EPT[G] Judging a Melee Map
Arguably more effective than a guide on how to make a melee map.
Not meant to replace actual testing.
Arguably more effective than a guide on how to make a melee map.
Not meant to replace actual testing.
I have decided to (perhaps temporarily) semi-retire from making and studying melee maps. If you're wondering, I'm doing UMS/Arcade mapmaking now so I'm not far ^^. This post could be viewed as a summary of most (aiming for all) of the more important things I understand about analyzing and judging melee maps - I am writing them here before I forget.
What are we judging? You mean there's more to this than pretty pictures?
If you put a lot of emphasis on aesthetics, you are not an ideal judge of competitive melee maps. Sorry. There are dozens of people capable of decorating a map from scratch to a better-than-acceptable level in less than an hour or two.
Every strategy works better on some maps than others because each map encourages different ways of playing, even it if may sometimes seem slight. Games are quick and bloody on Steppes of War for tangible reasons; likewise, there is no mistaking that macroing up on Metropolis is generally a safe bet.
But not every map gets the privilege of being played as much as those have, so we must use our brain to reconstruct as many games as possible. Make no mistake, actually recreating a single complete game in your head is incredibly inefficient and virtually impossible if you value accuracy.
What we do instead is get a feel for all possible scenarios (broad strokes, but as detailed as you can) and put emphasis on those which are most likely (from skilled players that are trying to win). You can't just look at a single replay and think that it encompasses everything about that map - not even close. The substance of a map consists of tendencies along many spectra that are nearly impossible to analyze completely due to an inescapable lack of information.
IMO, the key is to look at it from a variety of limited perspectives. The following is essentially a list of perspectives I have found most helpful, separated into stages purely for convenience (not a rule).
0: Prerequisite Knowledge
Game Data
- Actual (hidden) unit and building stats/abilities/etc.
- Map mechanics: Rocks, XWT, LoSB's, Pathable/Cliff/Air, etc.
- Economic understanding: supply cap, worker cost, workers per base, expansion cost, resource collection rate, minerals/supply of basic units, etc.
How the pros play
- Have experience with the game data in action.
- How do people actually play the game at a competitive level? How do they actually apply the game data in performance? What are they capable of (this goes beyond current metagame)?
- It's a good idea to be very good at playing the game yourself.
Six Separate Matchups
- Mirror: ZvZ, PvP, TvT
- Core: ZvP, ZvT, PvT
1: Sizing It Up
Map Boundaries
- All distances are within this area. Get a feel for it.
Initial Rush Distance (Main2Main)
- Extremely important, as it sets the pace and leeway for initial build orders.
- Shorter distance = earlier aggression becomes more viable, more easily punished if failed.
- Longer distance = earlier aggression becomes less viable, harder to punish if failed.
- All things are more important (have a larger impact on the game) the earlier they are used - the 'snowball' effect.
Central Map Width
- If the "Main2Main" distance is the length of a map, then the Central Map Width is it's width.
- How easy is it to cut off or scout the thinnest part of the center of the map? How long does it take to move from one side to the other?
2: 'Features'
Global Resource Density
- Look at all of the bases that can plausibly be held by a single person simultaneously. In general, these bases shouldn't be too close or too far from each other.
Local Resource Density
- Specific areas on the map with more bases are more easily defended and encourage more rapid expanding.
- Areas with less bases are harder to defend and encourage less rapid expanding.
Gas & Tech
- Gas availability is strongly related to players' ability to use more variety and higher tech units.
- This is very normal on most maps - but on the maps that have unusual gas it is worth paying attention to.
Aggressive vs. Defensive Expanding
- Generally, expanding away from your enemy is a defensive move and expanding towards them is an aggressive move. Keep in mind a great offense can be a good defense.
- Zerg usually has the hardest time expanding towards the enemy.
-There should be a good mix between aggressive and defensive expanding - going too far with either is often sub-optimal.
Expansion 'Flow'
- How easily does one base 'flow' into another? Is there a small path that directly connects the two bases? How hard is it to take?
- How hard is it to hold? Does it flow into more bases?
Attack-Defense Pivot (Army Positioning Gameplay)
- For each likely set of bases, identify the longest line you can make between the two outermost ground chokes that lead directly into them. How far away are the farthest two entrances?
