|
On March 05 2010 05:42 Bill307 wrote:
I think the random miss idea is good.
But imo they should tweak the random number distribution so that if you approach an extreme situation (e.g. a dragoon missing 5 shots in a row against a marine), then the probability of the situation becoming even more extreme is lower.
Example: suppose (for simplicity) units have a 50% chance of hitting units on higher ground. Then the probability of missing 5 shots in a row is 1 / 2^5 = 1 / 32. At that point, you have a 50% chance of missing the 6th shot in a row. I'm suggesting that the probability of missing the 6th shot should be less, say 25%.
The nice thing about random numbers is that, if you fired a large # of shots uphill, then 45% to 55% of them would hit (assuming you have a 50% chance). But, for a small # of shots, you can easily get hit %s anywhere from 0% to 100%. And SC has a lot of small yet important battles with a small # of shots fired. This is why it is useful to bring the actual hit % closer to the average % for a small # of shots.
Ultimately, it's all about controlling the variance of the random distribution. I agree that no variance -- no randomness -- is bad, but I think that SCBW has too much variance, i.e. you're too likely to end up with a 0% or 100% hit rate when a small # of shots is fired. I think something in between would be best. E.g. "If I attack up this ramp, I can expect 40% to 60% of my shots to miss. Am I willing to accept that risk?" (As opposed to exactly 50%, or 20% to 80%.)
This would make sense if the point of a 50% miss rate is primarily to halve the damage done. IMO, it's less about how much damage is done and more about injecting quick critical thinking into the game. If the distribution was skewed after x misses in a row, it's like the game is trying to make up the negative effects of the risk the player took. If the distribution stays the same after, say, 5 missed shots in a row, you still have to sit there thinking "crap, I already missed 5 times, is it worth staying?" as opposed to "oh i missed 5 times, the next shot is definitely going to hit, that's how the random number generator works." If you still have to make that critical decision with a 50% miss chance, on-the-fly no less, it lets better players who have a better understanding of the risks involved differentiate themselves from lesser players.
|
Gontech reminds me of another point as well:
If you have probability sets that are dependent on previous sets, you'll have this never-ending cycle of overcorrection. You're trying to reach 50%, but you accidentally go 70% in one cycle, so you compensate by going (for example) 35% in the next cycle, and then compensate by going 65% in the next cycle...
You can see where I'm going with this, you have alternating high and low cycles which will just be WEIRD if nothing else. I think the straight 50% by itself would do a better job and have a better spread of results rather than dips and peaks (and it would require no extra maintenance on top either).
|
I agree with Incontrol and Chill (especially about the good implications of randomness in this particular case), and i want to point out that most people still regard miss chance in BW as 30% or 1/3 wich is incorrect, it is actually a lot worse but people dosn't even seem to realize this while playing.
It is actually ~52% chance to hit vs high ground, wich is almost a coinflip each shot.
So if you can't tell the difference between 50% and 70% to hit (it means you get only ~2/3 of hits you would with 70% wich is huge) you think you should call randomness bad just because you feel like it, without even considering it's implications ?
People arguing about randomness = bad also seem to not realize in how many sports and games randomness is and that it also take skill to get along with it. Poker is big example, but even with sports like motorsports like WRC or F1, where tire choice is important when there is chance for rain - (even with best forecasters you cant be shure about track condition), racers make gambles on tires and win or loose just because of it, best ones have not only driving skill but also dare to gamble in tight situations.
edit: poor english made less poor 
|
I wholly support bringing the miss chance into cliffs.
Damage reduction skews with damage vs. armor stuff. And someone proposed some static "every 4th shot misses", which sounds ridicoulous - I take 3 shots and then retreat, how cool is that...
RNG baby
|
Totally agree with this, please put this back in Blizzard.
|
@DefMatrixUltra
"[1] This is almost certainly not true. Remove high ground advantage from BW and see if the game balance is the same."
