Page 1 of 1

High skill 1vs1 guide + an example game (medium+ 16x16 maps)

PostPosted: 06.11.2020, 05:30
by Dieego98
Hi. From some time we've been in need of players in the game, so I wanted to make a guide to help players around 1400 skill get to 1600+. For a bit of perspective, at the time of writing this (and for the past 2 years) the skill of the highest active player is ~1750 and there are ~5 active players with more than 1600, so the current skill cap for a very good player is at most 1800 (so skill in general is around 200 points lower than a few years ago).

I apologize in advance if some parts sound arrogant or maybe too detailed, english isn't my first language and I wil try to explain everything the best I can.

This talks about the mindset, heuristics (general gameplay), and concrete tips needed to win a 1vs1 game, in the form of a walk-through a usual game. It is very detailed so feel free to just skim over the parts that you like. The most useful stuff for experienced players is probably the defense tips. Those are mostly ideas from SoulMan and MJochen.

Special thanks to:

-SoulMan for his original defense and attack ideas, and general playstyle.
-NEb0 for his attacks and his aggressive gameplay, pushing income over lives always.
-501/dumboder2 for his low-saving gameplay and his agressiveness (attacking first, then see how to def whatever is on your screen)
-jvarg4do/jvarg5do for his general well rounded gamplay.
-MJochen for his defense knowledge.
-All the players that are still around, 0always down for playing a game :)
This is based on what I've learned on hundreds of 1vs1 games against them, and so it is a mix of the playstyle of those players and mine. A lot of this can also be applied to 4 players all vs all games, maybe with more care on lives and more pushing.

This is based on maps like Darkvector, Circlevector, Bubbles, Richtungs, Plox, Beringstrasse, Blueeye, Redcreep, Carbon, Blitzerino, and current tournament maps in general. Of course, different maps have different playstyle. This works well on current tournament maps and most fun maps (medium+ length)

This is also all my opinion and a description of my playstyle, so anything that says you should or shouldn't do isn't the absolute truth nor a critique of other styles, just my observations of what works and what doesn't while using this playstyle.

If anyone has suggestions or thinks something is wrong or can be improved, it will be appreciated.


-----


BASICS

Most 1vs1 games in medium+ maps with a similar skill opponent end in a full def endgame, exception being if one of the players screws up very badly and dies earlier. With good play on both sides there will be full def.

The objective of the game becomes then to get enough money to send the final attack on your opponent before he had enough to send theirs.

There's a few ways to do this:
-Have a lot more income (40%)
-Have a bit more income (15%-20%) and similar lives (+-5)
-Have the same income and significantly more lives (7+)
-Have the same income and lives, and attack first
-Have less income, but a lot more money in hand (saving for final attack for more time than the opponent)

For having the first one the opponent has to play poorly. For having the last one you have to play poorly. If both play well, it will probably end up in one of the middle cases. One must keep in mind then that strong attacks and good, cheap defense is just a means to get more income or take a significant amount of lives.

To achieve this, here's a few rules of thumb that must be followed at all times during the game:

-Do. Not. Save.
Except in counted occasions, this is the single main mistake a lot of players make. Saving kills your income, period. This doesn't mean "save a little", it means do not save at all except on those counted occasions. You have $800 inc? Don't save ~$300 to send 2 demeters next round, just send the $800 and wait one more round to send the demeters. Most of the time the money you get from killing creeps will be enough to make up the difference without saving. If you really need to send early, you can sell something that won't be used in a while (an extra slow for example)

-The answer to "when to def a strong atrack and when to let go?" is in general that if your opponent saved a significant amount (more than ~30% of the inc) to send the attack, it should be let go, except on counted occasions that will be noted below. If your opponent starts saving regularly, you will out-inc him fast and it will be an easy win for you.

-Sell little or nothing. There's 2 problems with selling: the one that everyone knows, you lose 25% of the tower's value.
I won't get into the math details, but assuming that you sell to send first row creeps (10% inc) that means that:
If you rebuild 1-2 rounds after selling, you lose money.
If you rebuild 3 rounds after selling, you break even.
If you rebuild 4+ rounds after selling, you win money.
I haven't done the math in the case you send stronger creeps (9%-8%-7% inc), but it probably is one more round needed (4 rounds to break even, 5+ to win money)
This also implies that killing creeps that pass trough your def with low hp, then selling those towers, is usually not worth it. Just let them go and build your towers in good spots.
Okay so the second problem: you have an empty screen. Is it good to sell? It depends on your actual income and your plans on combos on the next rounds, because you will have to rebuild and that will stop any combo that you are sending. That lead us to our fourth and last point...

