Hello hiveworkshop, I am mafe and in this guide, I would like to gather and elaborate on some topics around melee maps that come up of on a regular basis for new melee maps uploaded on the hive. Authors note: This will be a long thread, and it will likely never be finished. I will try to add to it on semi-regular basis over the next weeks/months and also improve layout. What is the purpose of this tutorial? There are several. Briefly, I want to explain why certain features of 1v1 melee maps are good idea to have. At least if you are aiming for balance. I try to give a lot of feedback to melee mappers, but I think that quite often, the same "mistakes" are made. However, explaining the reason for some of these errors is often rather nontrivial and requires diving into some deeper aspects of gameplay. So my goal is gather some of these points and elaborate on them in more detail than possible in the thread of an idividual map. What this tutorial is not: Strictly speaking, it is not a tutorial. I wont tell you what to do, I want to increase your knowledge about melee mapping. And then let you draw your own conclusions. Also, if you are not aiming for a highlevel competitively balanced map, most of the topics of this guide can safely be ignored. Also, maps that are not for 1v1 allow for more freedom. For now, I'm going with the style of asking a hypothetical question and then answering it. I will also update this introduction over time. Best regard, mafe So let's begin. What does a map need to be suited for competitive play? These are basically the ideas from which all of the other points follow. I believe a good melee map should at least achieve 3 different, but not unrelated, goals: It should be balanced, it should reward skillful players, and it should be fun (both to play and watch). Balance should go without saying for any wc3 fan. The challenging part is to have all 6 non-mirror matchups in mind. Be prepared however, that much of the feedback will be that any new map is a "not-my-race" map. Skillfullness means that ideally any skill that can lead to players winning a game wc3 should be relevant on your map. Of course, you can make some aspects more important. For example, even if it might be hard to swallow for ud and ne players, make sure expansions and tanks are still possible (otherwise you might also run into balance problems, if you nerf one weapon for humans without giving anything to them in return). Also, try to not introduce any larger-than-usual luck factors (itemdrops!) which can be a deciding factor instead of skill. Also, if players can apply usual tricks like AoW-creeping, walling their base etc, they will immediately be more positive towards the map than if they would have to change their entire playstyle because the map demands it. Fun is what most players play or watch the game for. If your map feels annoying to play, why should anyone play it? If it produces similar games over and over again, or if boring playstyles are more effective than usual, why would a tournament or league accept the map into the mappool? It is of course hard to define what consitutes fun, but as wc3 has provided entertainment for so many years, you are probably once again on the safe side if "normal" play works. Notice however that standard playstyle working doesnt mean that there cant be anything new. The key point is that new things should ideally come in addition to and not instead of the map having everything that is necessary for a standard game of wc3. Why does every competitive map have a tavern in the middle? If you look at the blizzard ladder or GCS maps, almost all of them have a tavern in the middle. Why shouldnt taverns be on different places of the map, like other neutral buildings? Well, there are three different major situations when a tavern is relevant ingame, and I'll look at all of them. It turns out, having the tavern in a central location of the map leads to more interesting gameplay: Players go for a neutral first hero: Tavern heroes can be bought later than a standard (race-unique) first hero can be produced. Assuming both players go creeping, a player with a standard hero will therefore always be a little ahead in the earlygame. So why get neutral hero if this means you will be behind? Well, you need get some compensation, and that's where the tavern location comes into play. If the neutral hero spawns in the middle of the map, it has more options than a hero that spawns in base. You can kill creeps in the middle, which would would be contested, and have your safe next-to-base creepspots later on, while the enemy already kill theirs (Of course, the enemy can simply run towards the mid and do the same, but then he loses some of his timing advantage). Rushing the enemy is also possible and you force him to consider it. You could even adapt your hero choice to what you've seen the enemy do. However, if you had to buy the hero from a tavern near your own base, you miss out on some of these advantages, and then neutral heroes become less attractive. Players go for a neutral second hero: Much of what I said