|
United States7166 Posts
On March 16 2010 13:26 Manifesto7 wrote:Nice job Daigomi. I love the way the problem is laid out, and the way the various solutions are presented. I think it is obvious by playing the beta that simply having vision is not a good enough advantage for the high ground / low ground. Something else has to come into play in order for the features of the map to have more impact. Without it, map design is less important, and maps are one reason BW lasted as long as it has. Here is another post from this thread on the same issue that I agree with. Show nested quote +On March 05 2010 02:53 DJEtterStyle wrote: I'd take things a step further. The high ground mechanic needs to be changed -- absolutely, and for me to agree so strongly with InControl is a bit jarring -- but in general, SC2 lacks the all-important defender's advantage you see in pretty much all strategy games. Static defense is too weak, buildings fall too quickly, unit AI and bunching are so good that choke points and ramps have little tactical significance, and too few units have abilities that reward a defensive posture, with the most obvious example being the siege tank. Lurkers and spider mines, reavers and high templar added a major tactical element to SC1 because of their distinct lack of mobility.
We're seeing Starcraft 2 reduced to a lot of one-base play because it only takes the most minute of opportunities to win a game. A meaningful defender's advantage would open up a wealth of new, viable openings and make the game vastly more competitive and fun. i'm very happy to see this article, it was also very well done. i'm hoping Blizzard seriously considers this issue and understands the significance of it
i have to also agree completely with DJEtterStyle's point here. in fact this point probably deserves its own article, because right now it feels positional tactics and strategy, in particular defensive ones, is a bit lacking and plays such an important role in strategy, tactics, importance of map control, and the excellent dynamics that made BW such an excellent game
|
You make a few good points, particularly in illustrating the difference between miss chance and damage reduction, but I still feel that several of your arguments are invalid or beside the points that you are trying to make. This is mainly a reply to those particular arguments, rather than my opinion on whether or not there should be a miss chance in Starcraft 2. So I don't necessarily disagree with what you're saying, it's just my opinion that your claims doesn't follow from the arguments you make.
In regards to "a bit of bad luck will cost you the game", you say that a Starcraft player is never all-in, in the same sense as a poker player can be. That may be true, but that doesn't mean that being unlucky in a battle that is partly randomized could not be game decisive. Sure enough, you may not lose immediately just because a few more dragoons than expected missed their targets, and you are right when you say that the more units that are fighting, the less the variation becomes (you will almost never miss 10 out of 10 shots). Yet you say, and I think we all agree, that small things in Starcraft can have a huge impact on the game, but you don't really explain why this is not important when it comes to the miss-chance. Is it not conceivable that, for example, a few more shots than expected hits a high-templar one second too soon (maybe 9 out of 10 times it would have enough time to cast storm, but this particular time, it dies despite the high-ground advantage), and the game decisive battle is turned around completely? Wouldn't that be a bit of bad luck that costs you the game?
I agree that it would be necessary to wager the probabilities against each other, much like you would calculate pot odds in poker, but that is beyond the particular point of the argument of whether or not luck would (or could) become an important factor in the game.
In terms of predictability I also think you tend to exaggerate the importance of the miss-chance. While you are right, perhaps, that a game that is completely predictable is uninteresting, there are many factors that contribute to a game being unpredictable. A game that is decided purely by skill does not have to be predictable, and the fact that neither Flash nor Jaedong win all their games doesn't imply that Starcraft involves some elements of random luck (Starcraft does involve luck, but it's not a consequence of the best players not always winning). Take chess, for instance. It is widely recognized as a game that consists of pure skill, yet the outcome, especially between evenly matched players is rarely (if ever) completely predictable.
A game that is completely predictable would probably not be very interesting, so far we agree. I would interpret predictability in this case as the lack of random elements. If there are random elements, the players would sometimes be forced into making "bets" on the battles, which, sans randomness, would otherwise be unnecessary. When somebody say that the lack of perfect predictability is bad, I guess what they mean is that they would prefer such gambling situations to be minimized. Whether or not a "gambling element" is preferable or not is not crystal clear to me - I can see advantages and disadvantages either way. I do think the question is more complex than you portray it, however - your arguments in that regard seems quite irrelevant.