-The decision making during the game depends, in my opinion, in what will you be able to send in the next few rounds. Say you get two raiders at 3300 income because your opponent saved, and you need to up one tesla to blue to def them. Consider that your opponent just made $720 inc sending those raiders, maybe he even sold something to send them. If you up that tesla he will have all that advantage on inc for at least 4 rounds, the time it takes you to get to up the tesla, then push to ~3800 inc, then send your pair of raiders, and then they have to get to his main def. That's a lot of time. If you don't up the tower, his saving doesn't hurt your plans (pushing for 1-2 more rounds then sending raiders). That is the third point, don't let the opponent's actions affect your attacks. You have $2000 in hand and $1850 inc, enough to send 2 rays, but a demeter just escaped with red life your main def. You will need and extra tesla or lots of guns to kill it. It isn't worth to do it, because you won't be able to send your rays. The same situation if you had $1500 in hand and $1500 inc, it would be better to kill it because that doesn't hurt your plans. The same has to be even more calculated for combos, because when you send a combo your money will be in use for even more time (no money to upgrade your def).

Okay so, this is all heuristics, let's hop on to the actual game, with concrete tips. This example game assumes that the opponent doesn't save or saves in few occasions. Also, the following assumes that your towera are built in the correct spots. If you build on bad spots (specially your green and white teslas), it doesn't matter how well you play, the opponent will have a big advantage. The best spots on most common maps can be copied from other players, but also there's vector vision if you have pro or use NEb0's client (more on that at the end).


EARLY GAME

Depending on map length and opponent's playstyle, there are a few valid starts (examples of maps where they are useful in parentheses):

-Mako/2 mercs at second 7-15 (depending on maplength) then 2 close novas. (Darkvector, Bubbles, Richtungs)
-3xmako at second 7-15, one mercury a few seconds after, and a nova as soon as you can. (Beringstrasse, Plox)
-4xmako, not that close to kill them easily with cannons, nor that separated to kill them easily with guns. (Redcreep)
-3xmercury a bit spaced, then a few seconds after one nova. (Blueeye)

Note that the first 2 use 2 rounds of income, and the last 2 just 1, useful the shorter medium maps.

Also note that you save $50 on the first 2. This is one of the exceptions to the no-saving-rule (But it's just $50!! Does it matter that much? Well yeah, $50 is 1/4 of your inc at that point)

If we both send the mako+2xnova case, and the opponent skips all 3 of them, usually I do the same. This might not be optimal, but the slight income lead he gets if I def and he doesn't makes for a uncomfortable early game for me.

Depending on map, a good def vs those are:

-2 green cannons / lots of green guns for 4 makos (depending on separation)

-Green guns vs 3xmerc+nova. A lot of them if you want to kill the nova, maybe a slow and/or a blue one. Defending all is expensive, depending on map I usually just put 3 guns on O and let go the nova. If they are too close one green cannon on O is the best.

-1 green cannon and a few green/blue guns vs 3xmako+merc+nova. No slow, 2 extra guns are worth more at this point. Guns should be on P then 0

Note on good def #1:
From 1300+ skill most people do it, but it's still worth noting. Defense vs more than one creep is more efficient if you use the P setting. The reason for that is that while on 0 your towers kill one creep first, on P more creeps are alive for more time, that means that creeps are in range of your towers for more time, and then the towers are being used for more time (as they have more stuff to hit). A usual trick is to put your def on P, then when the first creeps has red life and is about to exit the tower's range, put your def on 0 so they kill the first creep and they have more time to kill the second (that is still in range).

-1 slow, 2 blue guns and 2-3 green guns (or another blue, depends on map) vs mako+2xnova. But that... Doesn't kill 2 novas right? That leads us to...

Note on good def #2:
This is a trick that SoulMan discovered, maybe it was known beforehand but the first person I saw using it is him.
The manipulation of the creeps using slows is important during the whole game, and this is a way to separate creeps, different to the most common ones: the slow goes on P and the defense goes on 0. If done right, the slow will hit just once the first creep then only the second creep, allowing for the first one to go ahead and get separated. This has 2 advantages over other ways to separate the creeps: first is that if the creeps are already a bit separated, this way "wins" the space between them, meaning that the creep that is behind gets more behind (compare to using 9 where the first creep is always the one that gets behind in the end). The other advantage is that it can be done quite efficiently with only one green slow on the main def, that means it is cheap and can be used in the early game. Examples of maps where this works very well vs 2 novas are Blueeye and Beringstrasse, but it is really a trick good on every map and usable during most of the game, not only at the start. In the particular situation of mako+2xnova it can need an extra green gun if the mako is well timed (arrives to the def a bit before the novas) or if they are 2 mercs instead of a mako.

