Thursday, January 24, 2008

Fire and forget.

Leveling subjobs can suck if you don't like the subjob. As I do not like BLM, it sucks. :) SCH is currently behaving in much the same way as BLM though, which is troublesome for me. :/ It's nice to have some real cures and a regen available but honestly, now that Omoi has Curing Waltz I do very little curing and mostly just free nuking because I don't have the enfeebles that Iz has. The only benefit with doing this trio is that I get to hang out with Omoi and Iz, LOL, thats the awesome part. But I wish I could do that on a better job. What I need is the ability to go to sleep and wake up at level 37 SCH and BLM. No, I am not saying I need a leveling bot or something, I am saying that I need to be able to turn my brain off while I am leveling SCH and BLM. At least Iz and I have come up with a devastating combination move that allows us to do a large amount of damage to the mob and kill it much faster. It's called a skillchain! LOL Seriously though we have been doing Shining Strike to Burning Blade for a Fusion skillchain then we both burst with Fire, and it is at least fun and gives me a reason to pay attention for more than just Fire, Aero, Fire, Aero, repeat ad nauseam. Well, after a bit of running around and trying to make it back to Mhaura without getting killed, we all managed to get to level 20. Outside of leveling SCH, I did a couple of Campaign battles on WHM, which is actually pretty efficient for getting XP at that level. I get about 400 XP per battle but I can do it any time, and I am always active during it. I would still prefer to be in an XP party, but lately I haven't had a lot of time to build one and I can still search for a party while doing Campaign.

As a side note on Campaign, I have noticed that some people have been complaining about the fact that the game is thinning out because there are less people searching for parties. I really don't understand how these people can not put two and two together. Before Campaign there were people that didn't really want to party for XP but outside of XP scrolls they had no choice, now instead of waiting around for a party invite and dealing with crappy people they don't like they can now just get XP on there own. Now, I don't think this is the majority of people, but I do think that there is a significant group of people that prefer Campaign to partying. And this can in fact have an effect on the number of people available to make a party with. I do not think that it is an indication of anything.

The problem that I do have with the game right now, is not the game itself, but rather the non-existence of information about the next update. I am getting really antsy, and I am going to expect a lot out of the next update. My theory on the current update cycle is this: SE wanted to get the expansion out by the Christmas shopping window, and to do so had to limit or remove content that they wanted to have at launch but did not want to release if it did not function correctly. SE is different from Blizzard in this respect. SE does not use open betas to test game changes or content and although they do make mistakes they are more careful about implementation than Blizzard, who will put out content then fix it later. I guess what I am trying to say is that SE and Blizzard both will have issues that are caused by unexpected programming errors or conflicts, but Blizzard is more likely to release content that has problems that they know about than SE. So SE probably left a lot of content out of the expansion to allow for more testing and tweaking and will probably be adding that content that should have been released at launch in this coming update. I don't really mind this if the next update is a good one, but if it is as sparse as the expansion update then I will have an issue. If I am right about this, then it should be easy for them to have a lot in this update because they should just be wrapping up the things that they did want to include in the expansion.

Now for my last random comment for the day: I was reading a thread about how people view damage calculations and parses and I found some of the comments to be rather uninformed, and by that I mean that the questions the person was asking were the wrong questions. They wanted to know if damage calculations or parsing was a better way to determine gear selection. The simple answer is neither of them. They are both tools for helping to determine which gear is the best but neither of them will tell you in all cases what you should use for any particular situation. The question posed doubted the quality of the calculations in determining damage, and I had a problem with this argument.

Now I am not a computer programmer, but as long as the rounding is understood, the mathematic equation does not change if we have the assumptions of the equations correct. I think the questioner was confusing the variables within the equations for the equations themselves. We know (or at least theoretically know with the appropriate testing to back the theory) that the equations work, we do not know the way that SE derives the variables for the random factors in the equation. The only way we would know this is if we did have access to their random number generation algorithms. But the equations were derived from situations in which the variable part of the equation was stabilized. And the proof is in the pudding, the equations work. I have not seen an example for a place where they did not work. There might be places in which, in a certain variable range, there are certain factors that we do not know about, but as we do not know about them and for that there is no test, and that circumstance just becomes a Schrödinger's cat.

As for parsers, they can lie just like any other measure of data that is variable and open for interpretation. There are many factors that apply to parsers that make them a good tool for general analysis but should not be taken as gospel. They are like any measure of a large system with multiple variables apply, in that they provide a representation of a larger whole, but individual variables act on other variables in a myriad of ways making them always at least partially indeterminate. This isn't necessarily bad, but this indeterminate factor increases as the change in inputs decrease making it increasingly difficult to determine differences between smaller variations in input (i.e. differences in stats on gear). This is not limited to in-game variables like level of mob versus variable in the weapon damage equation, but also things like when you weaponskill, how fast you engage, the amount of time it takes the bard to get songs up, or even when your mage casts Dia. Too often people treat parsers as iron law, instead of a guide for making informed decisions. There is no error if there is a difference in parses. It is the nature of variables in a large system. There are a myriad of factors that determine damage for each swing and the damage it inflicts. This is then multiplied by a huge number of swings, applied to a large number of mobs with more variation in their statistics. Even if a player was perfect and managed to perfectly replicate their performance from one xp party to the next, they could not account for the actions of the 5 other people in their party and how their actions interrelate with the parsing player. A parse can only give you a snapshot of a very large and variable system and any differences between parses are just the results of variables within a large system.

Ultimately, neither equations or parsing will give you a complete answer. I have already stated the problem with parsers but the damage equations do not factor in same things as the problems with parsing (e.g. weaponskills, engage time, bard delay, Dia casting) and they also do not factor in accuracy, which is a major indict.

You should use all means at your disposal to come to a conclusion on decisions like these. Just using one or the other limits the amount of information input into the decision-making criteria and can lead to incorrect or flawed conclusions.

1 comment:

Anonymous said...

omg he refrenced Shroedinger's cat.