2008-09-14, 16:59:02
Quote:So what is a good play or good team?
A quite difficult question (which rather deserves a separate topic). I'd split it (playing good) up into some aspects, which are strategy(build), strategy(movement), (map) overview, micro, game theory, analyzing the opponents and team play (some might be missing, as this is just a short (and by far not enough reviewed) statement on my view on "playing good").
*Strategy (build): This means of course the basic strategy one is going to play. For example if someone plays scout+bombarding in mid, or light+basics on the side lanes at start and which way someone plans to go in future. This basic build is quite easy to learn, but it's hard to master, as it's definitely not something static (which it might seem to be). It's all about adjusting to the opponents. It's not only important to know all possible builds, but as well to know against which builds and play style from the opponents every strategy is good. For example: You know your opponent to play demolisher on lane (if from playing with him, or you waited in start some seconds to see the strats doesn't matter for the example) and you know that he'll probably stay near his own cp. As you know that you have to judge which build is good against this sort of strat. It might be some sort of creeping build, like scout+bow/hull or scout+bombarding or heli+bow or lt+bow/basic. Now you have to predict how the opponent is going to play in the future of the game most likely (you see, some points I mentioned separately come in here as well, that's all connected) to pick out which of those builds you should start with and how you'll should be changing tanks/getting weapons in the long run to counter his strat most effectively. Against this demolisher, you probably know him to play a kind of old fashioned style like demolisher->demon, so you can either try to push him with an early 2nd tank like thunder/air ship/heavy, or go for a creeping strat to get demon even before your opponent, or you try something in between by saving for sky tank and pushing then (you may as well try some risky tech straight to inf). Now you have to think about all this possibilities with their pros and cons (like getting an early 2nd tank will probably result in a switch from the opponents [or a change of tactic from the opponent] and you will have an additional gold spending in comparison to the demo->demon player, which you have to compensate) to pick the perfect build. Especially the compensation of the (at least 2k) gold is something which shouldn't be underestimated. You have to kill a demolisher with bow ~6 times (just with the kills it would be more than 6 times, but you creep better than him as well) before he can get the demon tank without getting deaths yourself to be break-even if you changed to a thunder tank!
Playing the right build against something doesn't mean playing something imba, as there is nothing imba.
*Strategy (movement): The thing most people consider to be connected with the skill most closely. The correct movement implies knowledge of ranges and the chance of being hit by weapons if x creeps are close to yourself, as well as knowledge of opponent abilities with their ranges. Players with the better movement will most likely win 1v1 fights with similar equipment (for example the lt+2basics against lt+2basics on the side lanes). Just “most likely” and not “for sure”, because of the random aim from the weapons if there are creeps nearly. But with a good movement and overview you should be the favorite to win the fight, even if it's just a 55-45 situation in your favor it's a good play (If you're bounty is equal, you mustn't forget about bounty differences if you're judging about good attacks on tanks. If your bounty is 550 and your opponents is 450 with 55-45 chances then it's a breakeven play). As this has quite a lot of variance someone can easily loose 3-4 such situations and get 0-4 stats in the beginning even if he played better than the opponent (might happen if the skill difference of 2 players is quite low).
*(Map) Overview: One thing I would consider as really important. You have to analyze situations in a short time, to judge how you should move/act. This involves the overview on the minimap and get's more important when players get teleporter.
*Micro: Micro is some sort of term for the control of the tank and units, which is especially important if you have some sort of summons. For example split summons close to opponent towers to minimize aoe dmg. Micro is closely related to the movement point and most people know the term micro, so I don't go into detail here.
*Game theory: Game theory is a key to improve everyone's game in general, as with analyses of situation you can find out which way of playing would have been the best and can act in a similar situation next time properly. As the success depends on the play of others it is important to know how your opponents play and will play and to give them false, or no information (as little as possible) on how you are going to play, to let them make mistakes in their builds and movement. A classic example is the armor upgrade one. If a team is going to get an infernal they should make the upgrades just when they got enough gold to buy an infernal, so that the opponents can't see earlier that you would be teching for it, not 10 minutes before. If you play with infernal you can fake chaos teleport or you can fake control point teleport to force your opponents to make mistakes they would probably don't make if they would know you're not teleporting (or sometimes they do think you're faking, but you're not). As both teams will try to hide their strats, guessing becomes an important factor, which brings us to another aspect:
*Analyzing the opponents: I probably don't need to say a lot here, as the information is kind of in the other texts implemented. You need to find out which start builds or which time someone buys item X implies strategy X. You need to know what the opponent is going to do if he starts flying from mid to a side, to give your mate a signal to retreat, or about opponent heli's movement patterns to plant mines correctly, etc.
*Team play: That's a difficult one to describe. Mostly it's about knowing your mates, know what they do in certain situations and play as synergetic as possible. It's about stunning when you're mates stun, staying close to each other, but not to close to be hit by aoe too hard, helping with a good timing (timing might be another aspect of skill/good play) as well as it is to play builds which fit to each others. If you go for an aoe strat it's good to have a player with a quite long stun playing with you, for example. Blinking for other players if you have the better map overview is another part of team play, or informing the other players about what you think an opponent is going to do (build wise, movement wise). As a team you can judge more accurately about analyzing opponents or which build anyone of your team should go for.
[btw: I don't understand what you're trying to point out with "using every imba condition", so please make another attempt to explain that more precisely.]
All those points are closely connected to each other and (without the movement) all are influenced by the team play quite strongly. Stats, especially in the early game, are mostly influenced by the movement and overview. To counter in the long run with good analyzing and game theoretical approaches often doesn't result instantly in better stats (but it mostly does in the long run), but rather in capturing cps, get more creeps or saving gold by changing tank less often and without a certain amount of micro any advantage in those aspects are not enough to win a game.
As you mentioned a game with 40k gold difference, I don't think that the effective gold difference was that much (taking tank changes, creeps and force gold into consideration as well and there was probably a lower average bounty. You have to keep in mind that late game bounty>>>>early game bounty as well, so better stats don't imply more gold from tank kills.) and even if it was, the team might lacked of good builds, or other decision-making. Maybe they should have invested this advantage into getting infernals earlier or capturing all cps to stop you from getting one more efficiently. I can't judge without any replay of that game.
Regarding the infernal: There had been other changes on infernal which made him balanced, especially the stop of his ulti by stun nerfed him a lot. He is still stronger than demon (as he should be), but I wouldn't call him imba. And taking my view on "good play" as a measure: Buying an infernal asap might be a good play, if it's a good strat against your opponents build. Killing people is a good play, if you're the favorite to kill the opponent in comparison to him killing you, with regard to the bounty. All using 3 sec stuns on 1 player is a good play, tho if you're playing 3v3 it's important to know which tank you should stun. Having 110-210 team stats is no play at all, it shows just that the other team probably plays better in some parts of the game, winning with 110-210 team shows that they lack of skills in some other parts of the game. Loosing game is allways a result of bad play.