Okay so after you start and def the first wave, next attacks will probably be poorer for some time. No need for extra def. A try to make the opponent build extra def can be made with some combination of lots of makos and 1-2 novas after them, but it is hard to pull it off without saving. You should never be sending 3 novas, not because it's bad but because the math doesn't allow it without saving, if you have near $500 inc it's better to send mantas.

Here comes the first combo: 2xmanta + ?xnova. As all attacks should be, it depends on your opponent's def. Let's say you're playing Darkvector and your opponent has only a blue cannon and 1-2 slows. If we wait 30 seconds to send 4 novas, the mantas get alone to the defense, and the 4 novas will die easily to the blue cannon. Opponent will need maybe just one more blue gun to kill the whole combo, so it's better (in my opinion) to send 2xmanta + 2xnova + 2xnova, with 15 seconds difference between each. If for some reason your opponent has lots of green/blue basics and no cannon or a green cannon, it's better to send 2xmanta + 4xnova.

On the other side, to def that, depending on the map they can be used 2 blue cannons, 1 red cannon, or 1 blue cannon and some blue/green guns (all with slows)

After this is mostly repeating the same until about $850-$900 inc, then you should get enough money from killing creeps to send 2 demeters without saving. To defend those, a blue cannon + green tesla should be enough. On some maps a red cannon is required.

You can join some novas 15 seconds after the demeters, depending on map it might be useful. If you get that, remember that slows can go on all settings just like the other towers, not only L :D. Don't be afraid to switch slows to P or even 0 sometimes to kill combos. Even mercuries are faster than demeters and take priority for your slows if they are on L.

Now you start having stronger creeps available, but that is a double edged sword...


MIDDLE GAME

You must now find a balance between making enough pressure that the opponent doesn't sell, and also push our income high. A good way to do this is to send repeated 2x(strong creep) only if he skips the first 2, or when you have almost the income to send the next strong creep. This makes your opponent unable to sell part of his defense to send the next strong creep early. Example: send 2xray at $2k inc, then 2xray again at ~$3.4k inc, so he can't sell one of his green teslas and send 2xraider earlier than you. Everything in between must be first row creeps in order to build inc.

Ok back to demeters, a nice combo to send at around $1400 inc is:
-Money from kills gets you at least $100
-4xmanta ($1.500 + $500)
-15 seconds after, 2xdemeter ($1.000+$1.000)
-15 seconds after, novas (~$500+~$1750)
Depending on map this can be too much for a tesla + blue or even red cannon to handle alone, and should leave you with $2k inc, enough to send 2 rays.

About sending 2 rays, or 2 white creeps in general, this is also one of the exceptions of the no saving rule: if you have $1.850 inc and $2.000 money in hand, it is useful to sell something or save $150 from kills in order to send the 2 rays one after the other, without waiting for the 1s delay of getting the extra money from the income. The reason is that 2 creeps very close together can need up to ~25% more dps to be killed than 2 creeps separated by 1-2 seconds*, so saving that little money ($150-$200 for rays, $2.000 for sharks, $15.000 for phoenixes) can actually make a difference on how much the opponent needs to def. The same can't be said on other creeps, because they can be manipulated with slows to be separated or joined.
*Yeah, even taking into account splash damage. Unless your opponent's def is 70% splash, something like 2 white cannons and 1 green tesla, then yeah it would be better to send them very separated.

Def vs 2 rays is in general 3 green teslas and one red/white cannon. If you have just a red cannon and a green tesla before you get them, consider just upgrading the tesla. It may be cheaper and can defend the raiders that will come soon.

To defend 2 raiders you need usually one blue tesla and one red cannon, and maybe a white cannon instead of red and or an extra green tesla. Blues or even one red slow. Also, you need to separate them, ideally before they get to your main def. In some maps (Bubbles, Darkvector) they can be deffed with only 3 green teslas and one red/white cannon, in case they are very separated)

Note on good def #3:
The usual 2 creep separation can be done in 2 ways: first is, as soon as you can, put one slow that will hit only the back creep of the pair. Then another one, and another, etc. Ideally sell them all before the current round ends, to recover the $ lost as soon as possible. If done repeteadly it can be expensive at raiders stage, with only $4k inc.
The second way is to put a blue or red slow on a turn at the start of the map (second circle on Circlevector, first corner on Darkvector, U turn on Monoblue, etc.). This slow should be on P (depends on what you wanna separate, but it's almost always the stronger creeps). When the first creep is about to get in range, use the 9 (lock) feature. That makes the first creep that gets in range be the only one slowed, and the other one runs free. Most of the time this can be used first, then use the other method to get them even more separated. Keep in mind that the first slow should be created before the creeps appear, because it takes 8-10 seconds to create, upgrade, put on P, and then put on 9 the tower. P may not be needed if there aren't more creeps around or if the creeps you wanna separate are blue ones. I like to build it at ~3k inc.
The first method is more flexible, and you can screw up more without consequence, but it is also harder and more expensive in the long run (buying and selling all those slows matters at around 4k inc)
The second method however is a one time chance, so if you screw up you will have to fight the creeps together. In that case it might be better to just put the towers on O and let one go, or if you have a well placed slow on your main def, use the slow on P + def on 0 trick to separate them there.