about neutral first heroes still applies here. However, neutral second heroes are more common than first heroes, so if there is only one tavern, players will often meet there at about the same time to buy a hero, and fights take place. This is creates and interesting, exciting and entertaining dynamic, which is important for highlevel gameplay if it streamed and commentated. Also, the player to first buy his hero has a brief advantage, which in turn makes rewards good build orders and execution, and gives players an incentive to fight for mapcontrol in the early game, so that buying a 2nd hero from tavern isnt dangerous (rewarding skill is something a map should encourage, I will write about this later separately). Many neutral hero strategies are build around establishing mapcontrol, and if your hero is right in the middle of the map, it helps a lot. Also, you can try to cancel t2 buildings which usually isnt possible with a second hero from the altar. On the other hand, if your enemy could buy a hero very close to your base, these building-cancel pushes might be too strong. Reviving dead heroes: Heroes are a key feature of wc3 and losing a hero should put you behind. But should you lose a game where you were significantly ahead because of one lapse in concentration, or because the blademaster got 3 crits in a row on your hexed hero? I tend to say no. So the tavern revive is some sort of emergency option to here. Again, a safe tavern revive requires some sort of mapcontrol, so if you are behind already or even lost the hero in fight in your own base, the tavern revive is more of risk, punishing you as it "should" be. On the other hand, hero focus can also be a tool for player to drive the enemy away from his base and turn a losing fight into a win. If the enemy has a tavern next to your base, this is suddenly no longer as good as on option, as the enemy can instantly revive his hero during the fight, whereas otherwise he would have to move to the middle of the map, giving you time to breathe. Of course, there are probably also reasons why taverns could work in other parts of the map, but I hope that you now understand why one tavern in the middle of the map is probably gives more interesting and balanced gameplay. Why dont we see more different item drops from creeps? In this paragraph, I'll point out why the nonstandard items are usually not seen. There are of course many other aspects of items (what item is appropriate for which spot, avaoiding the pitfalls of tome of experience and wand of lightning shield...), but these wont be addressed here, but maybe later. First of all, the following items can normally be used without causing unacceptable problems: Lvl1-6 permanent items, lvl 1 powerups, lvl 3 charged items. Other acceptable items, which can used, but with care, are lvl 2 powerups, lvl 2+4 charged items, runes of healing/mana/watcher. Everything else should be a no-go except when you have thought about it very well. Of course, this there are many more items in the editor and it is tempting to try to make them use. However most of them have some issues and while I cant make a complete of all of them, I'll try to point out general issues and address some individually which I've seen used before. General issues: Many of these items are not balanced for gold cost. Remember you can not only use items, but sell them. If you check in the object editor, you will see that many of the exotic items from the campaigns will seel for some ridiculous 4-digit amounts of gold. To a lesser degree, the selling reward for some of the smaller, innocent-looking items is still questionable. Of course, you can edit the gold value if you want to go down that route, but then you would still have decide how "good" such an item would be compared to standard items. Specific issues: Gold coins & Bundles of lumber: Actually, I think it should be possible to make these units work in competitive maps, especially given that get a gold coin is essentially no different from getting and selling an item. Of course, if players know they can get lumber/gold from creeps, this could have impact on possible build orders (for example, teching faster with 1-2 less lumber peasants or just 2 ghuls), and the consequences are impossible to foresee. I havent checked yet, but if gold coins are impacted by upkeep, gold coins will get lose value later on (unlike items, whose primary use remains unaffacted), but this could also be seen as an interesting mechanic. My conclusion is that it could be worth trying these items again, but to use them sparingly and initially only as a small footnote in maps. Also, it would be difficult to rate them relative to other items, and if you have a 100% drop chance with no alternatives, you also change another unwritten rule of the game that item drops should rarely be predictable. Runes of rebirth: Another interesting item. The problem here is that the units you get will take supply you might rather want for other units and that you might accidentially screw your build order by being supply blocked at the wrong time, like when