@Deribus said in Balance Thread Guidelines Feedback:
There are several methods to doing this and it isn't expected for someone to utilize all of them as certain methods are easier for certain things. I will rank them by their levels of respectability as an argument/rationale with the easiest/least respectable ones being first.
TL/DR: stats can sometimes be the best way to analyze things, and analyzing one, or even multiple replays might not accurately show you whether something is unbalanced anyway, so sometimes the well-supported opinions of the most knowledgeable players should be weighed more heavily than replays for balance changes.
So my problem with this is that just because something is the easiest way to explain the issue (stat review), does not necessarily mean it is the "least respectable" argument/rationale. It really depends entirely on what you are talking about. Simply talking about stats could easily be the best way to make some arguments. A perfect example would be the discussions of ras sacus vs mass fabs. Who needs to see a replay, when the numbers can tell you the entire story? You can very easily and accurately explain the pros and cons of these two economy options simply by crunching some numbers. (I do agree that for combat units the interactions are much more complex and pure stats are almost certainly insufficient.)
Of course there are some minor complications that you can't reliably calculate, like how exactly should you quantify the value of the mobility of sacus, or the extra risk due to the explosiveness of mass fabs? But I don't really think I can show a replay that gives you a more accurate, realistic explanation of how that interaction typically works in a game either. I could show you a replay where many mass fabs chain explode, and another where sacus do too. I could show you some sacus assisting a shield under fire from t3 arty and dying, and one with them assisting the shield and surviving because there are also a few t3 engies supporting. Do we need to do a meta-analysis of replays? How do we know that a sample of replays is not biased by noobs who don't know how to space out their fabs enough/don't know how to utilize the adjacency bonuses of fabs enough, etc.?
The point is that many issues are nearly impossible to prove through showing a single, or even a small number of replays (I think especially if they are something based on risk, e.g. explosiveness, acu snipes). How do we quantify a situation dependent risk (e.g. do you have air control or not? how good is your scouting? etc.), versus the cost to mitigate that risk? "Mercies aren't op because two cheap t1 maa counter them" or "you were just careless with your acu!" You really need to know how frequently the risk matters. One game can't show you that. Were you stupid for building mass fabs next to pgens for the adjacency bonus because in that game situation it was too risky of an idea, or because it is too risky generally? This can only be answered by analyzing many replays...OR we can just get a feel from people based on what they say because they will have a sense of the cost/benefit analysis based on what has happened in the games they have played, which should probably be discounted based on how noob they are. You should definitely put a lot more weight on Blodir's or Nexus' opinion than mine.
Another complication is that balance is certainly map dependent as well (wagners are good in certain situations, bad in others?), but I don't have a solution for that, just saying it to reiterate that no single game replay could conclusively show the overall balance of a unit, because it could be much stronger/weaker on other maps.