How many moves ahead...

[dab00z]

How many moves ahead do you guys think?

I always have a "general game plan" as to how I'm going to beat the guy, IE rush his cleric or something of that nature, but when I'm playing I have trouble thinking more than 2 moves in advance within the time constraints during early or mid game. This is because there are so many possible things to anticipate, and I lack the experience to rule out irrelevant considerations without missing some possible good moves by them. I also always try to think what their general game plan will be and work to position myself to counter it early on.

In the end game I find it a lot easier to pretty much anticipate how the whole play will go because there will be less factors and thus less things to consider. The only thing I find myself really missing with some frequency is a blocking play by my opponent, such as moving an assassin to block my way to an enchantress. Against a good player this can sometimes cost me the game.

I also find it harder than say, chess, because you must make contingency plans. IE, if I kill the scout then the play will go like this, but if the scout blocks which it has a 30 % chance of doing, the play will then have to go like this. Even though it is an open information game, because of the randomness factor you cannot clearly see what the board will look like in, say, 10 moves.

This is something I'm really working on because I feel that whoever can see more moves in advance will have a serious advantage in this game. So my question is, how many moves are you guys able to see in advance? What about the great players like bottle etc? Thanks in advance for your input.

[freefall37]

I have the same strategy as you. General game plan, and then small subplans for different situations that come up.

[Kyir]

at least 7 moves in advance, Ialways take 1 turn at any givin point in for healing those lucky blocks, and uselaly my plans arn't disrupted overwise

[Wayfaerer]

I usually have the entire game pre-determined before I make my first move

[Kyir]

I useally have my next game planned out before I make my move

[Noda]

i usually plan out what im going to have for next days lunch before my first move. but its always so simple, a nice chicken ceasr rap with a side of chily cheese fries...mmmmmmmm

but yeh, im always positive my last move will let me win...kinda where i plan too lol.

[BaxVarlet]

Don't really plan my moves out, just do the best move I can after whatever he does. Unless I feel imatient or brash, and then I just do whatever.

[Warcow]

I have calculated every possible outcome from now until the end of the game.

Incidentally, none of them have the other guy winning :cool2:

[22woger22]

How many moves ahead do you guys think?
Up to the point where I win. :)

:rolleyes:

[Office_Shredder]

Well... first I think about all those ways that my opponent can kick my ass. Then I say:
"Damnit! I need a better setup."

After which I proceed to win a surprisingly large portion of the time :)

[Walrus]

i think of what i would do if i were the opponent. ie what move i would make against my form, with his form. then i act accordingly.

[Bottle]

A lot of people who have posted here are just massaging their egos.

This isn't chess, in any sense of the word. Chess has three main differences; 1) smaller board, 2) no blocking, and 3) (and most importantly) units don't die in 1 hit. This means you can plan ahead by up to about 10-20 moves in a chess game if you're a good player, but in TAO, you have some many more variables that you can't plan any more than a few moves ahead with any more certainty than about a 10% chance it will happen at best.

The easiest way to plan ahead is in a turtle match when you have to juggle recovery times of a few units, to make sure you have optimal turn efficiency. And maybe you might come up with the odd combo that will result in you taking the enemy cleric. But in TAO you don't have many medium-term plans(about 5-20 turns). You have long-term plans of above 20 turns (I'm going to send my knights and scout down that grey player's flank, try to kill cleric and chanty and scout and retreat) or short-term plans of 1-4 turns (I'm going to send my scout there to threaten the stone focus, he'll move to block it with furgon, and then I can mudquake and kill the unarmored scout with my dragon that currently has 2 recovery).

Anyone who says they have the whole battle planned out from turn 1 is either a) lying or b) isn't a very good player, because what happens when that scout blocks a side shot or your opponent makes an unexpected move?

[Walrus]

what bottle said ^^

[Noda]

wow bottle...i think sarcasm is a word that should be added to your vocab.