you want to buy your second hero; however if players know where they get this rune, they should be able to adapt their strategies. Of course, you once again not only introduce a new item, but also again give a 100% guranteed reward, unlike with other items. It is kinda unavoidable that some rebirthed creeps will be more useful for one races than for another (For example, undeads cant coil-heal a creep). With standard items, usually any race could a useful item at any spot. Now if you know what reward you will get, the entire situation is different and where normally an reward has an about equal chance to be useful for both players, now it mightbe 100% clear that one player would benefit slightly more from taking one creepspot. This puts players in a situation which we normally dont see. Of course, this if you can make it balanced, it would be awesome, but the ramifications are probably hard to predict. Again, a first (and boring...) step would be to have just 1-2 runes on the map and on relatively insignificant creeps, in particular not creeps with potentially fight-changin spells like pruge, ensnare or slow. Other runes: Mana, heal and replenishments runes again can probably made to work out ok. The change to usual gameplay would be that while after creeping you normally have some hurt units/lost some mana, the rune quickly gets you back to almost full strength. This means that a strategy which tries to creepjack or or force a creeping enemy to fight is now nerfed. This might lead to more passive, creeping heavy, and less entertaining games. Items that can be bought from shops: For example, why not have creeps drop a heal/mana potion? In this case, one potion has a higher (selling value), so if both players decide to sell the items, one is better of. For racial shops, they are a big part of (early) game balance and often needs to be seen in connection with the units and heroes for whcih they are usually available. If undeads, for example, could get a heal salve from a green camp and suddenly heal their hero, the entire complexity of the early game would change. As another example, orcs have the scorll of speed in theri shop, but they tend to not have enough grunts to go for a surround. But if a human with 4+footies gets a scroll of speed from creeps, then you can bet that good human players would be able to use the scroll to get a surround on a regular basis. And again, you either have to violate the not-100%-principle, or find several equal-valued items, which is hard enough. Why each player should have an expansion that is clearly closer to him than to the enemy, so that humans can reliably expand. The reasons here is one matchup: human vs undead. As of 1.28 (and the patchnotes of 1.29 wont make much of a change), this matchup is normally unwinnable for humans in a 1base vs 1base situation, assuming players of about equal skill. I cant explain all details here, so it has to be enough that there is a matchup that forces map to have an early game expansions option that can realistically be taken and defended. However, because of the other 5 matchups (and in particular those of hu vs orc/ne), expanding cannot be too easy either. So what does this mean? Assuming a sufficiently skilled player on the undead side, there will usually be death knight around with a skeleton rod by the point the human is trying to creep the expansion. At this point the human has to fight both the enemy hero and the creeps. Normally, this is a creepjack situation which you typically want to avoid in other circumstances, but here the human cannot. So the human army must be strong enough to have a chance, and this typically is ensured by bringing enough militia to the fight. This last word is the key here, as the miltia should still have enough time left before expiring to kill at least some of the creeps. This already limits the distance an expansion can have from the main bases. However, a good undead can also scout and see militia leave the human base. If the distance from both the undead base and the human base to the gold mine would be about equal, the undea player can decide to call his ghouls to the fight at the same time the militia leaves the base, and suddenly the fight to creep the goldmine looks very tough for the human. This isnt an option when the human has a shorter distance to the gold mine, as then the undead either has to call his ghouls earlier, and risks losing much more lumber-harvesting time when the hu does a different-than-expected timing, or the ghouls will only arrive after the human-vs-dk-and-creeps fight has been going on for a while. If we make things a bit more concrete, the following factors can be used to tinker with the difficulty of expanding: A.Distance from hu base. B.Distance from ud base. C.Strength of creeps defending the expansion (and strength of items). D.Adding more gold mines. E.Availabilty of neutral buildings. I've explained A&B, while C should be pretty obvious. Regarding D, if there are further goldmines available, a human can live with more units losses in the early game, as he can safely invest