On Timing (A rant)

[BoB(MarLey)]

Today I felt like writing a little something about Warcraft and the topic I have chosen is timing as it pertains to UD, mainly for two reasons. The first is that Many replays I watch these days in threads like "what I did wrong?" often are a result of bad timing. The second is that timing seems to be less and less focused upon as an important concept; there is more to warcraft than simply micro and macro

The first thing to do is to define what timing is. The definition I have been running with is knowing when to fight and why. Imbalanced Video (a sort of espn sportscenter for Warcraft 3) had a really good video on this that seems to have disappeared with all their other great coverage when seed retired from doing the show. His idea of timing emphasized 5 major elements, which I will summarize and then add two I think are worth mentioning:

1. When the economies are unequal

2. When one party has a counter

3. Hero levels

4. Food count

5. If there is anything else worth doing

6. Critical mass

7. Upkeep


1. When the economies are unequal

Warcraft, like all RTSes, is a game of economy. If one player has an expansion while the other does not, that player will win in the long run. However, in order to get that expansion up, the player must make sacrifices in his army strength or in his tech. This leads to the conclusion that a player who is down in the economy game should be more aggressive in the short run. The HU v UD matchup is a perfect example of this. While the HU may opt for any number of ways to do it, he will fast expand 9 times out of 10. Since the UD cannot stop the expansion, he must try to slow it down by killing peasants. This leads to the UD creating timing; if he kills enough peasants, the HU will either sacrifice his army strength by ceasing production of units, or he will sacrifice his tech by not teching for some time. In the former case, the UD needs to dual crypt rush to take advantage of the lack of units/towers. In the latter, the UD should tech quickly with the goal of a destro push, but also be open to adapting to what the hu is doing. In both cases, the undead is gaining an idea of when he wants to fight.

2. When one party has a counter

A game of Warcraft can often be seen as a pingpong match; one player creates one army, the other creates a counter, the first transitions to counter the counter, and so on. It goes without saying that fighting with a counter is infinitely better than fighting when countered. Looking at the UD v NE matchup, once the UD has level 2, he has a counter. While archers can stand and fight vs ghouls without coil or aura, they cannot do such vs ghouls with both. This creates timing again; on maps with easy creeping, the UD can opt to invest heavily in ghouls, creep to 2 quickly, and hit the NE with the goal of damaging wells and archers, or delay t2 buildings. To adapt to this timing, most NE get a panda 2nd, which means the UD needs to adapt before continuing to push. Another adaptation for NE was to pick a summoning hero first, namely the BTM or DR. In both of these cases, a UD player may want to wait for t2 to pick up fiends and a Naga before engaging the army head on.

3. Hero levels

The one aspect that defines Warcraft 3 as unique is the hero unit. A unit that becomes more powerful as it increases in levels. Since the first hero is free for all races, you must get one in order to be in the best possible position. Further, the hero's abilities can be levelled as the hero levels. This creates timing right in the game mechanics; level 3 is a critical level in that your first ability can become level 2. So naturally, a push where one player has a level 3 hero and the other has a level 2 will be much more powerful than if the same player had a level 2 hero vs a level 1 hero of the same opponent. In the UD mirror matchup, where both players opt for fiend builds, one player builds his first zig a few seconds after the other because the other player sent an acolyte to block where he wanted to place the zig. Thus, his DK is slightly delayed, and he can no longer creep for 3 because he knows his opponent's DK will be 3 and begin to push his base while he's out creeping. The simple unorthodox move of sending a scout acolyte gave a good advantage to the second player in a matchup where any advantage can be game breaking.

On the same token, a player must be aware of the health and mana on his heroes. If the UD player with the level 3 DK crept recklessly, and spent all his coils healing fiends, the fact that he is 3 before his opponent is no longer an advantage since if he were to push, his fiends would have no coil, while the opposing player would at least have a level 1 coil. If the same player lets his DK tank rather than spend coils, he still cannot push since the DK may die under a manual targeted cold tower and fiend volleys.


4. Food count
While it is hard to look at an army and quickly figure out an exact food count, it is not wholly necessary. More experienced players can look at an army and roughly estimate if it is a 50 food army or an 80 food army and plan accordingly. More often than not, a player does not want to fight against a sizable food advantage; UD v NE where the UD is banking at 50 and the NE is piling on units to 70, the UD should not be fighting since his army is weaker now, but will be stronger later on.

5. If there is anything else worth doing.
Generally, if you have 5-5 heroes, the map has been crept out, and you have an 80 food army, it may be time to fight. If you have the advantage you want to force the fight on your terms by setting up an expansion and parking your army there. If you're at a disadvantage you must take the fight to your opponent, because he will be setting up that expand. This position in chess is known as zugzwang, where it would be better to not move at all, but you must nonetheless. In Warcraft, while you can sit and do nothing, that is a decision in of itself and the worst of two evils. This endgame was the result of earlier mistakes, and pushing is the better alternative than letting him expand.