One more thing in regards to gambling - even if all random elements such as miss chances and scarab duds etc. are removed, there would still be "gambling" involved, such as trying to deceive or outsmart the opponent with various strategies. Making a hidden expansion can be viewed as a gamble, in that if the opponent scouts it quick enough, you are at a disadvantage, and otherwise you gain an advantage. Whether or not your opponent scouts is partly beyond your control, and so it's a random element from the perspective of one fixed player.
Finally, you also make it sound as if cheese would only be possible if there is a miss chance. But it seems to me like most cheese builds are not depending on lucky shots on units on high ground to succeed. Take a 4 or 5 pool build, for example, or a bunker rush. Whether such a cheese build is successful is usually determined by early scouting, the opponents build order, and micro, rather than random elements such as the high ground miss chance.
You convinced me that it's easier to balance miss chance, as opposed to the alternatives you propose, and if there is no other means of getting some kind of high ground advantage, I agree with the first point of your list of pros as well. The others, though, I don't quite see how they follow from your article.
|
Excellent read. Very true though. It would a more realistic feel to the game. It would be like an actual war with the miss chance (or atleast moreso). As for the professional players not accepting this rule, i think majority rules should apply here. There are more SC2 players than there are professionals. The high-grounds were made in SC2 to make the game more 3D. Now, they should be used to make the gameplay more three dimensional as well.
|
I think the difference between miss chance and damage reduction can be explained with one mechanic: overkill. Miss chance does not affect a unit's tendency to overkill, whereas damage reduction reduces overkill, making heavy-hitting units more effective by reducing the damage they "waste" on overkilling targets. So miss chance affects all units equally, whereas damage reduction does not. Damage reduction would then act at both a balance shift as well as a tactical mechanic, when all we want from high ground positioning is a tactical mechanic.
|
I disagree with the second con of Speed Reduction.
The other disadvantage is that it is less logical to the average gamer than a damage reduction or miss-chance is. This might be because we are not used to the concept, but the idea that marines will suddenly start attacking more slowly when the enemy units get to higher ground seems strange to me. It makes the solution feel artificial, like a solution to a problem rather than a natural part of the game. I think in real life to hit someone with a positional advantage over you would require more focus, so I think to an extent speed reduction does make sense, however I still agree that miss percentage is the correct way to go.
|
Well written article, the only part i disagree with is your predictability header.
To me, a game is great if the player that plays better wins every game. Now this does not mean that the better player will win every game, but that the unpredictability lies in the players ability to perform in the given game. Instead off on a scarab going off or dudding.(not saying that the scarabs and such don't give amazing game play, but i wouldn't say it's required for an amazing game)
|
I haven't played SC2, so I'll use SC1 units.
The suggestion being examined is straightforward. When units attack from lower ground to higher ground, they should have a certain chance to miss. What this chance should be is up for discussion, but I feel that, combined with the current sight mechanic where lower ground units cannot attack higher ground units without vision of the high ground, a 25% miss chance would be a sufficient higher ground advantage.
So does this mean that when there's 2 vultures on high ground versus 2 hydras on low ground (all on hold position) the vultures will hit the hydras while these won't be able to attack back? And if an Overlord comes into play, and flies into high grownd, the hydras would make complete damage against the vultures?
My point is this, if the higher ground advantage is applied in SC2 by the factor visibility or by the actual difference in ground altitude. If it's determined by visibility, a scan or a flying unit would completely deny any altitude difference just as if both armies were battling at same ground level; while if its actually determined by the altitude the %miss would still apply when the overlord grants them vision (the OL having reached higher ground while hydras are still on lower ground), this is how it would work in SC1 I think, mixing both visibility and higher ground factors.
|
There is no way to completely balance the disadvantage of shooting uphill.
Lets say they go with -50% dmg shooting uphill:
Thor 90 dmg attack vs hydralisk 90 hp.
On level ground thor kills hydra in 1 attack.
Shooting uphill thor would only do 45 dmg. So thor shoots hydra, hydra goes down to 45 hp, but before thor can shoot again hydra has regen'd 1 hp and is now at 46 hp, thor shoots again hydra at 1 hp, now thor needs a 3rd shot to kill the hydra, thus meaning a simple 50% dmg reduction is taking the thor 3x as many shots as normal.