Ok... LOL Go play tau cross!
|
On March 05 2010 07:35 Card5harko6 wrote:
@DefMatrixUltra
"[1] This is almost certainly not true. Remove high ground advantage from BW and see if the game balance is the same."
Ok... LOL Go play tau cross!
I will if you go play Python. Or Destination.
|
I believe this was removed to get rid of as much of the negatively viewed luck factor as possible. But the fact of the matter is luck plays a huge role in many skill based games. That's what makes them so fun, overcoming that luck. (poker)
|
I never really understood Anakin's disadvantage from being on the low ground... which I guess the video beautifully points out. -__- Plot holes for the win! :D
|
United States17042 Posts
I strongly agree with incontrol. I read the rest of the thread, and I'm pretty sure that the strongest arguments is the strong prevalence of the lack of defenders advantage.
Some things to think about with this argument; the kind of flash/savior style of playing with the defenders advantage was something that was developed over several years, and it required some very tight timings, build orders, and reactions to pull off. In starcraft 1, there was a large amount of 1 base play with no expansions early on as well. We won't start to see macro/expansion based play until we start seeing tighter timings and build orders.
One other factor in the one base argument is that the current map pool is heavily weighted towards certain strategies (yes, the maps are different, an and they're probably better than what was included with the orginal starcraft, but they're still not amazing).
that being said, there are certain advantages to having ramps. It makes it so that there's an earlygame defenders advantage, especially since walling off is so much easier....the game would be more complex if there was a damage decrease while shooting uphill.
|
i highly agree with the importance of cliffs but i think the randomness could be altered to be % less damage when shooting up cliffs instead of random
|
Norway28911 Posts
definitely agreed with inc. some slight randomness is not bad. if the best player always wins then both playing and watching is a lot less entertaining.
|
For those arguing that straight % damage reduction is inferior to random % miss because it makes high armor units too strong vs low damage units, couldn't this be solved by applying the %dmg reduction after armor has been applied?
From a mathematical standpoint, these two alternatives provide the same effect given an infinitely large sample size of 'shots', whilst the former being 'fairer' given low sample sizes, i.e. in real games.
Edit:
Of course, an argument is that % damage reduction screws up unit attack / hp ratios, so that for some combination of units, the balance of power between the units is shifted.
Example from SC1: Marine vs Vulture
On normal ground, marine kills the vulture in 14 hits, Vulture kills the marine in 2 hits.
The ratio between the number of hits required is 7:1.
Given 50% dmg reduction, marine does 3 damage to the vulture, and vulture does 10 damage to the marine.
Marine kills the vulture in 27 hits, Vulture kills the marine in 4 hits.
The ratio between the number of hits required is now 27:4, or slightly less than 7:1.
So the marine actually improves vs vulture with % damage reduction, since there is less wasted overkill damage on its last hit.
Personally, I would rather tolerate this slight imbalancing of unit relationships than deal with random chance deciding which unit wins the battle, but I suppose other people will feel differently.
|
On March 05 2010 05:42 Bill307 wrote:
I think the random miss idea is good.
But imo they should tweak the random number distribution so that if you approach an extreme situation (e.g. a dragoon missing 5 shots in a row against a marine), then the probability of the situation becoming even more extreme is lower. All you'd need to do to deal with that is treat the RNG here as a deck of cards, instead of a dice roll. If you wanted it at 50% after exactly X attacks you'll have hit a 50/50 ratio. The more cards you put in this "hit/miss" deck the more stringy it can possibly be. Anywhere from 8-20 would probably fit well with how fast the SC rate of fire is.
I really do not think the luck factor of the high ground in SC1 was that significant because the odds were so ridiculously in the favor of the high ground that it conveyed such an overwhelming advantage that you simply don't attack up high ground unless you are absolutely certain you can break through. In all the games I've ever played I don't think I've ever once considered, "hmm, there's a slight chance that I'll get lucky and be able to barely break through if this the stars totally align for me", I'll set up a contain and wait until it's safe to push in.
|
you know how there's an attacker/defender dynamic in sc.. The miss factor lends itself very nice to this concept. one player can hold off a hoard because he has the high ground and wants to defend while he gets an economic advantage but the opponent can double expand.. It adds a Really interesting dynamic to the game. Imagine for a moment both players expanded equally and had armies that were of the same strength.. you now have something much more similar to wc3.. not sc2 and not the next best e sport in the world..