[Bottle]

Kyir's first post wasn't sarcastic (unless it's so deep that no one will get it). As for the rest, you guys were just messing around, yes. Except maybe Wayfnub, I'm never sure if he's boasting or just making a joke. :)

[dab00z]

Thanks for a serious answer bottle :)

[Noda]

lol, well obviously wayf was serius haha :)

and i was just messin, i know u have some sence of humour under that big hea of yours :)

but ehh watev, now i feel bad lol.

within the first move i have set my objectives, set what i need to do to win, then usually plan about 3-5 moves(this is in freestyle)

in freestyle i actually usually just dont even pay attention to what the other might do, if u do that (in the beggining) u usually jsuit end up trying to defend it and miss out on ur oppertunity to gain the lead.

as for turtle, i have a set first like 8 moves for every type of game that might happen(same side defensive, same side power, same side upper corner, same side lower corner, same side furgy then the same for opposite)

but as bottle said, the middle game is where u need to plan ahaid, but it is my opinion that unless u are very good at defensive game it is almost worthless to go beyond any point where a block might happen.

having said that i will now need to gloat alil. i have had about 15 games in the last 2-3 weeks(almost all opposite side turtle games) where not one shot was a blockable one and in all of these by about 12 turns in i had all the plans ready until the end(or atleast until a point where i would have a deffinet advantage)

having said that, it is often my prob that i plan well until i have a good advantage...then just do dumb shit hehe.

but there...the actual answer

[dab00z]

lol don't feel bad I thought the joke posts were pretty funny actually :P Thanks for the real answer too :)

[TurtleKaze]

It mainly depends on the type of game you are playing. Rush vs Rush it is nearly impossible to predict your opponents moves until a few pieces have been committed, or an obvious combination is imminent. In these types of games, I've never had to think more than 3 moves ahead. I try to focus on pieces that can't go anywhere (just attacked), or have no or low blocking first. That will make planning on your part a little easier. As a general rule, I never plan longer than a given unit's remaining recovery time (given unit could be one you are attacking/attempting to paralyse, or one you are trying to save/get an extra attack from).

[Lonely Tylenol]

I always think of my next 3 moves after the current one. Every move my opponent makes that interrupts my 4-move plan, I think of a new plan for my next 3 moves.

To be honest, not much foresight can be put into this game. As Bottle explained, there are simply too many variables to think many moves into the game. I'm pretty certain nobody can think more than 5 moves out, if even that. General strategies are much easier to predict, but you have no way of predicting the specific moves of your opponents for many turns. The only way I am able to think 3 moves ahead is by considering a plan that doesn't get too specific (freeze unit X, then attack with A, B, and D to kill it). And even then I'm wrong a lot, especially with more complex strategies such as Wisp/Furgon/Frost traps. :bad:

[dab00z]

I always think of my next 3 moves after the current one. Every move my opponent makes that interrupts my 4-move plan, I think of a new plan for my next 3 moves.



That makes alot of sense actually. That was kind of what I was looking for. Thanks pos rep :)

[Realist]

The amount it is worth planning ahead for is very dependent on the situation. I don't usually consiously plan ahead except in terms of general strategies; play by play planning only occurs in very complicated situations involving many units.

[Bottle]

(freeze unit X, then attack with A, B, and D to kill it)
Is unit C skiving off or something?

[Faulty Logic]

....d and c are right next to each other on the keyboard, other wise that's funny.

[Anarchy_United]

I usually think 0ne or 2 moves ahead, in this and chess.
But I am much better at this than chess, because of the reasons bottle mentioned before

[bludhoundz]

It depends on the situation. With fewer units, I can see a lot farther (fewer possibilities). Sometimes I make a move just from impulse if it looks good. Most of the time I plan out my moves to see if they will be fatal to my plan or not (so a general search into the next few turns, but not always a definite plan).