On the other hand lets do thor vs zergling.
Level ground thor kills zergling in 1 attack. Shooting uphill thor's dmg is reduced to 45, thor still only takes 1 attack to kill, thus leaving no advantage/disadvantage for either team.
Basically what I'm saying is that certain unit matchups are favored over others, thor gets completely screwed attacking hydras with a 50% dmg loss. On level ground thor kills both hydras and zerglings in 1 hit. But lowering his dmg 50% makes him take 3 shots to kill hydras, and still only 1 shot to kill zerglings.
With 50% miss rate: Thor can still kill hydralisk on his first shot (no advantage/disadvantage for either team). He could even miss 10 times in a row then kill hydra on his 11th attack, or he could just as easily klll 10 hydralisks in his first 10 attacks. The odds of him missing 10 shots in a row is the same as him hitting 10 shots in a row. While there can still be no advantage/disadvantage, odds are the attacker will be at a disadvantage but with good luck.. can still kill just as fast as on level ground.
With shorter range attacking uphill: There can be disadvantage or no disadvantage at all.
disadvantage: High ground has tank, low ground as a marine. Marine has to spend more time moving to the cliff to be able to shoot the tank, that's if he can even still have enough range to reach the tank.
no disadvantage:: High ground has melee or short ranged units. Low ground has siege tanks, even with reduced range the siege tanks can still attack the high ground without being touched.
Speed reduction: I'm not really a fan of this, I think it'd be really hard to even attempt to put in the game. As there would probably be some kind of bug that players would figure out by switching from attacking high ground and level ground units to mess with the units attack speed. I also don't see why it would make sense for units to attack slower uphill. Units attack slower will just be less entertaining for the viewers, especially when they already are slow attack units by default. And what would you apply the speed reduction to? The cooldown (the time after you attack and before you can attack again) or the cast time. If it's the cooldown then you could attack then run away so that you are out of range of being shot, then go back into range when cooldown is over thus defeating the purpose of it. Think of dragoon micro, where you attack then run away then attack again, in this case you'd just run twice as far due to longer cooldown.
I personally would go for miss rate. Make it around 75% hit, 25% miss. This means that while at a disadvantage you could still break his ramp if and only if you had the superior force. Having a miss rate too high deters the low ground player from attacking and thus spending more time traveling around the map to find a different location to attack. This from the viewers point of view is bad for the game. People want to see players attack each other not running around the map all day. At the end of the day the viewers are a huge part of the success of e-sports or just sports in general. Without the fans watching there will be little to no tourneys and little to no prize money.
75% hit, 25% miss still gives you a pretty good advantage. Remember YOU won't always be the person on the high ground, you will be attacking from low ground a lot as well. High ground should help you win small battles here and there, but it shouldn't be the only reason you win/lose games.
|
Nice read. An important thing to add is that Blizzard already has a method of reducing the likelihood of luck in the most extreme of cases (i.e. missing or hitting up a cliff 10 times in a row) through the use of Pseudo Random Distribution which is built into the Warcraft 3 engine. The general idea is that the miss chance would start at some value below the expected miss chance, and the chance would slowly increase with each hit until a miss occurs, at which point the number would reset.
Anyways this thread is worth a read.
|
90% of all games which have a highground have some advantage system implimented to increase the statistics of the high ground player the win providing they have the required unit build ect.
High ground is a spot of elevated terrain which can be useful in military tactics. Fighting from an elevated position is easier for a number of reasons. Soldiers will tire more quickly when fighting uphill, will move more slowly, and if fighting in formation will have little ability to see beyond the soldiers in front of them. Likewise, soldiers fighting on the hill won't get tired as quickly, will move faster, and will be able to see farther when in formation, aiding them in making smart tactical maneuvers. Furthermore, soldiers who are elevated above their enemies can get greater range out of low-speed projectiles like rocks and javelins. Likewise, rocks and javelins will have less range when thrown uphill.
|
i couldnt believe this wasnt in the game already. sc1 had it and it made sense, with the new ai its way easier to get up chokes and cliffs. i mean it really does make sense to me
|
I think the %miss option is by far the best choice. It just seem to make more sense that some units can dodge a full attack instead of percent reduction, which just means one army inflicts exactly X% less damage, and makes battles more interesting.