6. Critical mass
What I mean by critical mass is the point where an army is at it's peak. Your army is at it's strongest point, and it will only get weaker comparatively to your opponent from this point on unless you do something. Generally this is a point you want to fight, although sometimes expanding may be a better option. One point to hammer in is that your critical mass may not be ideal, but it's your best chance in the game. Consider the NE v OR matchup. The elf is at 50/50 with master dotts and 3-1-1 dh/bm/tinker. The Orc is at 55/60 with raider walker and 3-3 bm/tc. Tc has a mana pot, bm has an invulnerable pot. This is not an ideal fight for the elf, it would have been bettter to catch the Orc before he hit upkeep or got those items. However, the elf must fight, because it will take time to build a moonwell and continue production and the difference between a 60 food dott army vs 60 food raider walker army and a 50 food dott army and a 50 food raider walker army actually hurts the elf. The extra food means more dispell and more effective link for the Orc, and a few more cyclones for the elf. In the ud v elf match up, critical mass is hit when the undead has 80 food abom wagon destroyer, if you decide to bank at 50. once you hit that 80 food, regardless if the elf has hit 80 food and is beginning to tower, you need to hit. Otherwise, the elf will expand and continue to add dryad bear, which benefits his army more than the extra aboms benefit you.

This also means waiting for food in production. I have seen so many games where players push at 70 food vs 60 food with an abom a stat and a fiend in production. This essentially means the player is really fighting 60 food v 60 food while in another minute, it will be a 10 food advantage. Wait that extra minute, and a fight will be so much easier since you aren't wasting gold on units that will not be there for the critical junctions.

7. Upkeep

This is ridiculously important to keep track of. As you are all aware, when you hit 51 food, you lose 3 gold to upkeep. When you hit 81 food, you lose 6 gold to upkeep. As such, planning what is to be done when is critical to managing your economy. If a UD player is playing an elf and decides to gargs to contain the elf before pushing into a late t3 army, he must make certain decisions about the size of his army. A UD who goes about this goal with a 51 food garg army as opposed to a UD who goes with a 49 food army, is going to have a much harder time trying to raise that t3 army later on due to the fact that he is losing gold for that 1 food. Therefore, learning when to break upkeep becomes a crucial skill. For containing an elf with gargoyles, a UD player wants to wait as long as possible before breaking his threshold into 80 and then 100 food.

Once you break upkeep, you should have a reason why, ie: the elf has expanded, therefore I need to maximize my force to take it down. You should also have enough gold to push right into 80 food, in order to make full advantage of the low upkeep threshold and be ready to fight at a moment's notice.

Finally, utilizing upkeep effectively can also mitigate expansions in certain situations. If any race takes an expansion and pushes to 60-70 food (all races do this vs undead at different points), if you can delay a fight until you jump to 80 food, you can often mitigate the expansion economically. Imagine a hu player who has FE'd, then pushed into low upkeep. He really wants to fight, but as the UD down an expansion, you do not. Therefore, you futz around the map, harassing his workers, taking down towers, doing whatever damage you can without a direct confrontation. Over time, he is continuing to add to his army, pushing into high upkeep. All during the time he was in low upkeep, he was mining +7 from his main, and +7 from his expo, a total of +14, while you mined +10. Then when he pushed into high upkeep, he began to mine +4 and +4, so he was effectively being outmined on one mine. Of course, the ability to mitigate an 80 food army with a 50 food one takes very good knowledge of where your opponent is, but that's another rant. The point is, if you can effectively stop fights when you are ahead of upkeep, even down an expansion, you can push into 80 food and still be as competitive.


This is a list of things to keep in the back of your head as you decide to fight. That is to say, no 2 players are the same, and they will create their own priorities as to how to decide what is most important to their timing. However, if you look at any replay, and try to figure out why a player is pushing when he does, odds are 3-4 of the things up there are in the players favor, and he has decided that the timing is good enough to hit.

Feel free to add what you know, this list is based on my experiences on what has been effective, and there may be some overlap, but i guarantee if you have no idea as to why you lost your last game when your micro and macro were perfect, your timings were probably off.


EDIT(S): just random grammatical stuff that was annoying me.

[Lazzarrath]

nice little guide. Timing is often overlooked so its cool to have a little reference page. Just thinking about the imbalanced series brings back a lot of memories

[primobaby]

nice

[Shayan]

good post

[thehunter99]

nice read.

[TheBeatles-]

will read

[BaronFel]

I approve, something everyone can work on is getting better at hitting timings!

[xebec]

this is higher quality than most of the front page articles

as I understand, zugzwang isn't technically a choice between two bad moves. it is the necessity of making a move when it would be better to make no move at all.

in a chess game, a player who has only one legal move (and thus no choice to make) would still be in zugzwang if this move would leave him worse off than were he not to move at all (which would be against the rules). this doesn't happen in wc3 because it is not turn based.