After raiders it's just push inc and only send if you see your opponent sell. If he has very little def and managed well the rays, a well timed ray + 2xraider + ray attack at $6k inc can make him build a bit more.

At ~$7.200 inc you should send the first toucans (no saving, just money from dead creeps and/or selling slows or a bad placed green tesla for example)

Same as with mantas, whether to send raiders 15s after or 30s after toucans should be based on the opponent's defense.

To defend this combo it can be used a red tesla, white cannon, red slows, and some green teslas if needed (always vs separated toucans) or maybe 2 blue teslas, a green one, a blue rocket and a cannon, depending on map and if he sends 2 raiders or 4. If he sends 4 together then a blue rocket (or 3 white cannons) are a must-have (join the raiders together by putting some slows on 0).

At this stage you might be tempted to build slows on the whole map. For this style of play, it is a mistake: you want to have full control of what is on your screen and what gets first to your defense. That means that you slow only the creeps that you want slowed (for example if you get a vulture at $25k inc, it should advance fast in case you get sent 2 sharks next. The vult shouldn't bother you while you are trying to kill the sharks! Another example is one shark and a bunch of raiders, you would want the raiders to get to the def before the shark so they are killed without being a problem).

Around $10k-$12k income I put the three white cannons at the start to keep dead the mantas your opponent pushes. If you don't care about the mess, it can be done later. There's also some slow manipulation possibilities if you have mantas all over the screen (switching to O instead of selling, so you can have slows in the whole screen without selling). I don't think it's worth the mess, but you may think otherwise.

Do not save to send a single shark at $12.5k that your opponent will skip easily.

Anywhere around $8k-$12k income you might get a vulture + loads of raiders (opp saved to send the vult). This is one of the occasions where it is worth to kill the vulture, even if the opponent saved for it. The reason for that is that if you let it go and the opponent joins another one, you will have a hard time defending them due to the regeneration, and at this point of the game you wanna be comfortable in order to do the following steps...

We are getting closer to shark money so our play should be very certain here, this part of the game is on my opinion critical. At ~$14k inc you should send 2xvulture, again without saving (as always), just money from kills or selling slows. Depending on map and the opponent's def it might be useful to send 3-4 raiders 15seconds after the vults.

To def the vultures that you'll get at that point you want to use a white tesla. If you separate them, you shouldn't have any issue with only a white tesla and red slows, even if the opponent sends raiders after the vults (this is a good chance to sell everything else that you still have and send mantas: blue rockets, white cannons, blue teslas unless they are placed in a spot where you can upgrade them to white and have good range). You could probably def them for cheaper with a mix of 1-2 red teslas, 1-2 blue teslas, maybe a blue rocket, etc, but it is important at this point to have a white tesla built. Also it is important to build it after you send your vultures, otherwise it is quite expensive. The important thing of having a white tesla is that it can damage sharks and rapidly kill anything weaker than them.

Right after you build your tesla you should have $15k -$17k inc and no money in hand. Next round is another saving situation, we save the whole round in order to send 2 close sharks the next one. After that, we push. No more def, your white tesla is enough, no raiders or stuff, just push mantas. We have 2 possible situations here: either your opponent deffs the 2 sharks and you get a comfortable lead of around 5% inc, or he lets one (or maybe both) go. If we are in the second situation, you need to send 2 extra sharks to join the one that passes (else you just saved for nothing), and for doing that without saving you will need $23k inc. If the map is too long and you have $23k already but notice that you need to wait one more round until the first shark passes, one option is to send 3xtoucan. If timed correctlythey will arrive at the main def a bit before the 3 sharks, and they will be a problem bigger than it seems at first sight.

So to recap this part of the game is
$13k-$14k inc send 2xvult+raiders
$15-$17k inc upgrade red tesla to white
$15-$17k inc save one round and send 2xshark
$19-$22k inc push / 3xtoucan
$22-$24k inc send 2xshark, preferably joining the first one(s)

In that serie of actions you should get a slight lead on inc (unless your opponent does the same of course). This works on most maps except very long ones where opp can just let go the first 2 sharks and push (Richtungs) or maps that have amazing spots for rockets, because opp can just let go all sharks, maybe 2 times even, and then def all with rockets (Bubbles, maybe Redcreep or Monoblue, tough I still do it on those two).

After this, you should probably dedicate a couple rounds to def all the sharks that have been accumulating on your screen :) 2 white teslas and 1-2 red ones are enough in most maps (remember to attack on P).