[Warcow]

Alright to answer the question honestly,

I think nearly six or seven moves in advance at the start of a turtle game, because I know how I want things set up, regardless of what my enemy does.

In the later stages, I think three moves ahead. Those three moves are always based around scout recovery, or mud golem recovery.

[Lonely Tylenol]

I meant to put D, not C. I thought it would look stupid if I was reciting the alphabet in a discussion about strategy. :yes3:

[iceman2001]

Bottle is right, Kyir's first comment wasn't sarcastic, and I as well couldn't tell if Wayfaerer's second comment was serious or not. Anyway, how far you can look ahead depends heavily upon the situation. Many times you can predict your opponent's next move, and your response, ect. Taking into account blocks, healing, their own attacks, ect makes it very difficult to look very far ahead at all. The maximum I've been able to look forward is 5-6 moves (I usually only look ahead about 3-4). Although some claim to look forward more moves than that, their claims are false, as I can and usually beat them.

[JesusCraig]

I know none of you will believe this, but I plan my moves for nearly the entire game in advance. I do exactly as walrus said and think how my opponent will respond and can usually narrow the choices down, however I don't account for blocking when I do it, when a move is blocked I replan my entire game plan.

LT may be able to vouch for me, how definitive it is is another matter, I made a move early in the game (furgon shrub near the flank) which he chided me on as noobish, however subsequent play had me channelling him outward and the shrubs became an intregral part of the strategy much later in the game. I told him the reasoning when I placed the shrubs down that I would be using them to block his melee units, and he was inclined to disagree. It's the only time I ever beat him (I consider the grey game a draw it was so close) and I account it to foresight and my ability to foresee nearly all events in a game.

Totally unorthodox moves tend to screw me up a bit though and it takes me time to recover.

[iceman2001]

I want to explain what it means to have the entire game planned out. This means you know every move you would make in response to blocks, my own blocks, and lucky hits. Since I start with 10 units, there are already 90 unique outcomes for my units alone by the end of my second turn. (Assuming each unit has only one optimal move) You will have 90 as well. That means given any two starting setups, there are approximately 8100 situations by the end of the second turn. In another words, anyone who says they have every move planned is full of shit.

[Match Strike]

Someone should calculate the average amount of possibilities there are in an average TAO game. I wouldn't venture to guess, except it's probably in the mind-numbingly huge range.

[monkus]

Someone should calculate the average amount of possibilities there are in an average TAO game. I wouldn't venture to guess, except it's probably in the mind-numbingly huge range.

There are more positions a chess board can take than atoms in the universe. As bottle said, TAO is much more complicated, piece/rule/possibility wise. Thus, yes, I would guess it would be mind-numbing, scientific-notation-ain't-good-enough range, especially considering hp, armor, focus, unit facing direction, etc.

However, as for planning ahead, I find I almost always do best with a general strategy planned. It may be "play offensively", and it will probably include a hierarchy of targets, where the ones I place up top are much more decisive in the game and are only skipped for efficiency purposes. Very little counter-tactics are involved here, unless your strategy MUST shift to accomodate your opponent's strategy, such as taking the defensive against a surprisingly aggressive opponent.

Fulfilling this plan is your medium-term planning, to put it in Bottle's terms. If his cleric is at the top of the hierarchy, then planning on using my scout and muddy, with the dragon as a destraction, could be my medium term planning. This portion also includes most of your counter-planning, i.e. stopping your opponent's plan from coming to fruition. This does require a lot of intelligence as to figuring out your opponent's strategy, and many people (stupidly) forget to do this entirely.

The short-term planning is your micro-management. You figure out how you can do each step of the medium term, how you can stop individual ploys by your opponent, and how you can take advantage of any tiny little blip your opponent may have missed or allowed. This includes positioning a unit, facing a unit in a direction, shrubbing, or taking a shot at something. Many newer players focus ONLY on this, never knowing where the game may take them. It is also through your opponents actions in this that you usually learn of their medium or long-term strategy. Make sure you don't just defend against your opponent's short term; this is how many games are lost.