- Identify how easy it is to defensively maneuver troops in this area vs. offensively. What blocks the defender (is it too 'linear')? What blocks the attacker? What if the attacker splits up his forces? What if one split is air/drop and is sent into the main?
- Each player has an area like this on their side of the map. As time goes on, the distance between these two areas becomes more important than the initial rush distance.
Raw Base Vulnerabilities
- Consists of everything not directly related to it's distance to other bases.
- Get a good feel for each base on an individual basis. Obviously, more vulnerable bases are less likely to be taken because they are more easily lost: beware of "winner's expansions".
- Get a good feel for overall vulnerabilities in a late-game scenario. Put it in context with it's distance to other bases (see section on Attack-Defense Pivot above).
- Note that sum of overall vulnerabilities in context with distances is probably the number one factor (that you have control over) that affects the 'skill cap'.
Cliffs
- Pay attention to cliffs. Where are they (especially if near the main)? How much cliff? How far are they from chokes/entrances and base locations? If there is enough in the right places, tactics and build orders can be built around cliffs.
- Are there any protruding cliffs overlooking important paths that collo, siege, or infestor (fungal) can utilize?
Air Space
- Note where and how much are space there is. More air space (especially if in the right places) gives more power to air units.
- Some air space lets units sneak by... some gives leverage (especially for Carrier/BC/BL), even if as little air space as a cliff.
- How exposed are mineral lines to air/drops?
Percent Pathable & Obstructions
- How much freedom do players have when moving units around the map?
- Too much freedom and it gets hard to find chokes to defend your bases, which can actually limit strategic depth (by making many strategy unviable).
- Too little freedom takes away too much autonomy and choices. You want players to have a lot of potential, not be limited by a confining map.
Muta Harass
- Pay attention to how easy it is for mutalisks (and banshees/void rays I guess) to jump from one base to another (ex: main to third) vs. how hard it is for ground units to catch up to them.
- How easy is it to "trap" the mutalisks? How many turrets/cannons/spores(queens) does it take to do this efficiently (if possible)?
Scouting / Map Awareness
- Are there multiple spawn locations? They can make cheeses harder to scout, but they can also give more safety to very quick FE's against midgame timings.
- Consider overlord movement speed and sight range. Understand that bigger mains are harder to scout.
- How much can an overlord see before a marine/stalker/queen arrives?
- Where are places overlords can hide, and what can they see there? What don't they see?
- Where are the watch towers?
- How hard is it to hide a proxy?
- How easy is it to see any ground movement coming through the map (at multiple intervals? Air movement? How easy is it to sneak something into the main to snipe tech?
- Does a high concentration of bases make scouting for expansions difficult?
Defensible Positions - Openness
- Attempt to highlight all areas that seem significantly easier to defend than attack (high ground, chokes). Put these areas in context with where clumped up or important bases are, and get a feel for the likely potential interactions.
- Also consider areas with high openness; Zerg in particular have an easy time attacking these areas.
- All units without splash tend to perform better against units with splash in areas with more openness.
Map Control
- There are 4 main reasons to want to control the map; (1) secure expansions, (2) deny expansions, (3) map awareness, (4) deny map awareness.
- Are these useful goals in this map? How much do you need to control the map to do those things? How much does controlling the map actually help you do those things?
- You do not actually control an area on the map if your opponent can attack your position and win the fight. How well can units utilize terrain features to hold the positions?
- Are some attack paths easy to cut off? Are some very hard to cut off?
- How viable is a tank/turret/bunker push?
- How easy is it to spread creep to important locations?
3: Game Stages
Early Game
- How easy is it to take and hold your natural? Xel'naga Caverns hard or Shakuras Plateau easy? Build orders are formed around this.
- Understand that people aren't going to be flying around with fleets of battlecruisers at this stage - the terrain should be shaped accordingly (each tile really means a lot near the natural).
Mid Game
- Consider all possibilities at the 3-4 base stage. Pick out the types of strategies that seem strongest (there is often fewer of these than you might expect)
- This is arguably the most important way to look at a a map - second only to initial rush distance - because this is often the first stage a pro player considers when crafting a strategy for the map (source:Day[9] Daily #44). Only then do they look at what it best leads out of and the best way to lead into it.
Late Game
- Consider what the game looks like when the players have almost all the bases on the map.