Mambas, sent at ~$37k inc, are one of the easier creeps to defend. If your def can kill 2 sharks, then you probably need just one upgrade of a tesla from red to white, and maybe in harder maps one extra red tesla to kill 2 separated mambas. Key word: separated, as much as possible.

There exists a nice combo that can be done at exactly $38k+ inc. anything less than that and you will need to sell stuff for it to work. I haven't used it in a while in medium+ maps because I like more to push at this stage and don't like the idea of saving another full round, but it will be noted here because it might fit some players' playstyle better. It is also an excellent finisher on short (not very short) maps.
Save one round at $38k inc, then send one titan. At what second you send it depends on the map length and the slows of the opponent. On short maps (Keepoffthegrass, Limit, Hotround) always on second 1, on medium maps (Plox, Redcreep) on second 1 if opp doesn't have slows, second 15 if opp has full red slows, and somewhere in between if opp has only some slows placed, or green, etc. This, again, is why it is important that you use slows to control the pace at which the creeps get to your main defense, so the opponent can't time his attacks this way. 15-30 seconds after the titan send 2xmamba, and then 3xshark. Depeding on the initial inc ($38k vs $38.5k) you may need to sell a blue tesla or some cannons. You will then have around $55k - $58k inc.

Okay back to the game, we send mantas, then the game becomes push party for a few rounds. If the opponent is sending hard here, just pressing 4 can give you up to a 10% lead on income for the rest of the game.

Some people like to save one round at $40k inc to send an early 2 titans combo. A good strat against that is to just push to $55k-$60k inc and meanwhile defend: separate titans, separate mambas if they are 2 or join them if they are 3, create loads (~5-6) of white teslas if they are 2 mambas or some teslas and 2 white rockets if they are 3 mambas, teslas on P and 0 when it is needed, rockets L or O if it is needed to kill a big bunch of sharks. Consider upgrading slows to white if you need more time to push. After you kill the combo the situation is the following:
The opponent saved and lost inc.
The opponent sent strong and lost inc.
You pushed and won inc.
You bullt about 2-3 rounds of def and lost a lot of inc.
The final situation of all that is usually the opponent has some more inc (~10%) but now you have a great defense that can handle whatever comes until the $100k combo, and that means you are free to send a $60k combo stronger than his: 2xtitan, 1xshark+3xmamba, 3xshark. After you send it he will have to def a lot and you will be the one with an inc advantage.
Edit: after testing this for a while, it seems that the opponent gets advantage if you do it. A simple counter that cements your income advantage is to just go back to the basics: as the opponent is saving, just put all def to O, leak everything your current def can't handle (~3 white teslas + 1 red, blue/red rockets), push, and send your titan combo before he can join another combo to the first one that goes through your def.

The other situation is that the opponent doesn't save at $40k and you both keep pushing. Then a good attack (depending on opponent's def) is 4xshark at $50k inc. This makes him build a bit more than the weak def he had to use for mambas. Or just push, both are valid. In this situation it becomes important to send the titan combo first, or at least in the same round as your opponent, even if that means selling something to send titans at ~$53k inc.

After sending the combo, you should have $85-$90k inc. You can send one or a couple titans (and push the rest) to make up for the income needed to get to $100k inc, then 2xzeus and after that mambas then sharks. At this income you should create/up white slows on the whole map as default, and sell them if it is needed to break some combo. The slows position is not random at all and the spots should be chosen carefully (more on that later).
To kill the first or second zeus combo it is useful to build 2 white ions, as they will stay there and help kill eventual early phoenixes.

On longer maps, consider sending titans or zeuses twice for the first combo (2xtitan, 2xtitan, mambas, sharks). It makes for a stronger combo at the cost of being 30s slower. It is very useful if the opponent can manage well his slows (he can break a combo that has only one pair of titans/zeuses). If you need time because he only sent 1 pair and you sent 2, up your slows to white.


ENDGAME

The following is valid only if you still have at least 6-7 lives. If you have less than that you will need some kind of trick to win (unless the opponent is low on lives too)

