Ranged Units in Warcraft 3: Part 2, By MaDDoX

Difficulty: N/A
Type: Theory
Focus: N/A
This article was submitted by MaDDoX, a South American player, who was kind enough to share his research with zileas.com. This is the second portion of the article. The first can be found here

In the first part of this article, I presented some some theory and actual test results of ranged units (rangeds, as I called them). In this second and last part of the article, some considerations will be made about rangeds optimization and also for ranged siege units.

Rangeds use optimization:

To improve your actual game performance you should utilize the three adventageous properties of ranged units:

Range: If available, spend all possible cash to get the rangeds' range upgrades (long rifle, improved bows) as fast as possible. That'll make it much easier to pick fleeing units and heroes. Tests have shown that the range upgrades for mid-large ranged masses are more effective than the actual ranged weapon damage upgrade.

Attrition Factor: Never leave your rangeds unguarded by a melee force. It may seem obvious, but sometimes in the heat of battle, it is easy to overlook. Preplanning is key, as is being attentive during battle. I recommend using group shortcuts, to quickly retreat if necessary, or to try to cover them with melee units.

Damage Stacking: Once you have a melee force large enough to offer good attrition - dependant, as we've already discussed, on map design, choke points wideness, etc - start pumping as many ranged units as possible to maximize their damage stacking advantage. As always, focus fire enemy units, closest or more hurt first, killing them one by one.

Rangeds Tactical Analysis

Some strategic abilities must be evaluated in the overall game balance picture to assess the efficiency of ranged units in Warcraft III. In particular, the strengths of defensive enhancement, choke points, and anti-air capability are very important.

Defense enhancement: 5-6 towers with a few peons around on auto repair is a very serious static defense force which, if not addressed by siege, can in a very short time wipe an entire army. However, going to this extreme is basically an invitation for siege, and therefore a more realistic defense is usually 2-3 defensive structures in an effective configuration. Effective configurations use normal buildings as shields, while allowing reasonable repair access for your worker units. As explained in the previous article, ranged units work well in combination with static defenses in that they can hide behind buildings and attack without sustaining damage. In general, beacuse Warcraft 3 is an army-centric game, it is generally favorable to use ranged units for defense more than static defenses when possible.

Choke Points: Choke points can greatly increase the effectiveness of your ranged units. On the Dark Forest map for example, I find myself building a great number of rangeds since it's so easy to take advantage of the natural bridges that are a nightmare for any invading army. I used to say that maps make up for up to 40% of the gameplay balance, and my guess is that the final maps will be much more "closed", which will further bolster ranged unit effectiveness. Always keep in mind that, if there's no choke point in the map, or the map is just reasonably open in general, you can stlil make "artificial chokepoints", as Zileas call them, using static-d and other structures. Good tactical control allows you to taek advantage of these and other chokepoints. Its often important to draw an opponent into a battle to facilitate this.

Anti-Air(AA): Since patch 1.20, all rangeds perform adequately as mobile anti-air. The real problem is getting the rangeds in position to output their AA damage. Although Blizzard tried hard to not make the Air units overpowerful by slowing them down, the fact is that this makes them less effective for fast response roles. However, they are still adequate, in general, for base defense. When combined with certain spells, such as slow, it is possible to corner and eliminate ranged units in abundance. To counter siege-air like Frost Wyrms or Chimaeras, your best option is currently still making your own air, since rangeds many times won't reach them if they attack while over trees. It is unlikely that this will change, and for good reason.

The Siege Factor

In RTS games, siege can be used to force your opponent to respond with its long range. This typically takes the form of being able to outrange static defenses, and this is definately the case in Warcraft 3, where Siege's primary purpose is base destruction. As I mentioned in the first article, the more range the ranged unit has, the more powerful its damage stacking becomes, and less the attrition factor impairs it. When coupled with decent mobility, it also gains substantial "Skirmishing" effects.

In Warcraft Beta 1.21, siege attack damages normal and small armor units (melee and ranged respectively) with full power. That's what made the siege units the main "attraction" in the latest 1.21 patch, and the reason why I have included some siege units tests in this article tests chart, to demonstrate that siege is over-effective.

