|
Hi,
This idea came to me in a match when me and my opponent (both silver leaguers for now) were camping in each of our own bases each waiting for the other to attack.
The advice sounds pretty obvious at first but one may be tempted to just focus on macroing up and leave his/her units near their own buildings. Yet all the professionals seem to be always attacking. Why would this be more effective?
Don't we have at least equal chances, leaving our army at home? Or better yet, maybe it is optimal not to attack ourselves so we have more concentration to focus on the enemy in case he attacks us.
Ok, first of all, in order to win, one must have an army, otherwise the opponent's buildings can't be destroyed. You can also win by the other player resigning but that is not in your power. The perfect opponent does not resign.
This means that, in case you want to win, at least one probe should not be gathering minerals or one must have built at least one pure army unit. To win at least one unit should be ordered to attack. That unit can be either a part of your army, or your entire army. Suppose it is only a part of your army. Then you have a fighting unit that is not attacking at the deciding tick. Then it is impossible for that unit to contribute to you winning. So that unit should not have been built, saving you some APM (That APM could give you extra energy for other matches in case it is significant). So your entire army is attacking at the time you win. Maybe I'm indeed overthinking, so I'll leave it at this for now...Greets
From previous version of this post:
A part of your army attacking has less chance to win than in case it was you attack with your entire army (strong army defeats weak army). So having units that aren't ordered to attack is a losing strategy.
So lets suppose that the "ultimate" strategy has a army/economy ratio of 5%.
(5% of units are fighting units). On top of that, one cannot win without ordering that army to move towards the enemy and to attack that enemy.
Second, since a bigger army has more chance to win, one needs to always be increasing their total empire size as opposed to going for a stable or decreasing empire.
Third, a unit, when ordered to attack at an earlier time, outperforms that same unit when ordered to attack at a later time. If we assume exponential empire growth, an army of 5 units can be for example 100% of your total army at 120 s and only 10% at 180 s. So by the time your army arrives at 180 sec it is only 10% of the size of your enemy's army (who also uses perfect play).
Since an army consists of individual units that each are better when ordered to attack at an earlier time, it is better to send units continuously as opposed to discreetly.
Fourth an attacking unit that does not attack sets you behind because you could've built a worker unit instead.
Therefor, since a part of your units are attacking units, they should be moved towards the enemy. Maybe that's what Tasteless refers to as "playing against the game" as opposed to only playing against the opponent?
Note: if the enemy is in your base, moving means ordering them to go in attacking range of the enemy units in your own base.
|
A mirror match game of perfect information would result in never attacking, but the game is all about imperfect information and pulling your opponent out of position so that you can attack, otherwise you would always lose being the aggressor due to defenders advantage. The game can never be black and white like a game of chess can be due to the imperfect information (in chess we can see the entire board, know the turn order and therefore can logically conclude the winner or draw with enough computational power from any given state if played perfectly). Also there is asymmetrical balance in non mirror matchups where you have advantages by attacking or defending in each matchup so you will tend one way or the other in each.
|
"Preparing for a matchup" vs "preparing for an opponent" refers to philosophies of preparation, not the ebb and flow of combat units.
Preparing for a matchup: In a tournament like GSL the players have a lot of time to prepare between rounds. If you played a matchup on the same map 10,000 times, you would have a general idea of timings and would be able to formulate a robust playstyle that can address stuff that might try and catch you off guard and eventually build up a solid composition that can end the game somehow (whether it's winning a fight and pushing in for the kill, or starving your opponent out over time). So you construct an overall gameplan that doesn't have any glaring weaknesses and is generally well-rounded enough for you to survive long enough for your mechanics to shine through. No matter who the opponent is, you're at least somewhat ready for almost anything they can throw at you, but you sacrifice unpredictability and your "pattern" can be recognized if you do it enough times.
Preparing for a player: Since there is a lot of footage and since top level pro players often run into one another a lot, it may be possible to discover some preferences or predictable patterns in a player's general approach to a matchup. Some players might favour an all-in in game one of a series to gain an early lead, for example. In that case, you would try and formulate a defensive or counter-attacking plan based on that likelihood. In weekend tournaments, or in situations where you have no information to draw off of, you're not able to adopt this approach. Some players never or seldom try this type of approach, because they aren't able to carry over the benefits of practicing such an approach into later rounds, or never get to play in tournaments where they have enough time to examine their opponents' games.
EDIT: On the idea of always moving toward your opponent, the reason why people don't always attack, is that they don't know what they're attacking into. Generally, it's the idea that you're able to win a fight that pushes you into taking it. It isn't just a leap of faith. Pro players fight more often because they're mechanically capable of doing so. As soon as they realize a battle isn't going their way, they fall back or commit deeper, but only after they weigh the cost.
|
One pro's comments: 'if you're not attacking you're probably losing' is apt.
|
I believe you are overthinking. For me the easiest ideas are better for lower levels (below master 2) (I'm diamond one so also low level). I think having an overall idea based on the build is better. Like I'll go fast 3 cc so I'll defend until my macro kicks in. Or I'll open 2 base and be agressive.
Ofc the style you go for depends also on stuff like multitasking. I've had a lot of opponents who went super agressive and micro intensive but once they got stopped they had nothing on their side.
But generally I feel better having my army on the other side. It let's me see their army composition, get a better feeling of who's winning the game, and then it helps assesing if you can take fights or not (by just trying and learning from the outcome).
|
You could try and synchronize the timing of your attacks with the completion of a set of upgrades, or with some other indicator (like noticing your opponent has begun to take an additional expansion and should therefore have a smaller army than you), but if your macro isn't on point, and you've got an accumulation of resources or you've been supply blocked at points, your timing attack will be weaker.
The problem with a commited attack is always the fact that you're (mostly) on their side of the map, meaning their units reinforce faster. If their army is smaller when you leave home, it may be equal or larger by the time you arive, with the newly produced units always adding into the fray (unless they mismanage their rallies). This is why an attack eventually loses its potency as it pushes deeper into enemy territory. Typically, people don't push to end for this reason; they push until they have a lead, and then attempt to maintain their lead.
At silver league, we can assume that there are some holes in the players' macro, which if they overcome, they will win more and rank up. Attacking will almost always strain your ability to macro (assuming you actually watch and attempt to control parts of the fight); simply rallying into defended positions without control will typically result in failure (density of damage output being reduced when units are spread).
The go-to advice most people would give is to focus on your macro. As you get better at growing your economy and army, you will find more time between those tasks to scout, attack, harass, etc. Every opponent is different, and no single philosophy or style simply beats everyone. Some people will not crack under constant pressure, and will simply grow their lead by defending perpetually.
|
i think the sentence you are looking for is 'a fighting unit immediately begins losing its value the second it is produced'
while true, this is not helpful in 99.999999999999% of in-game situations.
any conclusions you draw from this are completely set on other assumptions
|
Hyrule19202 Posts
It goes without saying that in a military RTS game you need to attack your enemy. So let's not say it.
|
|
|
|
|
|
EPT