Attack from $100k to $800k inc:
After $100k income, all the pushing stops. With the exception of some very long maps where it may be better to push toucans (Richtungs), you should only be sending 3rd and 4th row creeps. The main thing you should send now is titans and zeuses. We want to spam the opponent' screen with red and green.
$120k inc? 2xtitan+1xzeus
$160k inc? zeus+titan twice in a row, etc.
If the opponent sells, just send more mambas and sharks for a strong combo.
If the opponent saves at $130k for 2 phoenixes, kill one and hurt the other. It is easy to do if you don't have zeuses or titans on your screen.
If the opponent is very low on lives (2-3 for DV, CV, Bubbles, 5-6 for Beringstrasse, Blueeye), it can be worth to save and send 2xphoenix at 130k inc, and join more if the opponent lets them go.
Keep sending titans and zeuses, phoenixes only if the opponent' screen is too empty and he could easily sell.
Don't save at $270k inc to send 2 raptors unless opp has 1-2 lives. This is a big mistake, raptors are too easy to skip in my opinion and at this stage a lot of income is created.
At $350k inc (with your opponent's screen full of titan and zeuses), sell some ions/teslas and send 2xraptor. If the opponent is very low on lives, the raptors should be followed 15 seconds after by 2xphoenix. Else, they should be followed 30 seconds later by another pair of raptors. The phoenixes are more likely to take lives in the short term, but the double raptors make him build more def and you get a more tangible inc advantage. Sending a single pair of raptors is weak, as the opp can just put all on L or 0 and kill just one and hurt the other.
After this, any money that you have left after sending strong, should be spent on titans and zeuses.
At $600k inc, send 1xphoenix+2xraptor then 4xphoenix next round. Don't save $200k to send 3 raptors.
At $700-$750k inc, sell a few ions and send 4xraptor (2+2 spaced by ~3 seconds or 4 together, depending on your opponent's quantity of rockets).
With $800k inc we should make the decision on how we'll make our final attack.
If at some point of this stage you see toucans pushing on your screen, sell ions and attack, and keep attacking (even if it means not defending at all 2 phoenixes or 2 raptors that your opponent sent when he saw you sell). Build a nice income advantage, then wait, because he will sell at some point and attack hard. If for example you have $600k inc (from selling when you saw toucans) and opp has $400, you're winning $200k extra for every round he waits to kill your attack and sell his towers.

Defense from $100k to $800k inc:
By $160k you should have, depending on map, 6-12 teslas, 6-12 rockets, and 2-3 white ions. Green ions are good only when you need damage fast, like to catch a shark that is about to exit your screen with red hp.
For 2 phoenixes, 5-6 white ions plus your teslas and slows should be enough. Scale accordingly if more phoenixes come after them (2 extra after 30 seconds means you need more def because your ions will be still hitting the first ones)
Don't be afraid to defend a lot while you're sending titans and zeuses, as it doesn't hurt your combo (sending 2 titans and 3 zeuses is not too different than sending 1 titan and 3 zeuses and building an extra white ion).
At around $300k inc, start building green and blue rockets on L at the first half of the map and/or strategic spots on the second half. They will be needed at around $600k inc but there you will be too busy on attacking and upgrading ions to make and upgrade 50 rockets on time.
If at $400k inc you get 2xraptor + 2xphoenix and want to kill only one raptor and the phoenixes, you can put all on 0, kill the raptor, then sell the slows in front of the second raptor so he leaks fast, and your towers start attacking the phoenixes. This is easier done with blue creeps as they are fast, but it can be expanded to every slowable creep, for example if you get a single early titan and lots of sharks, you can attack the sharks on P if you let go the titan by selling slows.
If you get loads of zeuses and titans, you will need to kill them with rockets. Remember that titans have 1.2M hp, zeuses 1.4M hp +10k/second regeneration, and phoenixes 2.5M hp (for comparison). Zeuses and titans are really strong. So what we do with them is kill them all at once... For that we will make use of our last slow manipulation trick: compression**.
If you get 3 raptors at $500-$600k inc, kill just 2. Attacking and creating more inc is more important now, the one who gets first to $800k inc has the advantage in the endgame.
At $600k inc it's still simpler to build ions, but at $800k inc (def vs 4 raptors) it is easier to compress them and have lots of rockets.