InControl is a professional gamer.. and out of all the controversial topics in SC2. he chose this one to write a thread and express his concerns.. obviously there is something that needs to be addressed.
|
Hm, I wasn't aware that high ground advantage was taken away. It certainly made things a LOT more interesting in sc:bw. Players shouldn't have equal footing if you're fighting from low to high ground as long as they have vision. Just the sight advantage isn't enough. Without a distinct damage advantage from the high ground player, there would be less fighting over strategic positions. This is a strategy game after all, and players should definitely be rewarded for taking strategic positions.
I'm in favour of bringing back the decreased accuracy of low ground vs high ground. The lower accuracy against high ground also added a degree of realism to the game. I thought that the advantage was great in both the observer and the player's perspective because it added more nail-biting situations like when a defender just barely holds his ramp to a rush.
The range solution is interesting, but it might give too much of an advantage to high ground. Imagine you have TvT, where it's a battle of millimeters. If you give the high ground group extra range, you kinda force the low ground group to charge up.
Damage reduction would work too I guess, but I wouldn't want to see that some units suddenly get an additional hit in some match up on high ground vs low ground. In the decreased accuracy only some units will be able to take an extra hit while others won't. If all the units get an extra hit on high ground, that seems a little strong.
In any case, I hope blizzard implements and tests some of these solutions in beta. I really hope that high ground advantage isn't limited to vision.
|
Bill307
Canada9103 Posts
On March 05 2010 07:17 DefMatrixUltra wrote:
Gontech reminds me of another point as well:
If you have probability sets that are dependent on previous sets, you'll have this never-ending cycle of overcorrection. You're trying to reach 50%, but you accidentally go 70% in one cycle, so you compensate by going (for example) 35% in the next cycle, and then compensate by going 65% in the next cycle...
You can see where I'm going with this, you have alternating high and low cycles which will just be WEIRD if nothing else. I think the straight 50% by itself would do a better job and have a better spread of results rather than dips and peaks (and it would require no extra maintenance on top either).
I like how you just assume it has to be implemented in a stupid overcorrecting way like that.
Maybe try asking yourself if it can be done in a good way instead of just trying to prove me wrong?
|
United States44387 Posts
I really liked the way ramps worked in terms of tactical decision making. It just create options to be exploited by the intelligent player who could correctly analyse all their assets and combine them effectively.
|
Bill307
Canada9103 Posts
]On March 05 2010 07:04 gontech wrote:
This would make sense if the point of a 50% miss rate is primarily to halve the damage done. IMO, it's less about how much damage is done and more about injecting quick critical thinking into the game. If the distribution was skewed after x misses in a row, it's like the game is trying to make up the negative effects of the risk the player took. If the distribution stays the same after, say, 5 missed shots in a row, you still have to sit there thinking "crap, I already missed 5 times, is it worth staying?" as opposed to "oh i missed 5 times, the next shot is definitely going to hit, that's how the random number generator works." If you still have to make that critical decision with a 50% miss chance, on-the-fly no less, it lets better players who have a better understanding of the risks involved differentiate themselves from lesser players.
I'm pretty sure most people would think their 6th shot has a higher chance of hitting when it doesn't.  But you do make a good point.
On the other hand, if you just missed 5 shots in a row, then you're already a lot worse off than you expected, so you've already suffered for taking that risk.
My personal preference is, I'd trade some of the sharpness of making that 50/50 decision in order to have a lower probability of ending up in an extreme case like that.
|
I think that they should implement it again, it definitely took away from strategic placement of troops, and as well as the importance of you positioning your army. It hurt defenses as well alot, as sieged tanks on a ridge in SC2 will not strike the same fear as it did in SC1.
|
|
|
|
|
|
EPT