There are no turn limits. It's all relative.

[Lonely Tylenol]

I know none of you will believe this, but I plan my moves for nearly the entire game in advance. I do exactly as walrus said and think how my opponent will respond and can usually narrow the choices down, however I don't account for blocking when I do it, when a move is blocked I replan my entire game plan.

LT may be able to vouch for me, how definitive it is is another matter, I made a move early in the game (furgon shrub near the flank) which he chided me on as noobish, however subsequent play had me channelling him outward and the shrubs became an intregral part of the strategy much later in the game. I told him the reasoning when I placed the shrubs down that I would be using them to block his melee units, and he was inclined to disagree. It's the only time I ever beat him (I consider the grey game a draw it was so close) and I account it to foresight and my ability to foresee nearly all events in a game.

Totally unorthodox moves tend to screw me up a bit though and it takes me time to recover.

That one was painful. That is the first and only time I have witnessed someone who wasn't me manage to completely wall in a flank and get away with it (I did this to beat SpLiFF's rush and aside from that the only time it has served me any benefit was in my game with JesusCraig the very next day). JC carried on to flawless me that game, using a four-attack turtle with no Mud Golem.

Painful indeed... :confused:

If you guys want me to fully back his statement, he placed a shrub move in such a way that the only way his non-teleport units could get out were the two spaces between the LW and the wall... At this point he rushed me with the four attackers (2 Scouts, Dragon, Knight), making sure to go specifically for my Frost Golem and making sure a Scout was free to break Focus... After that, he killed off my wisp and Cleric reverted all his energy to the two panels between the LW and the wall. Forced to go on the offense (and not really being able to kill off the four damage units concentrated on that point), I advanced my Knights and mud on the Furgon flank. This is where he really gets me... The shrub placement, having boxed himself in, meant that if I were to go in with the Knight, he could freeze without moving much, and if I moved the Mud in, he could reshrub and freeze, and I'd be slaughtered.

And I was. I lost a Knight and Mud trying to break that one boxed-in shrub wall and the units behind it (although I almost got the Cleric and Furgon at separate times), and he just swept the rest. :( He must have planned his entire game on the shrub placement on the second move, and, funny thing, it worked.

My experimental 100-game formation ended at 92-8, he's the only one that flawlessed it.

[Match Strike]

But to say he planned each move for the whole game... that's a bit much, no matter how good he is.

[Lonely Tylenol]

Err... No, I believe I said he planned the entire game based on that move.

Or had the intention throughout the entire game of using that move to help him... Or...

Ugh. I'm so confused! I need sleep!

[Almantas]

Ok , this is very interesting , but being decent chess player myself i found one of Bottle's comments very wrong. He said that a good chess player is able to calculate 10-20 moves ahead. No offense , Bottle , but you probably dont know much about chess. Yes , grand master can calculate that much , but ONLY if the variation is fully forced . If position isnt razor-sharp no-one , even computer wont be able to calculate that much.

[TurtleKaze]

It is possible for a human or computer to calculate 10 moves ahead in chess. Sure, I don't have the patience to do so, but it is possible. 20 is a stretch for a human, but Deep Blue can do so in a matter of minutes. It literally evaluates every possible position.

Almantas, seeing as this was your first post, I thought that I should tell you that Bottle is the same person as The Exile. If you look at The Exile's player profile you will notice that chess is listed as one of his interests...

[Almantas]

It is possible for a human or computer to calculate 10 moves ahead in chess. Sure, I don't have the patience to do so, but it is possible. 20 is a stretch for a human, but Deep Blue can do so in a matter of minutes. It literally evaluates every possible position.

Almantas, seeing as this was your first post, I thought that I should tell you that Bottle is the same person as The Exile. If you look at The Exile's player profile you will notice that chess is listed as one of his interests...