**Note on good def #4:
Compressing creeps is the contrary of separating them, we want to get them all together so rockets do maximum damage.
Scenario 1: suppose we have an empty screen and the opponent sends 10 zeuses and 6 titans, one at a time during 30 seconds, then stops. We have a thick line of creeps, and we want the back ones to catch the first ones. How do we do that? We start deleting all the slows that the first creep in the line already passed, and we put the slow that is currently slowing the first creep on 0, so it actually slows the first creep and not the back ones. After the last creep passes, we start building slows again, and in the meantime the non-slowed creeps catch the first one(s) that is(are) slowed. If done correctly, by the time the zeuses and titans get to our main def, we will have just a bunch of 10xzeus+6xtitan, easy to kill with P rockets (and P teslas/ions if you have them).
Scenario 2: end game attack. The same can be done with a stream of raptors, and it is very powerful. In Bubbles on a full map of only rockets I have deffed 100 raptors in sandbox, all rockets P and a full compression. The longer and more space/stops the map has, the more efficient it is to compress: Richtungs, Bubbles, Monoblue, Verlaufen, etc are some examples, but it works on shorter maps too like Plox (lots of space to build). A 8M attack can be defended on Monoblue, and (!!) they will die on the stop.
For this to work rockets should be on L so they attack the first raptor of the bunch, then switch to P when they are all together, and be ready to switch again to L if the opponent sends a mothership, or more raptors after you rebuilt your slows. Edit: turns out, rockets on L hit the raptors that are behind the group, as they aren't being slowed. A mix of 0 and P is better, or even K if they are phoenixes in front of the raptors.
Making 2 groups is very hard and it's best to just wait more time to rebuild the slows and make 1 big group. On a good Darkvector compression I usually sell 9-10 slows, and the creeps get fully compressed only at the middle of the map.
This is why the position of the slows matters, you want to keep two things in mind when you put them: first, when you put slows on 0 they should hit the first creep for a long time, so they are better placed at the loops (Circlevector center of the circles, Monoblue U turn, Richtungs first change of direction, etc). And second, when you delete one slow, you don't want any other to be hitting the same path. Slows should be placed such that they only hit one spot of the map, not two separated. Example: Bubbles first slows should be at the outside of the first line of path, not at the inside because they hit two paths. Some times it is impossible to place them only hitting one way (Verlaufen), then you'll have to rebuild them. On the games on my profile there are some examples of slows placed to compress, just pick any game on a tournament medium+ map. Also, thanks again to SoulMan, as most of the slow positionning on tricky maps are ideas by him that I just copied (for example Verlaufen).
A simple counter to compression is to send a few phoenixes splattered between the raptors. This will make the slows on 0 not hit the raptors and the try to compress will be a mess, unless we compress them the hard way (idea also by SoulMan): slows go on L, and when the first raptor is about to get on range, we toggle 9 (lock), so the slow keeps hitting the first raptor of the bunch, and ignores the phoenixes that could be in front of it. Then we sell it, and do the same with the next. It is hard to pull it off and easy to fuck up, but with enough practice it is doable on long enough maps or maps with enough loops. For example on DV or Plox it is more probably to fuck up by trying to use 9, but Richtungs, Bubbles, and CV are easy to do.
If you compress well your opponent is forced to sending creeps slower, but that gives you time to attack him too :)


FINAL ATTACK

We're back to the start now, planning our final attack. The whole objective of the game was to get here in one of the situations described above. We will go through each of those situations, but this part are just suggestions and nothing is set on stone, because you just need one chance here to win. Opportunity > plan.

How to attack vs each def, if the opponent has 10 lives (all this depends on map of course, they are just somme general guidelines):

Full rockets - at $5M start sending raptors slow, one every 5 seconds or so. If he starts selling slows and compressing, keep the pace. If he keeps the slows, you can send a big bunch of raptors at the middle of the attack, then a few more. The bunch is there because he can rebuild some rockets into ions and easily kill a too slow attack. The raptors in front don't allow him to use L to kill the bunch, the ones in the back don't allow to use P, and the first ones are too advanced already to try to compress them with the bunch.
It is never a good choice to be full rockets in 1vs1 in my opinion, because you tell the opponent what you wanna do.

75% rockets 25% ions - almost the same as before, but the first raptors should be less separated, and start sending at $6-$7M inc. A few phoenixes first and in between are useful too.
This is a good choice if you wanna compress, some maps where this works well are Bubbles, Richtungs, Bluevectorwater, Verlaufen, Stones, Monoblue. Some of them may need even more rockets, but it is always good to have some ions to kill the phoenixes that are faster than the raptors.

50% rockets 50% ions - this is more expensive, but depending on map it can be very strong, the best overall def. You can't send a big bunch of raptors because it will be compressed and killed with all on P. You can't send too spaced because ions will destroy the attack. So you need to do a mix, but that needs a lot of money: at around $8M send first 2-3 phoenixes so no easy compression, then some raptors in a "spaced bunch" (in a bunch, but not that close that rockets will damage 100% to all of them), then another bunch, always with some phoenixes in between, and a third bunch in the end. About 7s separation between each bunch.
This is a valid but expensive choice on most maps, specially useful on Plox, Darkvector, Circlevector, Copy, Blueeye.

25% rockets 75% ions - as the opponent doesn't have enough rockets to kill the raptors even compressed, just pressing C with enough money will kill him. How much you send depends on the map length, some maps can defend only 6 raptors with ions on 0, others up to 10 or more.
This def should only be used if you have lots of lives but less inc or less money in hand: build the whole def, kill whatever you can, sell a lot and send your attack (with more money as you saved for longer) and then lose ~10 lives to the first attack of your opponent. You will then have a load of raptors on your screen, but he will too and your attack wil be first. The 25% rockets are only there in case he sends too many phoenixes.

