Dps and how it interacts with StarCraft 2

In this first example, we take a look at the firing rate and DPS of a Mutalisk in BW and SC2. In Brood War, the Mutalisk has a firing rate of one attack per 2 in-game seconds with the same damage per attack as the SC2 Mutalisk, 9. In StarCraft 2, the Mutalisk fires an attack every 1.5246 seconds in-game time. Even if both game values at their fastest speeds are translated into real-time DPS, the SC2 mutalisk still deals about 20% more DPS.

This means that the rate at which the damage value of 9 is dealt, increased in SC2. Their DPS increases. The effects are clear: The Mutalisk kills things faster in SC2, reducing the time of a fight, and reducing the reaction time of the player on the receiving end of said damage.

When player A harasses player B with Mutalisks, player B only has a few seconds to react before his whole mineral line is wiped out, or buildings are sniped. This is not too bad on its own, but it can cause other problems. The prime example being production time compared to damage done in a certain timeframe.

In a hypothetical situation, the receiving player B is a Terran with 10 Barracks worth of production in his base. His army is on the move to secure a strategic location on the map so he will use his next production wave to deal with the Mutalisk threat while his two Missile Turrets die a swift death. In BW, the speed at which damage is dealt is lower, which means the Terran has more time for his wave of 10 marines to finish that can push back the Mutalisks attacking his mineral line. This in turn means that the Mutalisk harass needs more time to be extremely destructive and game changing in terms of the outcome.

Now in the same situation in SC2, it will take the Mutalisks less time to wipe clean your mineral line. The units in SC2 have an equal or slightly faster build time than in BW, but not enough to compensate for the difference in killing speed that the Mutalisks now bring to the table. It now takes less time for the damage to be critical to the outcome of the game, and relatively longer for the next production wave of marines to spawn to fend off the Mutalisks.

One could argue that this DPS increase buffs the incentive to harass your opponent in multiple places, multiple times during the game. This is true when looking at this specific situation. But as I add more examples of bigger attacks and differences in time, something interesting will reveal itself: That in the big picture, harass is oftentime deemed less effective in SC2.

In SC2, most siege units have a very consistent firing rate with strong splash damage. I will take a look at the Siege Tank and the Colossus in this example. Once again I will use BW to help explain my musings on this particular type of unit and damage type.

In SC2 the Siege Tank has a fairly fast firing rate when in Siege mode: 3 seconds (2.175 Real Time seconds). To balance this out, the frontload damage of its siege attack is pretty low at 35 (+15 against armored), and the splash radius is fairly large. This means that when a player makes a positional mistake against this unit, they will take some area of effect damage, but the mistake will not be too punishing in cases where retreating is a possibility. To keep it simple, this situation will be an accidental engagement where the engaging player did not see the Siege tank line beforehand but still has the option to retreat after the first volley.

While it will not be ideal to move in a clumped up formation when engaging this siege line, it will not be punished too harshly. The penalty for engaging with a bad formation is moderate.

It is time to check out the Siege Tank from BW. When it is in siege mode, it will deal 70 damage per shot, with a cooldown of 75 frames (on fastest settings it will be ~3.75 Real Time seconds just to clarify). This means that the frontload damage of the BW Siege tank is extremely high, but it takes a while before consecutive shots are fired.

Now in the same situation where a player engages a Siege Line incorrectly or by mistake, he will get punished a lot harder for engaging with bad positioning. Even if only hit once before correcting the mistake and retreating, the damage is severe. Because of this severity in frontload damage, and only slightly lower sustained DPS over time (mostly due to splash increase of the SC2 tank), the emphasis on proper positioning prior to engaging increases.

If we compare this to a Colossus, which has very high sustained splash DPS, a problem arises. It does so much damage consistently (near 75% more than a Siege Tank) in such a short amount of time, with the beams always hitting, that it lessens the incentive to micro. A micro move against these units barely has any effect at all compared to the time it takes to execute said move versus the damage dealt by the colossus. A concave can help, but the Colossus is so mobile that it can force an engagement on favourable terms with comparative ease. The damage is so high that units die extremely fast and battles are over even faster. Due to the high DPS combined with the mobility of this unit, any mistakes made while fighting against it quickly snowball into a loss. There is less time to respond and rebuild.

Because of the nature of the Colossus, it is generally favourable for the Protoss player to keep these units close together with the rest of their army for maximum damage output.
In short, high DPS siege splash units put more emphasis on muscle over micro, where lower splash DPS with higher frontload leans more towards positional engagement priority and rewards micro more.

The power of AOE spells in SC2 is quite strong, but it also has some problems when it comes to the distribution of the damage over a certain amount of time. Take Psi Storm for example, a spell that deals 80 damage over 4 seconds in a relatively small area. This means that the damage builds up extremely fast but that the area of control is relatively small, and the DPS lower than a BW storm (double checked the numbers). Here is where the comparison to BW Psi storm comes in.

