Jump to content

Macroeconomics

Members
  • Posts

    826
  • Joined

Reputation

10 Good
  1. People can call the phenomenon by whatever term they like. But in layman's terms, when people are able to migrate from one group to another to take advantage of the group's characteristics, it's just a bad conclusion to believe that the two groups have the same underlying qualities (player skill in our SWtoR case). I'll give you a real world example. On an average month in the US about 500,000 residential homes are sold. This generates loans (mortgages) each month, with a aggregate value of about $300 billion. Those mortgages are bought by a variety of investors (bank, mutual funds, the US gov't, etc.) and it goes without saying that those investors don't like it when the borrower defaults on his loan. So those investors measure the quality of borrowers prior to the loan being made. But they also measure the quality of the borrowers after the loan is made. They use a variety of metrics. For example they measure what percentage of borrowers in various loan categories are behind in their loan payments (i.e. delinquency rate). Ooooo......meta statistics.....here we come..... Can you tell that mistakes will be made and that a butt load of money is going to be lost here? COUNT ON IT! See, just like our SWtoR players had a choice between a variety of advanced classes, those home buyers had a choice between a variety of different home mortgages such as 30 year fixed, 15 yr fixed, 7/1 ARMs, etc. Each of the mortgage types appeal to different types of borrowers. So 15 yr fixed mortgages appeal to borrowers that don't like to have a lot of debt and want to pay off their mortgages quickly. ARM (adjustable rate mortgages) appeal to borrowers who need an initial low interest rate in order to qualify for their loan. Needless to say that these people tend to default more often and all else being equal are worse quality borrowers. So even if I assemble a group of ARM borrowers that has the same delinquency rate as a group of 15yr fixed borrowers, I should NOT assume those two loan groups have the same quality. This would be EXACTLY the same mistake as assuming that because the W/L ratio for Merc healers is the same as the W/L ratio for Op healers, that the two player bases have the same player skill (and thus the two classes are appropriately power balanced). But back in 2006-2008, investors didn't really understand the magnitude such borrower choice (i.e. SWtoR player migration) could have. Investors said, yeah I know that ARM borrower is twice is bad as the 15yr borrower. I know that subprime borrower is 3x as bad. But I am taking that into account with the lower price I am paying for those mortgages. Well, that was a bad conclusion. Because it was more like 30x when all was said and done. And so Bear Stearns, Lehman, AIG, FNMA, Freddie Mac, AMBAC, etc. happened. Trillions of dollars up in smoke. Very smart people. Using meta statistics. The wrong way. So now you see why you can get paid very large butt loads of money if you can do it right.
  2. Remember when BW posted a dev note saying they were going to initiate the first of a series of nerfs to Op heals? That first "nerf" - it was a buff. They never even attempted any further "nerfs". lulz....
  3. People say that dps Merc has too much burst to warrant giving them increased survivability, but dps Merc burst is still not sufficient to kill an opponent. Dps merc burst is higher than that for Sorc dps. But it doesn't reach the thresh hold where you can kill a full health opponent in the time span of two stuns. Nevermind the fact that Merc doesn't get two stuns. So dps Merc winds up in the same bin as dps Sorc - a damage dealer than needs to work down his opponent in order to kill them. Except the Sorc has about 3x the off healing capacity, more get out of jail cards and more team utility. And of course better base damage output, which is the primary desired trait of a sustained damage dealer. Some will of course reply that Merc shouldn't be able to kill enemies 1v1, it is best when used in team focus firing. Well except that there are plenty of dps subclasses that can kill 1v1 AND have better survivability that Merc dps, AND have better base damage output than Merc. Or simply compare to the Sorc dps which is a better team dps bro, with significantly better root/snare/target fixing abilities and again, higher dps.
  4. It's not that simple. No one was playing Assassin Deception, but not because it is underpowered. Rather because its stablemate, Assassin Hatred was the most OP class in the game. Buffing Assassin Deception because it wasn't played much was a horrible decision. Implementing it with a 30m, low CD, mezz/root....well LOL. And you can not look simply at % of each class that winds up with rating higher than a fixed figure. That would just be a close proxy for W/L ratio, with all the same flaws. On the other hand, a dev somewhere has read this thread. We know this because I wrote this thread pretty much in direct response to EricMusco's post in DevTracker on 4/6/15: "We target a 50% win rate for each Discipline in PvP, and Carnage/Combat and Annihilation/Watchman are slightly above this ideal win rate..." Here is EricMusco on 4/8/15: "Our intent behind exposing win rates for Carnage/Combat and Annihilation/Watchman was not to suggest that they are our only data point for decision making ...However, we think the most important part of that data is not who is missing from the top, but who makes up the top. Looking at the top spread of players, almost all of them are spread across only three Advanced Classes: Powertech/Vanguard, Sorcerer/Sage, and Assassin/Shadow. That is the real problem that we are working on addressing right now..." So who knows? Maybe something good, or at least nothing further bad will come of this after all. This data analysis/statistics thing - it's actually not as easy as it might seem. You get paid an 'effing butt load of money if you know how to do it properly. You kiddies still in school keep that in mind. Butt loads. BIG butt loads. Seriously.