[BaronFel]

Well in WC3, not doing anything is bad (opponent sets up expo), attacking is bad (we've said that opponent may have some advantage of some sort). Thus, your only move should leave you "worse off" but that the lesser of the evils is fighting cause then you'll have more of a chance than allowing to give him more income to replenish faster

[vwavve]

I like to play ping pong

[Sarah_1988]

Very nice read, thank you very much!

[BoX]

excellent, this should be on the front page!

[Holcan]

i personally think this is my favourite aspect of warcraft 3. (timing, that is)


and baronfel is right, anyone who thinks they cannot improve anymore can always improve on their timing

[Vall3y]

good post. there are a few things I dont agree with I will comment on later

[xan]

Originally Posted by xebec

as I understand, zugzwang isn't technically a choice between two bad moves. it is the necessity of making a move when it would be better to make no move at all.

in a chess game, a player who has only one legal move (and thus no choice to make) would still be in zugzwang if this move would leave him worse off than were he not to move at all (which would be against the rules). this doesn't happen in wc3 because it is not turn based.

true

[gEcko-boY]

wait a second.....

i remember when uds used to counter human expo with an undead expo and sit in base running left and right to kill tanks..

now its all map dependant and ted fiends and destro pushes ;]

[BoB(MarLey)]

I'm glad you guys liked this . I might start doing a series. I really actually wanted to do these in AC form, but my mom broke her ankle a few weeks ago, so I haven't had the time/silence needed to make really good ACs.


@xebec, you are 100% right. My comparison between warcraft III and chess is a little hard because there is no waiting in warcraft III, but I consider the player down in the endgame to be in zugzwang because while he would much rather "pass" the situation and maintain a slight inequality, he must either A) fight now in an unfavorable position or B) do nothing and allow the expansion to get up, creating an even worse scenario. These are both terrible options.

@vwave, I like ping pong too. But I'm super noob .

@vall3y, I'd love to hear about your disagreements because its the only way to learn .

@gecko, yea those were terrible days. I sucked at stopping the third expansion from getting up. Now I can just push at t2/t3 and have hu's crying imbalance! YAYYYYYY!

[syrto]

well finally a nice post in the whole section or actually in the whole forum.

wish your mom fast recovering and hope for some more posts like this one.

would like to see arguments as well

[xebec]

Originally Posted by BoB(MarLey)

@xebec, you are 100% right. My comparison between warcraft III and chess is a little hard because there is no waiting in warcraft III, but I consider the player down in the endgame to be in zugzwang because while he would much rather "pass" the situation and maintain a slight inequality, he must either A) fight now in an unfavorable position or B) do nothing and allow the expansion to get up, creating an even worse scenario. These are both terrible options.

you know what would be kind of interesting? a chess program that wasn't turned based, ie, you could move as fast as you wanted to, as long as the move was legal at the time it was completed. Pieces could have movement speeds like in wc3, maybe a 1 second delay in which a just-moved piece cannot be moved again. Obviously it wouldn't work in real life because it would quickly turn into a fistfight, but it could work online.

I think one reason timing is so hard relative to other skills in wc3 is that it's much easier to diagnose problems in other areas of a player's game. For example, if someone's micro sucks, it tends to be immediately obvious, whereas if someone mistimes an attack slightly, the full effect of the mistake might not be apparent for quite a while, and then the loss will be attributed to other more obvious factors (eg, bad micro). thus some players never recognize the problem that needs to be fixed.

[BoB(MarLey)]

Originally Posted by xebec

you know what would be kind of interesting? a chess program that wasn't turned based, ie, you could move as fast as you wanted to, as long as the move was legal at the time it was completed. Pieces could have movement speeds like in wc3, maybe a 1 second delay in which a just-moved piece cannot be moved again. Obviously it wouldn't work in real life because it would quickly turn into a fistfight, but it could work online.

I think one reason timing is so hard relative to other skills in wc3 is that it's much easier to diagnose problems in other areas of a player's game. For example, if someone's micro sucks, it tends to be immediately obvious, whereas if someone mistimes an attack slightly, the full effect of the mistake might not be apparent for quite a while, and then the loss will be attributed to other more obvious factors (eg, bad micro). thus some players never recognize the problem that needs to be fixed.

Definitely, I think it was Rotterdam who said in some interview that the reason he didn't think much about micro was because 98% of all fights have already been won before they start based on earlier macro and timing decisions. It's a big reason as to why I have a hard time relating to players like Happy; when he first started playing everyone loved to see how he beat 80 food armies with 40. The problem is, for almost any other player that's a lost game. If you seek to emulate your game solely on micro, unless you're an absolute monster it's going to be a tough tough road.

I definitely like that idea about chess though. I always thought it would be cool to have an RTS that was essentially always a mirror match; no creeps, no heroes, just a map with even expands and two races exactly the same. It would be like chess because it would be perfectly balanced but it would require all the skills of an RTS game.

Of course replays would not really be entertaining but it would be an interesting game to see how it would evolve..