- At this stage, both players have reasonable access to all tools within their arsenal, and are nearing maximum upgrades. Anything goes - with a lot of bases to defend and attack.
- Aside from the metagame's ideal late game army position, it's hard to predict what will probably happen here; you might be better off considering whether or not one player probably won by now and if this stage is just formality.
4: What is normal? What is needed?
Metagame
- As previously mentioned this is all about trends and tendencies, but only one of them is truly dynamic in a given map. Players try to use map features to help them win, but in a freshly new map they also start off using build orders and strategies meant for other maps - things that have worked for them in the past. A smart player will expect this and use map features to help them punish it. An experienced player will anticipate this innovation and respond accordingly. A smart and experienced player will... you get the idea. This is known as the 'metagame'.
-It is important to understand the map's role in this metagame. Even brilliantly interesting maps are shunned if they stray too far away from the currently established metagame.
Compare to Other Maps
- Try to place a category on the map. [Big map? Medium? Small?], [Short rush distance? Long?], [2 players? 3? 4?], [10 bases? 12? 14? 16?], [Rotational symmetry? Mirror? Mixed/shifted?], [Square? Rectangle?], [Easy nat? Hard nat?]
- Try to find the map(s) in current use that is most similar in category to the new map. What details of the new map set it apart? What is genuinely new/unique? Is the newer map worth it?
- We should keep using accepted maps until we're sure they suck or they run out of steam, but keep in mind that innovation keeps the game alive.
Map Pools
- Brush up on map pool philosophy here: [G] Map Pool Creation Guide
- Some maps are already established, great, and new enough and don't need maps of a similar category overlapping with it in a map pool. A map can be a little better than others but still not worth putting in the map pool over the others (throw it an honorable mention).
5: Icing on the Cake
Balance
- Try to have a reasonable mix/balance of defensive and aggressive expansions (in terms of distances)
- There should be a nice mix/balance of open and tight areas throughout the map.
- Siege tanks that can cliff a natural or third hatchery are highly discouraged (etc.)
- Make sure protoss has a reasonably easy time with sim city at nat and third.
- Make sure zerg has plenty of flanking opportunities.
- If all matchups are within 45-55%, we consider the map balanced. 40-60% can be worth trying.
"Fun" and "Spectator Value"
- Do not make too many new things for players to learn; it can overwhelm them which is essentially what stress (kinda opposite of fun) is.
- At the same time, it's important for the map to stand out. People should be able to look at a feature or set of features and be like "woah, that looks cool," as their mind races to imagine scenarios/strategies playing out on it.
-Heavy macro, turtle-based maps quickly lose their novelty/replayability/watchability. They have deceptively low skill caps.
- People like to see variety (a lot of viable strategies) - they like to see very different types of forces clashing.
Clean & Pleasant Aesthetics
- Progamers don't care about your waterfalls, rivers, jungles, etc. They want different areas of the map easily distinguishable; they don't want to guess where the pathing is (or isn't).
- At the same time, we like our special shiny SC2 graphics (in contrast with BW graphics). The bar isn't that high, but it is true that some maps have ridden on stellar aesthetics.
6: Finish
Reflection & Meditation
- Try to think about everything you understand about SC2/mapmaking and how it applies to the map. Or just pick one feature and focus on it. Maybe clear all other thoughts first.
- Maybe do this when you are falling asleep, or taking a shower, or in a car ride, (or after you smoked a bowl ^^), or actually playing on the map.
- Always keep your mind open to new perspectives.
Map Concept vs. Execution
- A map concept is the idea you have for it in your mind - what you were aiming for. Execution is how close you actually got to the concept (most errors are in proportions).
- When judging a map, try to get an idea for what it's concept probably was or probably should be. If the actual map isn't that good, compare it's execution against what it's concept probably was or probably should be. Sometimes maps with poor execution have really, really good concepts - do not let these concepts be forgotten.
Conclusion
- Think of maps kinda like fruits. Can it be rotten (poor execution)? Yes, though thankfully you can keep trying.
- But is an orange better than an apple (map concepts)? No. But I have an idea of which one I'd choose if I wanted something juicy.
- Barrin
PLEASE REPLY WITH ANY ANALYZING PERSPECTIVES NOT LISTED ABOVE
"