In BW, a Psi Storm deals 112 damage over 8ticks of 14 damage. It deals this damage over a larger radius as well and deals its damage at about the same speed as Psi Storm from SC2. In BW it was really hard to cast multiple Psi Storms when compared to SC2 (no smart casting in BW), so it was relatively harder to blanket armies. Due to the larger area of effect, it was a better tool to control the board to make up for it.

Battles that involve the old Psi Storm last longer and allow for more control for both the casting and receiving player. There is more time for production waves of units to finish and for reinforcements to arrive, further increasing the duration of fights and their aftermaths. Comebacks are also slightly easier to be made when more production waves can finish before the enemy stands at your doorstep.

It seems that in this case, the higher DPS of the BW spell combined with the time needed to set up storms actually lowers the odds of the same DPS being put out by storms in the first place. Perhaps a slower buildup of damage but a larger radius would create the same type of effect for SC2 Psi Storm and its ability to truly lock down areas. In this case lower DPS vs larger radius could be something to test.

Now that I have outlined some of my musings on the DPS of different units and AOE damage, I want to write about the correlation between DPS and production. The DPS of almost all units and spells in BW is lower than that of SC2. The production time of most units is only slightly longer (5-10% on average) in BW than in SC2, but the rate at which damage is dealt in SC2 is close to 15% more than that of BW in most cases.

This is where a problem arises: Comebacks are harder since there is less time for a player to recuperate through production waves before the enemy can seal the deal, almost nullifying the defenders advantage of production vs arrival of the enemy force. The defender cannot buy much time with AOE spells to control the field either, since they do not last long and affect small areas.

This also creates a bigger incentive for a Death ball to be created, since the risk of losing a game often lies in one battle. Because of this, a player wants his force to be as strong and deadly and compacted as possible to deal the most DPS in the least amount of time, increasing his chances to win the one deciding battle.

In Starcraft 2, you can select a ton of units at once, and move them across the battlefield with but one command. The game also has smart casting, disallowing multiple spellcasters to cast the same spell simultaneously on the same target or in the targeted area, wasting valuable energy unless micro'd individually.

When I then draw a comparison to BW, where smartcasting did not exist and the maximum selection was 12 units, something interesting happens. When there are more actions required to execute certain mundane tasks efficiently, including army movement, spellcasting and control in general, there will be more time needed to execute all these actions. As a result, there is a difference in the rate at which damage is dealt per real-time second when the armies grow bigger.

In other words, if you would let 2 pro players play out the same battle with a maxed army twice in SC2, once with all the new UI and smartcast functions, and once without, we will find that the latter battle will simply take longer. So in a very indirect way, damage done per Real-time second has increased through the UI and behaviour improvements going from BW to StarCraft 2.

This unfortunate side effect probably wasn't intentional, and might explain the difference in feel when comparing fights from both games. The problem is that this increase in damage per Real Time second is hardly measurable, as it differs per player, per situation and per skill level difference.

Now couple this with a slight increase in game speed, and all the sudden no one can effectively keep up with all the damage being thrown around in large scale battles with high DPS density (Deathball), because sadly there is a physical human limit as to how many actions we can perform per second in a game such as StarCraft. As I will mention later on, this physical APM cap is already close to being reached and there shouldn't be much room left for "skill and speed improvements" for even the best/fastest players.

There is no easy way to account for this indirect increase, as we simply shouldn't want to go back to limited selection, no smartcasting etc. But hopefully we can snowball some thoughts on what could be done that would not affect the UI or fundamental AI elements.

With these examples in mind, it is time to take a look at the macro mechanics of StarCraft 2, including income rates, supply and production mechanics.

In SC2, a proper income rate is reached on relatively few bases. Mineral lines get fully saturated quite early and due to the supply cost of units, only few bases are needed to reach a maxed army of high DPS, cost (in)efficient units. Because supply is limited, it is generally not favourable to harass an enemy with more than a small amount of your army supply, since you simply do not have much to spare before the balance in a big engagement greatly favours the player with just 15 extra supply worth of units. The only exceptions are cost and supply efficient units like the marine, which explains why Terran is the race that has a bigger incentive to harass. Whenever a race has cost and supply efficient units with high dps, there is a bigger incentive to distribute them over the map for skirmishes than when a race has even higher DPS, yet supply and cost inefficient units.

When I take a closer look at the race specific macro mechanics, it seems that they all promote the high density DPS ball behaviour. We have one race (Terran) that is extremely resilient to economic harass through the use of MULES and the cost efficiency of their low to mid-tier units in general, which decreases the incentive to harass their mineral lines and promotes late game, high DPS high density engagements for their enemies.

Then we have the Zerg race, which due to the inject mechanic, can bank production and increase production rates to the extreme. This promotes the use of high DPS, high density armies in the late game to combat the armies of a Zerg, to increase the odds to survive a clash well enough to push through and deal with the remax before the whole wave of production has spawned. This in turn causes the Zerg to build spine walls and play a turtle style late-game to ensure the chance to remax to higher tiered units/renewed armies.

