Are PvP drops actually helping PvP?

You misunderstood, or I wasn't clear. What you have suggested there is an extreme version of what using the median calculation already does - gives low weight to the outliers. There is a difference in ignoring outliers and making them not matter in the broader context of the game. Dropping outliers still allows fat cats to get fatter, which means outliers still matter. What I meant is that there's no algorithm we can devise which makes outliers magically go away or get smaller in reality - only balance changes can do that.

Now, here's a dimension of all that which I didn't touch on earlier ... effort to implement. Some ideas there are going to be pretty easy - some of those non-combat hazards, or using mobs that can see stealth, for example. (Stealth may not actually be that big a deal, actually. I speed missions on an Invuln Scrapper with no stealth and his taunt aura on. People speed the LGTF all the time, and those damn drones see everything. But all that's an aside. ) Anyway, the devs would have to be convinced that making outliers shrink is important enough that they would prioritize some of those kind of non-combat challenges, or just make enough of a TF non-combat oriented.

One of the challenges facing them there is that people really seem to like heavy combat. When it came out, the ITF was widley hailed as the most fun TF the devs had created to date (except for the lag). The ITF contains what would seem to be speed throttles in the form of very large foe defeat count requirements. But raw combat is one of those areas people can speed up a lot.

Somewhere the devs have to balance what they think people enjoy against both what they want median (or whatever) times to look like and what sort of outlier distribution they want. If they are satisfied that the median times represent most of the playerbase, they may not put in direct effort to shrink outliers.

That is very considerate of you, Uber.
I've been wondering, who was that player that tried to make his teammates killed and aggroing whole cave, while stealthing to Reichs room on Khan TF.
I've been thinking of you then, but now I know for sure. I'll wait for Vet TP next time, instead of taking gamble whether I can make there alive or not.
/sarcasm

Very solid point. That's why on the one of my ideas that included technology, I mentioned only something that already exists. Application and story use would be a consideration, but the tech is there. Also, they did at least do some work on a detective system. I'd be very curious to know what that was all about. But there are inspirations for it.

But at the same time people complain about "everything is the same, kill all, click glowie". I think a lot of the appeal of the ITF were the unique new enemies and maps combined with exciting combat. The Cimeroan's -def abilities and high damage means you kill 'em fast or get chewed up. That's exciting. Still, it is a good point, the community (and me) do like a good fight. That's one of the areas I've often felt let down, in teh climactic battles. Most games those fights go in stages with the boss using different tactics and powers at each stage. CoH epic battles trigger ambushes and the occasional Tier 9 "not dead yet" power, but that's about it. However, as much as I don't like it, it does favor the casual player who doesn't need to master 2-4 different strategies for each TF and can just tank and spank every time.

Very true. Again, agree completely. Thanks for your insight.

If the distribution is multimodal which I would hazard it is, the median is likely a very bad way to do things. I don't recall if they actually came out and said this but it seems very likely that one of the purposes of merits is to control the rate recipes make it to the general population. The median if the average and the median aren't close the median won't serve this purpose well.

I remember asking this back when we got the news about merits, just how are they counting times ? Are they just taking it as a simple count of tf runs and speed ? Are they counting it on a per player basis ? Do they take a players average run and have it count as a single point ?



That is not necessarily so. It depends on the shape of the distribution. As to gaming the distribution, It would be interesting to see just what it would take to do so profitably. It sounds like an activity people would do to see if they could, not because they would get some net benefit. (Then again if you can get other people to do it for you, its a different story).

The median was said to be chosen specifically because they wanted something that would be less sensitive to outliers than the mean.

It's notable that all the standard tools for collapsing a distribution to a single descriptive number (or just a few numbers) provide poor representations of multimodal distributions.

From what they've said, which isn't much, it sounds like it's based on completed runs. It doesn't seem like it's anything to do with individual players. Failed/incomplete runs may be counted separately and considered as part of the difficulty fudge factor for merit rewards.

If people thought gaming the distribution for profit was easy, I feel certain they would do it. A simple average calculation would be relatively easy to game. Addressing that ease would require the devs to make it a not so simple average, doing things like trying identify and then discarding outliers. My guess is that, rather than figure out the right model for that, they went with something that purportedly did that "out of the box".