into towers to defend his mine later, and not risk running out of gold. E is hard to rate, but typically humans benefit from a shredder, a zeppelin for harass/savind hero, mercs or boots of travel much more than the undead does. A formula how to balance (pun intended) these factors, isnt easy. I would recommende to look at current maps and see how they to it. If I subjectively rate the factors from 1 (bad for expanding) to 5 (good for expanding), i get about the following scores for some maps (1.28 versions): Echo Isles: A -> 2; B -> 2; C -> 1 (weakest creeps out of all maps); D ->5 (no further goldmines avaibable); E -> 2 (Merc camp is there, but not that easy to creep). Total score: 12. Turtle Rock (cross position, going for expansion on the unoccupied starting position): A->2; B ->3; C -> 5 (red camp); D-> 1 (virtually infinite goldmines); E -> 4 (only shops). Total score: 15. Last Refuge: A->2; B-> 4; C -> 1; D -> 3; E-> 3 (all there, but realtively hard to creep). Total score: 13. Amazonia: A -> 1; B -> 4; C -> 3; D -> 4; (no further gold mines, but more gold in mine than usual); E->1 (very safe lab). Total score: 13. Terenas Stand (Expanding at the next-to base mines): A -> 1; B -> 2; C-> 4; D -> 3; E ->1. Total score: 11. I'm not sure which map I would rate as 5 for A. For B, 1 would be TM and 5 would be TR close position. Of course there are still some other factors which are hard to quantify however...... Nevertheless, these scores pretty much coincide with the winning percentages in the hu vs ud matchup as you can check them only liqupedia (ok, maybe a tinkered with the values here, but just a little bit). The bottom line here is that you should probably aim for a 12-13. Try to compare the expansion on your map and see how these factors look on your map relative to what you see on competitive maps. (Now that I see what I've written, it looks slightly different from the header of this paragraph. However, the distances are still a large factor in balancing maps, but in theory, it might be possible to rebalance unusual distance layouts by other factors). What about all those other mercenary camps? All competitive melee maps use only the Lordaeron Summer mercenary camp. Why not add some variety through some of the other merc camps? Igoring the fact that most maps are based on Lordaeron tileset and therefore Outland or Ashenvale mercs dont really fit the style, there is usually something very wrong (gameplay wise) with all the other merc camps. It really only depends on the mercs being available, so I'll give a quick overview about the different merc camps to check for each if it could work, why it doesnt, or if it is something in between. The general idea is that mercs can bring abilities that usually are not availbale to armies, which usually matters in particular in the early game: Until t2 there normally are no unit-based disables available, so if there is slow/ensnare available at t1, it changes the entire complexity of the game. Even if it not benefit one race, it is way more likely that some will benefit more than othere, as was the case with the initial versions of TS where the murloc huntsman was available. By default, I'll be comparing mercs with the established Lordaeron summer mercs or basic race units to see how they would do. Because the disclaimer is that much of what you're about the read is very much theorycraft as most mercs have never been seen in a serious 1v1 competition. A list of which mercs are available at which camp can be found on liquipedia: Mercenary Camp - Liquipedia Warcraft Wiki There you can also find all stats for the mercs. The analysis is based on 1.28/1.29. Ashenvale: The satyr shadowdancer offers curse, which undead players often use against the Blademaster. With this merc, they would have to invest into a temple, and it would also be available to all races (although in a weaker version than banshees). The furbolg shaman also brings a strong spell with rejuvenation (strong heal on heroes, which is normally only available to ne), all while being only slightly weaker than the standar Forest Troll Berserker. The Thunder lizard can dish out insane amount of aoe units is not comparable to any standard unit that I can think of right now. Footies will hate him. Verdict: Despite the mercs not being available early on, all signs point to these mercs having a strong and unpredictable impact on gameplay. Better avoid. Barrens: Centaur Outrunner is fine. Harpy rogues are unremarkable except for being flying units, which could be used for scouting, but I dont see any races getting a particular advantage from them. The Razormane Medicine Man offers interesting skills, and the healing ward can be countered easily during fights. Unpredictable, but could be fine. The Harpy Windwitch has fairy fire, which unfortunately could make for very strong hero nuke if undeads combine it with orb. Verdict: Less problematic than Ashenvale, it could be worth a try, but I'm not overly optimistic. Cityscape/Dalaran/Dalaran Ruins: Rogues is should be fine. The Assassin is about equal to the Forest Troll