Well , we could argue about that for an eternity ;) As i said , this is possible in sharp position where often you dont have many choices but just for computer, theres just too many variations to calculate. For every "real" move you'd have to calculate 1-10 sub-variations as well and evaluate resulting positions. And with current time controls this is probably impossible for humans. For computer - sure, but only if its some 32-procesor beast like Hydra . On PC fritz often changes his evaluation of position unless i allow him to analyze position for quite long time. And fritz's (and others engines) positional evaluation is fairly inaccurate. Not like its telling nonsences , but still it's evaluation is based on purely tactical moves. You can always make second rate move which is little worse then others and it wont cost very much.

On purely strategical positions it is impossible to calculate that much because theres just too many plans and counter-plans . And you can also add those plans that are just a bit worse than main ones. Chess isnt calculation game after all , only beginers think so.

P.s Yes , yours post second part is correct , i really didnt look at his profile , sorry.

[wolf-boy]

I havn't read this entire thread, but sometimes you don't have to anticipate your opponents moves for the most part, instead, make him think. When I say this I mean threaten him, make him make a hard decision. Like, Do I want my dragon, or my scout....Either one could do soon though, where Would I put it? That is very effective, putting them in tight spots, but then after I get that first threat, the eventually lead into others, but never get too offensive, especially with an anti or a turtle.

[.Vash.]

I play like if I was the other person seeing where I can make moves inthen I just go from there and if they make a move that I didnt see then I just try to make the best moves possible.

[wolf-boy]

Last time I remember Vash, didn't we finish at 1-1? You were wanting training from me or something.

[.Vash.]

Actually it was me 2 you 1 and yeah, but thats all over with now im doing my own thing. :P

[TurtleKaze]

Edit: to AL

Humans have the advantage of distinguishing between good moves and not so good ones, and working from there. This still isn't going to allow them to see 20 moves ahead though. But rather, they will know that they have some sort of advantage 20 moves down the road if a certain coarse of action isn't taken. Good computer programs also sift through relevent moves in the same manner, just not as well. An example of this would be the game where Kasparov actually beats Deep Blue (I can find the link to the game if you are interested). Kasparov trades a rook for a Bishop and pawn to create two passed pawns, which find their way home after about 20 moves. I'm sure Kasparov didn't see exactly how the pawns were going to queen, but he knew that his resulting position would be worth the sacrifice. It is believed that because of Deep Blues "sorting out irrelevent moves" that it wasn't able to see this. However, Kasparov only won 1 of the six games. As computers get faster, they will eventually be able to calculate every possible move in advance. The amount of possibilities is a ridiculously huge number, but it is a finite number, and will be reached.

[wolf-boy]

You lost, me. Please simplify that into simpler terms...

[Learz]

Personally, I play by reflex and experiance. He moves, then I move for the next turn, i.e., not really planning ahead.
However, in 1200+ games, I usually plan a couple moves ahead. Usually 3.
(No, I didn't read any of the 3 pages.)

[Bottle]

I havn't read this entire thread
This is never a good idea. :)

[FryLock]

I don't know about the rest of you, but I tend to close my eyes and click my mouse randomly when playing. It seems to work quite well for me.

That said, I think Monkus did the best job of describing "planning," such as it may be.

[Wayfaerer]

In another words, anyone who says they have every move planned is full of shit.
I wouldn't expect you to understand http://jconserv.net/swforums/images/smiles/icon_cool.gif

[Noda]

lol wyaf, ur my man.

[banditto]

Word wayf. I usually plan my laundry list for next month along with all of the football plays that are called on tv.

[Alpha Helix]

I think as far ahead as I have to. When I play someone good, I think about 5 moves ahead, depending on how hard I have to think. When I play a noob I chat on MSN and watch flash movies until they enevitably move all of their units into the LW range, take out the scout, and send the chanty in for the leftovers.