I would guess the perception of RV having "higher drop rates" was due to the fact that there weren't many places (before the recent changes to mission difficulty/team size settings) where you could find access to a high volume of lvl 50 mobs that you could kill solo.

Add to this the availability of heavies to allow any build to not only solo farm but do so unattended, starting at level 40(!) and RV quickly gains a reputation as a good place to farm purples. On a low population server you could park a heavy in there and go for hours with no effort on your part beyond making sure you didn't time out.

This is exactly what I've been saying all along, but the median, while being less sensitive to outliers, has a problem with a non standard distribution, which is what we have here. The speeders are at a rank that is far different from the average player times, which means that either the speeders get a fat reward, or the average players get ripped off, depending on which group is completing the TFs more often.

Again, I will state, the speeders will run the desirable TFs way more often than the average completers and will be over represented in the sampling data.

The reward system simply doesn't work well in the way it is implemented, so either the statistics have to be mapped to the reward system in a more comprehensive manner, or a better way of mining the stats need to be done (much more difficult in many ways).

Instead you propose to rework the content to compress the 2 groups, while I was saying to alter the reward system.

Yes, now I am going off my memory here as the original sources were all on the old boards, the devs repeatedly used language in terms of throttling the total flow of items into the game. The average is actually better, as is also the mode for lessening the effect of outliers. The median is probably the worst choice.


This is as always a situation where how you ask the question and how you define your entities can have a gigantic bearing on the outcome. If you say take grp task force times as your measure, you are saying players are indistinguishable entities. This really doesn't matter if your goal is to deal with the aggregate.

Anyway its no secret I think they have their counting off. If you look at the statesman tf, the lord recluce, or the abandoned sewer trial they have much higher failure rates, and the potential to be horrendously long but have very low merit rewards. Its a scratch your head kind of moment.



There are over 100k people that play the game, on 11 servers in the U.S. alone, even if you were to say that 10% of the people acounted for 75% of the tf runs it would still take a major effort to move the average in one TF. Seeing as ? they use data over a long time horizon ?? the task would be even harder because they would have to overcome historical data.

If you're the outlier, all you have to do is balance out your own best times. If you are running a 2 hour average TF in 30 minutes, all you have to do is make sure you add about 4 hour run per 30 minute run to basically erase the effect of your own best times. In fact, you'll slightly deflate the average doing that. If you make the offsetting runs 8 hours long you only have to do it once every 5 or 6 times to erase your fast times, meaning that you could run the TF most nights of the week and finish one spent logged in overnight just once to counter your efffects on the average. Of course most people would turn "logged in overnight" to logged in for 24 hours, which would erase the 30 minute runs if you did it only once every 16 times.

If that sort of thing became common knowledge, I think enough people would do it that they could keep a TF from being adjusted to reflect gradual increases in speed. Of course, if it became common knowledge, the devs would probably notice and try to do something to the calculation to adjust for it, either tossing out gross outliers or coming up with a way to tell if characters were logged out or sat in bases for most of their TF run. If the devs don't do any of those extra steps today, that would all take either greater analysis or new tech, or both. It becomes easy to see why they might try to pick an algorithm that does this work for them. Whether they succeeded or not is, obviously, up for debate.

The straight mode calculation seems an unlikely candidate, as TF times are likely to all be unique. They could round values to something like nearest minute or nearest 5 minutes and find the mode of that. You still need a tie-breaker if you end up with a true multi-modal result.

Lets look at an example

100 tf runs avg time 120 minutes for the sake of argument 1 person is running 10 of the 100 runs at 1/2 hour. This gives 12000 minutes as the total amount of time and 11700 minutes as the time the rest of the people take or 2 hours and 10 minutes without the speedster. If he balances out his runs the total goes to 14400 minutes and 110 runs. This shifts the average to 131 minutes/ run. The median though is most likely shifted more. In the original case, the median will be the run that comes in 40th after him. By just running the tf at slower than the average he can move the median to the person that is 45th after him in half the additional time. Seeing as the as the average for everyone else is 130 he is likely to get more bang for his buck doing this, depending on the shape of their curve of course.