Berserker, with less attack damage but poison and hide. He should be fine, unless hit and run with poison proves to be surprisingly effective. The Kobold Geomancer is a hybrid of the Forest Troll shadow Priest and the Mud Golem. Should be fine as he is first available at the same time as the Mud Golem. The Forest Troll High Priest is the main critical creep here: He is expensive, but has 3 abilties, including a heal that isnt halfed in value like the one from the Shadow Priest. Also while you cant spam innerfire with one unit only, using it on your heroes could be strong. However, his relatively high price could balance this out. Verdict: If the High Priest is fine, then this merc camp as a whole should be. There's no guarantee, but unless with Barrens, I'm relatively optimistic here. Dungeon/Underground: Wildkins are fine. Kobold geomancer, see Cityscape. Burning Archer: With their searing arrows, they will deal a noticable amount of damage. However, I think this is balanced thanks to their relatively low hp. Verdict: Looks fine to me. If you think the weird composition of creeps fits into your map, then go for it. Felwood: Murloc Flesheaters equal ghouls. No problem here. Sludge Flingers equal Kobold Geomancers with more hp but no dispell. Unfortunately, they are available a bit earlier, which might be trouble because of their slow spell. Satyr soulstealer have lvl 2 manaburn. The consequences are beyond imagination. If nightelves buy them, they can drain all mana from hero like mk/tc within the first seconds of a fight, and for the other races he could also change the game in unforeseen ways. Verdict: There is no way this wont cause any trouble. Icecrown Glacier: The Barbed Arachnid and Blue Dragonspawn Meddler are ok. The Polar Furbolg Shaman has frost armor, which orcs will hate. I dont see this going well. The Maganataur Warrior is slightly more expensive than the Ogre Mauler and has comparable stats. But then he is also magic immuni, which can be huge. I think that because of this, he is much more useful than the Ogre Mauler. And if the Ogre Mauler is balanced in its current state, then this must mean that the Magnataur would be too strong. Verdict: Somewhere in the middle of all merc camps, but still not balanced enough for 1v1. Lordaeron Fall: Mud Golem is fine. Gnoll brute isnt, he is way too expensive for what he does. At least this means he is no balance problem Gnoll warden gives an early disable which is also particularly mean to water elementals. And the Ogre Magi can bloodlust your heroes. I think this also forces everyone to go for dispell and therefore limits strategical options. Verdict: Balanced creeps dont necessarily make balanced mercs. Lordaeron Summer: The mercs you see on the standard maps. Verdict: The safe and balanced option. Lordaeron Winter: All fine except the Ice Troll Berserker. Ensnare 60 seconds into the game? Verdict: Nope. Northrend: Ice Troll Berserker only looks different from his Forest Troll relative. Nerubian Warrior should be fine. The Frost Revenant can cast the scary blizzard, but the damage value are nerfed and the mana cost increased. Still, it could be used for siege situations, in which almost any other merc would offer nothing. The Nerubian Webspinner has some interesting abilities (web, raise dead) which are unusual for mercs but dont seem to enable any noteworthy combos. His price also looks quite cheap to me, so he could be a strong merc. But I dont see yet how he would one race more than another. Maybe humans webbing destroyers, but then again, undeads can nuke him quickly. Verdict: Probably both the most interesting and most balanced creep camp from the 3 winter-camps. Of course, there is uncertainy, but it is one of those camp where I see potential and liked to see it tested. Outland/Black Citadel: Fel Beasts deal chaos damage, which looks scary until you realize that this is the same or worse than normal damage unless you hit buildings. So they should be fine. Voidwalkers are only available at the same time as Mud Golems, and can be nuked easily. I dont see a problem with them. Dreanei Darkslayers are another units that is completely different from any other regular unit/merc. They have immolation, but they lack tankiness. They are also available relatively late. I have no idea if all this makes them strong or useless. The Draenei Disciple is who I think is the biggest issue here: Compared to the Forest Troll Shadow Priest, he lacks dispell and has less initial mana (also slightly higher stock delay, but both will be available by the time you cleared the merc camp), but better hp, damage and medium armor instead of heavy. And he costs 155 gold instead of 195. Looking at the ne vs hu matchup, it appears to me that nightelfs much rather would have the Shadow Priest available (to dispell water elementals), but humans would prefer the Disciple (more hp, takes less damage from archers, missing initial mana doesnt matter much with aura, they dont need dispell). It is all theorycraft, but it seems to me that the predicted human-friendliness of the Disciple would have to