[Noda]

i think i plan the same moves ahaid no matter who im playing.

its the type of music im listening to while i play that is purely based off of how good the kid is...and what type im playing lol

[Vader13]

A lot of people who have posted here are just massaging their egos.

This isn't chess, in any sense of the word. Chess has three main differences; 1) smaller board, 2) no blocking, and 3) (and most importantly) units don't die in 1 hit. This means you can plan ahead by up to about 10-20 moves in a chess game if you're a good player, but in TAO, you have some many more variables that you can't plan any more than a few moves ahead with any more certainty than about a 10% chance it will happen at best.

The easiest way to plan ahead is in a turtle match when you have to juggle recovery times of a few units, to make sure you have optimal turn efficiency. And maybe you might come up with the odd combo that will result in you taking the enemy cleric. But in TAO you don't have many medium-term plans(about 5-20 turns). You have long-term plans of above 20 turns (I'm going to send my knights and scout down that grey player's flank, try to kill cleric and chanty and scout and retreat) or short-term plans of 1-4 turns (I'm going to send my scout there to threaten the stone focus, he'll move to block it with furgon, and then I can mudquake and kill the unarmored scout with my dragon that currently has 2 recovery).

Anyone who says they have the whole battle planned out from turn 1 is either a) lying or b) isn't a very good player, because what happens when that scout blocks a side shot or your opponent makes an unexpected move?Great post, one that will help me out a bunch. Before this thread, I hadnt ever really planned what my opponent might do, just what I want to do. Thanks Bottle.

[razor2007]

A computer in chess...nowadays......given unlimited time can calculate a "horizon" (the technical term for move-planning) of hundreds of plys. Now, a grandmaster, with a rank of about 2300, has a horizon of about 17 plys (I'm takin all this info form one of my chess books about a computer v. granmaster match). Normal, intermideate players like myself go 5 plys maybe.

Now, that said, in TAO I usually don't like to work any more moves than I can in chess. SO never mroe than 5. Usually 3. This is in a freestyle type game. In a turtle game I pla recovery several moves in advance, but I rarely plan actual unit moves at all in turtle games. For those I make the bets possible move that I can at that time.

So yeah, don't get all in a tizzy about planning WAY ahead, the more important thing is to sue the best tactics on every single move using things like "pins" and "forks" and "driving off", like in chess. I realzie it's harder to see but things like "If I do this I can take his cleric or his frost" or "I'll move here which forces him to move the kngiht to block the chanty, but then I'll witch burn him" are really what you should be looking for, not advanced plannin of mediocre moves.

EDIT: As a side note, the whole "ply" thing makes a difference. In chess, a ply is only one players move, not both. So when you hear that grandmasters plan 14 or 16 plays ahead, that's actually only 7 or 8 "moves" as in actions by both players.

[skylore]

Thinking ahead... hmm... ussually when i think ahead it involves my chanty or my scout. so ill give what i think ahead for them. for my chanty, i rely on guessing where they will move. but heres an example of how i would think ahead, is mainly when they move their scout onto my back line to take a shot on my cleric, they have to wait 2 turns (i think) so first i think ahead to where are all this ranged units (ppyros DWs scouts) if none are in range of hitting me within' one turn i look to the knights, just a general range if theyre close or not, and if everything is clear i move my chanty over twords the scout (never reaches first shot) then it just turns into baisic thinking ahead, but i ussually think ahead on that with what ever is necasarry. my chanty has to wait 2 turns and my barrier ward has to wait 2 and i see a knight coming ill think ahead to stop it and i think in one mode turn until i have another plan. and with a scout i ussually use him just to take out priest, DW, and pyros and with him i think ahead however many turns it would take to kill him. so tht would be keeping him in b-ward range, knowing when to heal, just general thinking ahead.

sorry if that post was a lil long and kinda rambling but its late and im not up to writing a good concise post :-p

[iownu4money]

i kinda plan my game before i play it but it always ends up changing after acuople of moves