Also, I think what most people do not know is that game designers have the ability to skew percentages. For example, many games that involve luck will have it so that if there is a 20% miss rate, if a unit misses the first attack, the second attack is actually less likely to miss, and so on until finally there is an attack where the attacking unit is 100% guaranteed to hit even from low ground. See Sirlin's summary of Sid Meier's discussion at GDIC 2010 here:
http://www.sirlin.net/blog/2010/3/12/gdc-2010-day-2.html
Sid Meier talks about scaling the luck factor so that the luck behaves a lot more like it "should" behave rather than being entirely luck.
|
The chance to hit should be calculated in real time, equivalent to the exposure percentage of the target's hitbox.
Marines shooting from a cliff exposing 25% of their bodies (also meaning that they're visible and no spotter is required) would only be hit 25% of the time as compared to 100% exposed marines. Explosive or ricocheting projectiles can increase those chances by some factor unknown to me.
No unit should be able to shoot through walls or the ground. That means even units on the high ground would have to get close to the cliff to shoot below, unless they have grenade projectiles of sorts, but then so can the units from below shoot targets above with sight. Depending on how these projectiles move and how "hard" it is to throw a grenade from below, chances can be <100% by a factor I do not want to think of.
There should also be a chance to hit marines inside bunkers in this sense, since it shouldn't be impossible, only "really hard".
Uniform chances are the dumb way of approximating uncertainty, theres a hundred more clever and realistic ways to give the high ground a defensive advantage, ways that go beyond what I tried to demonstrate and are certainly possible if blizzard even bothers.
|
I think the author has a severe misconception on how likely it is to win the lottery, unless most lotteries have like a million percent increase in chance of winning compared to the super lotto in the United States, which is the lottery most commonly referenced.
If you bought 15 lottery tickets and the chances to win were greater than 1/1,000,000 then I assure you people would be buying lottery tickets like crazy simply because the odds would actually be in your favor...
|
This post isn't spammed across bnet forums why?
Seriously, more valid input needs to be fed to blizzard forums instead of people complaining about how races are OP.
Great post, this nailed the issue on the head.
|
Osaka26959 Posts
Odds of winning the lottery...
it follows that each pick of six numbers has a probability of 1/13,983,816 = 0.00000007151. This is roughly the same probability as obtaining 24 heads in succession when flipping a fair coin!
... worse than missing 10 shots in a row.
|
On March 16 2010 16:42 Manifesto7 wrote:Odds of winning the lottery... Show nested quote + it follows that each pick of six numbers has a probability of 1/13,983,816 = 0.00000007151. This is roughly the same probability as obtaining 24 heads in succession when flipping a fair coin! ... worse than missing 10 shots in a row.
Exactly... compare it to
"To put this into context, if you buy fifteen lottery tickets, you have a better chance of winning the lottery than missing 10 shots in a row"
You don't have a better chance at winning the lottery. You still have a better chance at missing 10 shots in a row.
|
The problem of a fixed miss chance is it not only advantage high ground, it also advantages less costly units that you can mass vs bigger more powerful unit, because the chance of getting fucked by the RNG is higher if you have less units(even more if they fire slowly). In the end, over a thousand shots, they will even out, but probabilities work in such way that you can get bad streaks; not necessarily missing every shot, but missing 50% of the shots on one fight instead of 25% would make a big difference and is very possible if you only have 7-8big units shooting every second/second and a half.
Could use a system that scales the miss% based on cost and/or attack speed so tanks for example are less likely to fail shooting at a high ground over a small sample than marines(who shoot faster, and cost like 1/4 of the price so theorically you'd have 4marines for each tank). Not a big difference, but for example, if marines had 25% to miss, tanks would only have 20%(or 15 or whatever).
|
|
The only "problem" with a chance to miss is that it's nearly impossible to make this visually intuitive.
Unless you want a big "MISS!" to pop over unit's heads like some old RPG, you will simply see a unit shooting "normally", but not killing anything like normal. This confused me while I played SC1 in single player mode.
My suggestion? Cause the projectiles to actually fire off-course slightly when the miss, creating a visual representation that spectators and players will have a better chance understanding.
|
|
|
|