads in that only one will be able to execute at any given time despite most computer systems having parallel thread execution capabilities. You are also limited to a certain number of threads at any given time and exceeding that number will cause errors (which will likely break your map).

What new threads are useful for is that they have a separate op-limit and thus are very useful for iteration intensive tasks. For most actions you should be running them in the same thread (as function calls).
 
Level 7
Joined
May 15, 2009
Messages
168
Over at SC2 mapster I have been getting all sorts of advice, and reading tutorials and such on action definitions. The 3-4 people who have offered me good help have all insisted that action definitions increase performance. The reason I have gotten from... all of them is that say you have 5 triggers running at once for some reason; then they can still only run 1 action at a time; causing the actions at the bottom of the list to lag. With separate threads, those actions can run side by side.

Again, after taking much hassle for my "Wc3 style of editing" this is what many people told me. Also, I found on my searches, 1019 is the max number of threads; and it doesnt crash the map; it just gives an error and says "naw, we aint gonna do none more, thanks!"

The only support I have for what I was given was I broke all of my initialization triggers up (I had separate triggers because they reached the text limit per trigger) and ran through through different action definitions on init; and it decreased map loading time by about 25%.

I am actually very interested in the topic, and can find almost no true information on it. All of my information came from people on sc2mapper who firmly believe that no trigger should have more than 1 action; that 1 action should run an action definition that runs a switch action to check conditions and such. Do you know where any of the true data for this might be posted, other than just word of mouth from various opinions?
 

Dr Super Good

Spell Reviewer
Level 59
Joined
Jan 18, 2005
Messages
26,664
The reason I have gotten from... all of them is that say you have 5 triggers running at once for some reason; then they can still only run 1 action at a time; causing the actions at the bottom of the list to lag. With separate threads, those actions can run side by side.
This is not possible as that brings in horrible synchronization problems which plague professional programmers such as atomic operations and resource locks. There is no way Blizzard could have possibly expected their user's to cope with such complexity so only 1 thread will execute at any given time (you can assume each thread is atomic in nature and executes immediately).

There is no way it will influence lag at all as that is network related (unless you mean poor performance which is different from lag).

When using wait calls in the middle of a thread it will reschedule the thread to resume at a later time. Creating new threads means you can have many such waits scheduled in parallel but be aware of the thread limit. An example is you could have a thread per missile but if there are enough missiles you will encounter the limit (try and bunch many missiles per thread).

Daelin also agrees that only 1 thread can execute at any given time. For professional software developers you can imagine that the scheduler only has one core and that it does not pre-empt any active threads unless they reach a certain instruction threshold (the oplimit).

Again, after taking much hassle for my "Wc3 style of editing" this is what many people told me. Also, I found on my searches, 1019 is the max number of threads; and it doesnt crash the map; it just gives an error and says "naw, we aint gonna do none more, thanks!"
I never mentioned anything about it crashing the game. It can however break your map as critical threads that are needed for the game to progress might not be able to be created and thus a session will become un-playable as you cannot progress. An example could be a hero revival thread failing to be created and so a player never has his hero revive when game play is designed that it does revive.

The only support I have for what I was given was I broke all of my initialization triggers up (I had separate triggers because they reached the text limit per trigger) and ran through through different action definitions on init; and it decreased map loading time by about 25%.
You did make sure your file caches were cleared between these tests right? Something as simple as starting a map for a second session in a row can decrease load times by 25% as all the necessary files are already in memory. Unless you have an insane initialization script that uses a lot of time consuming operations (such as those that are throwing a lot of errors) it should load the script almost instantly.

I am actually very interested in the topic, and can find almost no true information on it. All of my information came from people on sc2mapper who firmly believe that no trigger should have more than 1 action; that 1 action should run an action definition that runs a switch action to check conditions and such. Do you know where any of the true data for this might be posted, other than just word of mouth from various opinions?
Look at the underlying Galaxy script file the GUI compiles to. All action definitions that are set to make new threads are is another trigger. When you run the action definition you tell the trigger to execute. Thus you are making a trigger run some code that fires a trigger that runs your code. Worse, it generates a function to start the trigger to run your code so it should be even slower.
 
Top