  5. For a long time now BW has relied on statistical analysis rather than actual play experience to judge which classes are overpowered and which are underpowered. Given the low staffing and the complexity of the pvp combat in this game, this is not unexpected. However historically, and unfortunately even currently, BW uses flawed statistical analysis to do this play balancing. I have said this multiple times before, but you can not use meta statistics drawn from the entire population for play balancing when the population distribution across classes can migrate and is not fixed. And that is what our player base does - it migrates to the OP classes. Currently BW uses win/loss ratio by class in ranked pvp as its play balancing metric. This is not a supposition. This is BW's avowed methodology. But W/L ratio in ranked is sporked because the OP classes are played by the broadest range of players with the broadest range of player skills, i.e. more lower skill players play them. The most underpowered classes are only played by those most dedicated to their class (and thus tending to higher skill/experience). Attempting to equate the W/L ratio across asymmetric qualities of player bases is why BW's class balancing always fails. The thing is, statistical analysis of class results CAN actually be used effectively. But you can NOT draw from the entire population's results. The correct way to do this is to look only at the teams that win in ranked pvp. Then collate which classes are overrepresented and underrepresented. The overrepresented classes are clearly favored by subset of players that win in ranked pvp - players whose judgements on such matters is probably the most relevant. These are the OP classes. The underrepresented classes are the underpowered classes. But the key is to look only at the data from winning teams. That gives you much better control over the population and makes it less susceptible to population migration issues. Or BW can keep on using meta statistics. Since three straight years of that methodology has led to our current state of affairs....
  6. Tanks in dps gear have been a "feature" of this game since release. What has changed in 3.0 is the ability of a true tank (in full tank gear) being able to do reasonable dps in arena setups. On my tank (again, full tank gear including tank oriented augs), I can do about 90% of the dps of a true dps in an arena - and there have certainly been instances when I have been top dps scorer. But, there is a downside. To really get full dps output from the tank, you need to avoid guard swapping too often, since that eats up GCDs quickly. So the cases where you score high dps are the cases where the other team isn't pressuring your squad, i.e. you are going to win anyway. In arena matches vs. competent enemy teams, my dps drops like a rock as I am constantly guard swapping.
  7. The main problem for Merc dps is the same problem that has afflicted it since 1.3 - almost no one plays the class. The very few that do are the veteran players. They score well and thus the meta averages for the class look stellar. Bioware then concludes no help is necessary for the class. This is true even in solo ranked. I easily win over 50% of my solo ranked matches while playing Merc dps. But I do even better when I am using a real class. Bioware still doesn't understand how player migration across classes taints their in game statistics that they rely on for class balance. It's been this way for 3 years. It's highly unlikely to change.
  8. A fully equipped toon's HP increased by about 40% from lev 55 2.10 to lev 60 3.0. Total damage output increases in the magnitude of ~40% are thus REQUIRED in order to keep the same relative balance between attack and defense. We already know that healing output has increased by more than 40% for sorc and operative healers. The bottom line here is that there has been output inflation with the level cap increase. If your toon is NOT outputting about 40% higher than in 2.10 your subclass either got a (relative) nerf or you have yet to master your new skills.
  9. 1) Downtime is not uniform. The big issue here is expected time until resumption of combat. That is what drives the efficient use (or non-use) of Med Shot during any downtime. We get that you were completely unaware of the relic proc issue. But now even as you are aware of it, you still fail to grasp that even in a combat pause during a boss fight, the time until combat resumption is the key to determining whether you should use Med Shot. Ignorantly spamming Med Shot regardless of circumstances does seem to fit you though. 2) 6-7 damage, not dps, 6-7 DAMAGE is virtually nothing in a fight of any length. It certainly is worth less than a single relic proc. Yet that is what you are gaining by using Med Shot in a tight window before combat resumption.
  10. No, you just said downtime. You didn't say boss fight. You didn't say downtime in a boss fight. It's pretty clear you weren't aware of the relic proc issue and you were one of the guys that spammed Med Shot before a boss fight to get Supercharge stacks w/o taking off your relics first as you were completely ignorant of the side effects. No one is forcing you to be cognizant about the small details of how your character works. And I encourage you to continue to spam Med Shot before boss fights w/o taking off your relics first. I'm sure it will make you happier. Again, it depends on circumstances. Consider that in the example I cited above, you are not forgoing your entire 10 stacks of Supercharge, you are forgoing one stack. In no way shape of form is that worth wasting a relic proc. Once the time horizon until combat resumption lengthens though, the opportunity cost of giving up multiple stacks grows. Also, on a mathematical note is is worth understanding that not using a single Med/Kolto shot prior to combat does not eliminate your Supercharge Celerity/Cell effects, it simply defers it. For example let's say that it takes the Merc X GCDs to go from 0 to 10 stacks, and then consume the full 10 sec of Celerity/Cell. For Arsenal/Gunnery, X might be equal to 20-30. But for any rotation at least equal to X+2, The Merc that did not use Med/Kolto Shot while behind the Voidstar screen still gets all the Supercharge benefits that the Med Shot happy Merc did. Except that for the first two GCDs (presumably Priming Shot and TM/Grav) he didn't have that initial 0.1% buff. You literally are gaining only 6-7 damage by using Med Shot behind that screen in exchange for probably losing a relic proc. Not good.