Once again due to the high DPS in the game, combined with the macro mechanics of SC2, time becomes the crucial factor in determining the most successful way to play: High DPS, high density maxed engagements.

On to the Protoss race, where we find that most of their units tend to be supply and cost inefficient, but extremely high on DPS and AOE in the higher tiers. This in combination with the ability to speed up crucial upgrades via chrono boost to enhance timings or economy promotes high DPS, high density engagements once again.

Due to the warp-in mechanic, Protoss also nullifies the defenders advantage of production waves vs travel time and reinforcement travel time completely. But due to the cost and food-inefficiency of the Protoss lower to mid-tier army, they cannot afford to have much of their supply strayed away from their main army.

This leaves the Protoss with two options, playing defensive to build up a high DPS, high density army, or hitting a timing attack. Harass is very limited due to cost to food ratios and the inefficiency of spreading your army apart too much.

This brings me to the last of my musings. SC2 is an extremely fast paced game, faster than BW. On top of that the DPS values have increased a bit when translated to real-time DPS compared between the two games while micro incentive diminished due to that. No one wants to micro more than the minimum as to not risk losing too much DPS time in big battles.

While SC2 still needs a lot of micro at the highest levels, it is limited not only through the high DPS, high density problem, but also the problem of the physical human limit. Many times I have read the argument that we need to give SC2 time, that pro players will “ learn to use more control groups and micro more and more”. The problem with that statement is that the top players are already close to their physical limit of what they can do within a certain timeframe in the game.

Due to the speed and DPS output, only few actions can be executed to change the outcome of a battle, while for some units and against some units, there isn’t even an incentive to micro at all. The speed and DPS weaken or nullify the effect of subsequent production waves, hindering comebacks and defenders advantage.

The Deathball issue of Starcraft 2 fails to be tackled by the FRB, Dynamic movement or Breaking up the Deathball alone. It seems that the game speed and DPS of units combined with the macro mechanics of the game play a big role in the efficiency and incentive to play a deathball style. The best players are close to maxed on Micro APM and cannot get to a level where the higher DPS and game speed would not be a handicap to the dynamics of the game.

It might be a good idea to look into the damage, macro mechanics and time values of SC2 in order to come up with a solution that gives players the ability to accomplish more with micro, buy more time and gain a better defenders advantage. This will lower the incentive to play a deathball style, and increase the incentive to skirmish and prolongs the duration of battles and the effect of subsequent production waves.

Note that all of the above is my personal, current opinion. If there are any errors in the values I used, please feel free to correct me. I am not in a position to correct grammar or rewrite (well apparently only once and it hurt!) parts since I have arm injuries, and I wrote down my musings in the past 5 days as they came. My apologies in advance if it is all over the place.

I hope that this thread will help educate people and bring progress in trying to help improve the game to be even more enjoyable, and more based on skill, epic battles and multiple skirmishes. Perhaps it will help us find solutions to stray away from 1 clash of doom in certain matchups and on certain maps/metagame situations.

I have been thinking about things that could be changed that would not affect the game engine or require a ridiculous overhaul of the game, since anything that requires said overhaul needs the full support of Blizzard. I am not sure if they even want to tinker with the game engine right now.

That being said, there are some small things. I haven't been able to fully work them out to know how they affect all areas of the game though, so all feedback is more than welcome!

1) Increase of the supply cap to 260+
Pros:
-No need to recalculate supply values of all units
-More supply to distribute over the current maps
-Takes longer/more bases to get maxed efficiently

Cons:
-Low-end computers/laptops might not be able to handle this. Blizzard wants this game to be highly accessible by many specs, including low tier.
-Might increase power of end game deathball and impact comebacks negatively.


2) Increase the mining time of minerals (this is easily done in the SC2 Mapmaker, just go to the mineral values)
Pros:
-This makes it so that any worker past 16 on a mineral line will add way less efficiency, which makes it less effective to stay on a low base count, and more attractive to get more bases.
-It will free up supply in earlier stages of the game, since you need less supply invested into economy per base.

Cons:
-It will slow down the early to mid stages of the game. Blizzard wants the game to be action packed from the get go. Will they support a change that reduces the speed of the earlier stages by a bit?
-Does it change the efficiency of 1-base all-ins? Not sure on this


3) Give more units the ability of moving shot, which means instant turn rates to fire (like how a marine can instantly turn and shoot, and a stalker cannot since it has a delay on turning speed).
Pros:
-Increases micro incentive/rewards the skill of a player in skirmishes and early to mid game.

Cons:
-Will not increase micro incentive enough in large scale end game Deathball battles, core problem stays around.
-Requires Blizzard to alter unit mechanics. Not sure if support can be found.

Added these to the original post, after answering the question "what would you do to facilitate positive change?"