Full ions - expensive and easy to send vs it, press X to send phoenixes and win. Don't understimate the ions, a bunch of $8M phoenixes (40 phoenixes) will be fully killed with ions on 0 on Darkvector, and $9M will be killed with ions on P. The real advantage of this def (for the player that doesn't have it) is that it is really expensive, on Darkvector it costs $18M to go full ions.

A note on map space and rockets: if the map is medium but on the shorter end, and/or there isn't enough good spots for rockets (Headquarters), you should just press C and attack before your opponent does. If the full map with only rockets can't kill a single raptor in sandbox (Headquarters, Virus needs 2 raptors spaced by 2 seconds), there is no problem in sending the creeps together. On the def side of this, they should be built 90% ions and some rockets at the start on K, only to stop an eventual titan or even shark spam.

Some very long maps need motherships to make space for raptors, in case the opponent can compress well. The only current tournament map that I can def a good $8M-$10M attack without mships is Verlaufen, but there are others non tournament that also need motherships to be killed. Also consider sending a mothership after your main attack if you see your opponent is defending all on P, to force him to change on L or 0.

If the opponent has 16-20 lives a lot of what comes now isn't valid, because you'll need around 1.1M inc to kill them

How to win in the situations described at the start:

-Have a lot more income (40%)
This one is when you heavily outplay your opponent and it should be enough to not screw up the final attack. Beware of the compression of the opponent, and send slower if needed. You should save around $800k-$900k inc, while the opponent has $500-$600k inc

-Have a bit more income (15%-20%) and similar lives (+-5)
This is a very common scenario. Assuming you have around the same def, you should start saving at $800k inc (while opp has ~$700k). This gives you a small advantage but enough to win if played right. A good rule if you have more income is to just copy the opponent. He builds half rockets half ions? Do the same. He doesn't build anything extra and just goes for a faster attack? Do the same, but yours will be stronger with your greater income. He upgrades only his rockets? Same, etc.
Note that you will need to be precise with the attack timing here, because if for example a 6M attack is enough, and you go to 7M to be sure, your opp could send 6M at the same time as you, then the end result can become luck or random or who sent first.

-Have the same income and significantly more lives (7+)
You have to use your life advantage, that means focusing on the defense and just save a lot to send a stronger attack, even if the opp attacks first. A good choice is 50% rockets 50% ions, compress the first bunch of raptors with rockets on L or 0, ions on P and then 0, and aim to kill only most of that bunch of raptors. Send when you feel you have more than enough money to kill the opp, except if you see him selling when he is attacking, then you probably need to attack instantly because that selling can make up for the life difference and finish you.

-Have the same income and lives, and attack first
First attack or best defense wins. Consider selling your ions (or everything except slows) to have enough money to be first. Only try to compress if you are sure you can kill all, because if you sell slows the enemy creeps get to your end faster, so it is the same as if you had attacked second.

-Have less income, but a lot more money in hand (saving for final attack for more time than the opponent)
This should be done if you are being outplayed (10%-15% less inc) and the opponent has low lives. Start saving at $700k or even $600k if the map isn't long, then defend all the stuff the opponent will keep sending, and then do a good attack when he is at the middle of his saving (but doesn't have enough money in hand to counter attack) by selling. This is a last resource, and if the opponent can kill this attack then you're dead.

A final situation can arise when you have a lot less income. It isn't ideal of course, but it's another last resource trick. Send a final attack in the form of mainly phoenixes, not like pressing X but close in general, and a few raptors before and after them. The enemy rockets on P will attack the raptors, and if he doesn't defend well you can take 15+ lives with this method. It is more useful vs a balanced def, too many ions or too many rockets can kill the phoenixes. Consider selling to send them (remember that phoenixes are faster than any raptor you could be getting), and remember to send a raptor or mship after 15 seconds and maybe another one after 15 more, to screw up his P defense.


-----


That is everything I currently know about 1vs1 games in medium+ 16x16 maps.

All this can be seen in practice in the gameplay videos in SoulMan's Youtube channel, there are loads of 1vs1 recorded there.

As a final piece of advice, watch replays. I didn't start seeing my mistakes and learning from the other players until I watched replays of losts 1vs1 matches. The mistakes and the parts when the opponent got a lead on income are very clear on a replay. I use NEb0's client to analyze the games because in the replays you can see the income, money in hand, and tower settings of all players. The client allows to use pro features without actually having pro, as it can't be currently bought (choose playing position, record and watch replays, different sounds and graphics for towers and creeps, vector vision - no veto). It is somewhere in the forums for download.

Thanks for reading, gl&hf!

Re: High skill 1vs1 guide + an example game (medium+ 16x16 m

PostPosted: 11.11.2020, 12:56
by Seraph07
Good guide. Maybe mark your topics in bold so you can access them easier.