be balanced out by the Darkslayer. But as one is available earlier than the other, this looks unlikely, or at least impossibly to predict. Verdict: Interesting creeps, but in all likelihood they will upset balance in some way. Would like to wrong. But does has anyone seen a 1v1 outland map anyway? Sunken Ruins: The Giant Sea Turtle is fine (despite missing on liquipedia, which apprently isnt perfect. Yet). Weaker than A Troll Berserker, but more expensive and available earlier. Makrura Snapper is roughly equal to the Ogre Mauler (less hp, more armor and attack), but also available much earlier and only slightly more expensive. Still, he doesnt have any unusual abilities, so I dont think he's a problem. The Murgul Snarecaster is Kobold Geomancer...... but with better stats and cheaper prize? His only downside is that he gives more exp when killed. I feel something is wrong here. He might even be more fittingly described as a Troll Berserker with slow and dispell. If the Troll Berserker is balanced, then again this should mean that the Snarecaster is too strong for a merc. He is only available relatively late in the game (same time as the geomancer), so that might be his saving grace. Makrura Deepseer: With summons and a entangle (weakened though, cant be cast on heroes), he is one of those mercs with an unusual skillset. At 500 gold, he is also the most expensive merc by far. His entangle gives undead and humans the ability to prevent bears from being staffed back home, which could be strong. Or you can easily add a strong disable to an army which normally wouldnt have one But does this justify his price? I'm not sure. Verdict: With two of the mercs having Lordaeron summer counterparts to which they look superior, and one being entirely unusual, it should be unlikely that this merc camp doesnt change the balance in some way. None of the mercs is available early though, so there is a chance it could work out nevertheless. Village/Village Fall: Kobold: Irrelevant. Assassing, Kobold Geomancer: see Above. Murloc Huntsman: Ensnare in the earlygame? Verdict: Nope. Why do we always see the same creeps? There are dozens of creeps available in the world editor, yet only the same few 30ish can be seen in the competitive melee mappool? Why is this the case? I believe there isnt one big reason, more like several small and sometimes highly subjective ones. Maps are often created to fit in the wc3 lore: And as it happens, for better or worse, most maps are Lordareon Summer maps, or from a closely related tileset like Lordaeron Fall or Village. Most of the non-Azeroth creeps would not live in such a place. Players are used to these creeps: Creeping routines have been standardized, creep AI is well known (regarding the standard abilities like ensnare, LS, purge) and if you see a new map, it simply helps getting used to it if you dont have think for every single creepspot how you approach it in the best way. Blizzard never really bothered to balance some of the more unusual creeps: There are some creeps who flat out are too strong/too weak realative to the standard creeps. If you use them, you have to be very careful not accidentially create a map where creeping is much more/less rewarding than on standard maps. It would take days to make a complete list as I did for the mercenaries. Just saying that you should also look at stats (dps, armor class and value, dps) of some creeps, which define how fast you can kill them and how much damage you will take. I think there is a lot of variance there, and it matters a lot if you need 50% more or less time to kill a creep, compared to an equivalent standard creep. If you dont want to mix different "families" of creeps, it can be surprisingly difficult to create good creepcamps: On most standard maps, different classes of creeps are rarely mixed. Normally you only have troll/ogre and bandit/renegade wizard as default combinations. I personally find that a bandit/harpy or a hydra/ice troll creepspot, for example, would look silly. Of course, there are some combinations which are rarely seen but imho would "make sense", such as renegade wizard/skeletons. But going on to some more unusual creeps, I often find that many of the families lack "natural" partner creeps to make up for good camps. Of course, this is very much my own opinion. Many of the creeps are just flat out bad and cant be balanced without changing them entirely: Some abilities are just not made for 1v1 creep. Manaburn (unfun, requires completely different creeping techniques to dodge), plague/death coil (one race is completely immune), possession and probably a few more. You just cant use those, and would need to mix them with others... see above. Despite this, I think more variety should be possibly here without too much trouble, as long as one avoids some errors like flying-only creeps or using the same creep too often. Other topics for this guide which I have in mind so far: Mainbase layout Why a map should not have "paths" connecting areas of interest. Why forests need to be 99% dense. Also to come: Nicer formatting of this post.