  11. Disagree. If I have 9 stacks, then yeah I will usually chance the relic proc loss to get that 10th stack. If I have 0 stacks? No. Not unless I have visibility on remaining out of combat for awhile. But at least you have recognized the trade-off here between Supercharge stack progression and relic proc loss. The big variable here is the probability of combat resumption in the near future as that is the event that triggers the penalty for relic proc loss. Early resumption of combat also cuts short your pre-combat stack progression benefits. Let's put this situation in a concrete example that will help to illustrate the issues involved. Let's say you die in Voidstar (your team is defending the first door) and when you respawn you see the door is about to open. You have time for a single GCD before the door opens? Do you use MedShot? If both your relics are off CD, then you most certainly should not. Doing so would give you a 0.1% output increase in exchange for an expected loss of 0.6 relic procs. That is not good. On the other hand if the door isn't going to open for 4 more quarter increments, then yes, Med Shot usage is a good idea. Bottom line here - Supercharge stack progression via Med Shot usage is best done with relics unequipped or on CD. Otherwise there is a trade-off involved and an efficient Merc will take circumstances into account when deciding whether to use Med Shot.
  12. Preparing Supercharge prior to a boss fight is irrelevant to the discussion - which I remind you is about the complexity of IO Merc gameplay. Prior to a boss fight you simply remove your relics, ramp up Supercharge and then put the relics back on. The complexity of managing Supercharge comes from deciding whether or not to ramp charges at the cost of possibly wasting a relic proc during pauses in combat - a situation which occurs in almost every wz. And that decision as will be shown below is NOT in favor of spamming Med Shot regardless of circumstances. And I did not call you a retard. You did that to yourself. Way to not understand the complexity of the picture. At the start of a wz (or pve boss fight), that decision does not have to be made as noted above. But in the middle of a wz, there will often be times when you are out of combat, and have less than 10 charges (or their timer is about to expire). Correctly deciding on whether or not to use Med Shot depends on the circumstances - the claim that one should spam Med Shot all the time is naive and shows a lack of understanding about the class. The best circumstances in which to use Med Shot (for Supercharge buffing), is when both your relic procs are on CD. The worst circumstances in which to use Med Shot is when you are both relic procs are off CD and available. In that case a single use of Med Shot gives you a 0.1% increase in damage output for a 42% chance of losing one relic proc and a 9% chance of losing both relic procs. That is clearly unadvisable unless one has clear visibility on remaining out of combat.
  13. Every time you post, you post about something you have no idea about. I've cast Static Barrier 20 times in a wz, making certain to cast only on full health teammates, and then have gone out of my way to not do direct heals. No healing score in endgame stats. Medals? Does anyone even care about those? The devs balance based on meta average output. Which is biased, on top of being already grossly unbalanced for off healing dps subclasses.
  14. I'll share a secret with you. The best way to use Rocket Punch's root in open terrain is not to turn and run away from an enemy after you use it. That invariably leaves you in range for a Force Choke or Force Push or other attack that negates the tactical advantage you get from the root. Instead I use Rocket Punch at close range when I am moving TOWARDS the enemy melee. Then I go through him. Given the vagaries of server lag, his client will not detect that I am behind him until I am already 5m or so on the other side. Then he must manually rotate to face me. I am usually 10m away by this point. After the Rocket Punch I can use a simple Rapid Shot (or Rail Shot) and then follow it up with Explosive Dart. He will be rooted for the full 4sec, and I will be a good 20m away when the root ends. This is harder to do with IO of course (due to the DoTs), but works fine on Arsenal. Other key uses for Rocket Punch's root - use on bridges/heights/near acid/fire before you Jet Boost so you can be certain your enemy gets pushed in the right direction. Otherwise that aforementioned server lag can cause the push to go awry. Rocket Punch's root is also great when you outnumber an enemy stealther who might try to escape. In those cases get in close and root him. Exception would be if your teammate is a melee, then you might not want to get everyone close in to the enemy and have everyone vulnerable to an AoE mezz. I use Rocket Punch a lot.
  15. It is in fact NOT a good idea to always spam Med Shot. Yes it does give you supercharges. BUT, doing that will also likely put your relic procs on CD. The sudden appearance of an enemy player ( a stealther for example) makes that a dubious strategy. This is not about complexity. I actually like that Merc IO is probably the most complex dps class for pvp. But with complexity comes the need for tactical finesse. If you believe that spamming Med Shot is your best tactic, then you have not grasped the complexity of the subclass.
×
×
  • Create New...