The improvements to 1.21 siege, which clearly were too great, were probably a result of the Blizzard designers wanting to make siege more used and needed in actual battles, since siege unit cost did not justify their use in battles, but only justified their use in base destruction. Unfortunately, base destruction almost always occurs when you already have a large advantage in battle.

Fixing Siege

I don't believe there is an easy answer to fixing siege, though there are some answers that solve most of the problems. From my experience during the HTAK design process (a TA: Kingdoms redesign and balance tuning mod), I developed a ranged balance theory I call "damage range compensation". The theory is that damage should be inversely proportional to range, and when one unit exceptionally needs more attack power, the extra damage is proportional to an extra vulnerability. A correlary Zileas suggested that the greater the range of a unit, the slower its movement speed must be to keep it non-abuseable. I add that that's also generally true for raw attack power, as demonstrated with 1.21 raiders. Fitting this theory, its extra damage in 1.30 is proportional to an extra vulnerability to pierce damage.

This theory accounts for many subtle interactions in the armor/damage system of the units, which are listed below:

1) Melee: One obvious consequence of this system, if literally applied, is that the melee units becomes the undisputed basic unit. That was usually the case of WC3 beta from 1.03 up to patch 1.13. The problem is that, if melee works so well vs. ranged and siege, it is difficult to not end up nullifying the usefulness of ranged units, sinceu siege normally counter ranged, and ranged counters melee. units again. This is partly related to the fact that in large numbers, ranged will almost always have to outperform melee by a good deal, to keep them fair in "normal" game situations.

In 1.21 although a small group of archers seems ok, you can reach "insane" level pretty quick. One obvious way of avoiding this is by increasing rangeds buildtime. To stop people from simply putting up more barracks, some sensible cost increase for the rangeds should be necessary. The other way, is to use convenient area effects, such as siege.

2) Siege: Siege shouldn't simply defeat rangeds. Since both ranged and siege are supportive (ie. secondary) units, they shouldn't mutually annihilate before being able to deliver their damage to the "star" melee units. If there aren't major limitations to the longest range unit, all other units become obsolete. Making siege units damage against other units very weak (super-weak for melee, a bit better for ranged), reducing its food cost and maybe increasing its AOE would make artillery siege interesting although much harder to abuse. Other good option, similar to the Warcraft II formula, is increasing siege buildtime and increasing its wood cost. Any of these options would improve the current situation and are probable changes for next patches. Zileas' recommendation is to simply increase their minimum range, and slightly reduce their usefulness vs units, and moderately reduce their effectiveness vs heroes. He also notes that a lot of the problem is that there is a major bug with ballistas, which won't be specifically named.

3) Air: One important aspect of air units is that it logically should counter siege, and if it has this property, serves as a good way to balance siege. Currently that doesn't happen appropriately really. Optimally, air units should be faster, slightly less resilient, and harder-hitting so as to suit as a mobile quick-reaction outpost defense.

4) Siege vs. Siege: Siege desperately needs reduced damage vs. other siege. Otherwise, countersieging is too potent. To make the situation worse, defensive siege currently also kills ranged units, so air defense is over-effective if you turtle and under-effective if you play "correctly", ie. using a balanced attack/defense force which strikes at siege directly.

I have included some data regarding ranged units in 1.21.

Conclusion:

Its pretty clear that while ranged units were not useless in previous patches, that they are, in general, too powerful now (especially siege). Most of what is really hampering the current beta balance can be summed up as rangeds, siege units, hero abilities and MAYBE mid-game unit strengths. It would seem that Blizzard believes the overall problems are minor, since they seem to have code frozen (remember Rob Pardo's post declaring that they would make no further major changes?). Depending on the nature of the final adjustment of the ever challenging melee-ranged-siege matchup, the game balance probabaly will head towards a melee- or ranged-prevalent game which should please some and dishearten others. I'm generally optimistic, but its bound to be interesting...