Whats more lets say someone finds a very speedy way to run tf x and doesnt want the rewards to change. All they have to do is for every speedy run they do slip in one run that is just slower than the average and the median doesn't move

It seems like you want a game where everyone has exactly the same chance to win in a competitive situation. The only 'game' I'm familiar with that would satisfy those requirements is coin flipping, which imo, isn't very fun, lol.

In any game you compete in, be it computer game, board game, sports game, etc., there will be 'imbalances' based on skill, knowledge, equipment, and ability. And that imbalance degree will be greater based on the complexity of the game, the more complex, the greater the imbalance. If the devs dumb this game down to 'coin toss' levels, who will still play it? I won't.

In regards to you talking about people thinking they have the 'right to wail on people because they paid their dues', that's just how games work, hell, that's how life works. The more someone works at something, the better they get. And someone who has worked at pvp for a long time is going to 'wail on' someone who is trying it out for the first time, and that is as it should be. Who wants to play a game where a total newb has the same chance of winning as someone who has played the game for years? Who wants to play a game where a player who has spent zero effort setting up his character has the same chance in a fight against another player who has spent months setting up his character? If that's for you, again, let me suggest you go flip some coins, lol.

And I'm not one of the pvp experts, I'm one of the newbs who goes into zones and gets wrecked by the pros. But again, that is as it should be. If I want to get good at pvp, then I need to put the time in to learn how it works and build a toon that can compete at the highest level. That is what competition is all about. It's not about everyone being equal. If everyone was equal, competition would be pointless, kinda like flipping a coin...

But with pvp as it is now, I'll probably remain a newb. First of all, the way the devs made powers so drastically different in pvp means I need a completely different build, which means I need to outfit a toon twice. The extra effort of doing that and having to learn a whole new set of rules for powers just isn't worth the effort imo. The pvp io's are great, and they absolutely could lure me into pvp, if they dropped WAY more often.

And ultimately, I enjoy playing melee toons, which have always been at a severe disadvantage in pvp. And the 'don't bring a knife to a gunfight' argument doesn't fly with me in a superhero game, or you'd have the hulk getting owned by wasp in every comic book. I realize the movement changes were in part an attempt to close this gap, but clearly it was the wrong way to do it. As others have said, it's better to bring the bottom end up than drag the top end down with universal movement suppression. One way this could have been accomplished (without changing everything and ruining the movement so many pvp'ers enjoyed) would be to give some melee attacks range by mimicking shield charge. With some melee attacks, you could target an enemy at range then 'auto charge/tp' and attack them so melee at's don't need to be right next to people to initiate an attack. Granted, true ranged powers would need to retain an advantage, but 60-80 ft ranged vs 5-7ft melee was way too big a gap, and still is actually, even with the gimped movement. With my suggested change, you could retain the movement everyone loved and make melee toons viable in pvp. Instead the devs gimped the whole system and pissed everyone off. The worst part is they are seemingly too proud to admit their mistake and rethink their actions.

No what I was suggesting was at the start of the game everyone gets to start with the same material. I also pointed out that if you force people to play another game just to play the game they may or may not want to and many less will participate.

If you think there is no skill involved making a build, and playing it doesn't involve any skill then you are correct, starting everyone at even in material would be like coin flipping.

I think what AF was saying is what I've seen posted and said in-game by MANY pvpers: they don't want to have to level up a 50 and then IO out the toon just so that they can pvp. To them pvp IS the real game and PVE is a complete waste of time.

The only mmo to do something close I think was WoW and it's arena system, and maybe Guild Wars (though not exactly).

HOWEVER, with that said I don't see this game doing that EVER. This game was/is/always will be a pve game FIRST, with pvp tacked on. Yes, pvp was planned from the begining, yadda yadda yadda, I've seen the vids. HOWEVER the IMPLEMENTATION makes it pretty clear that the planning for pvp "from the beginning" was done with no REAL planning (read: AT balance and figuring out what type of pvp system they actually wanted in the first place) at all.

The more realistic suggestion is for cross server pvp, or at least cross server arenas.

As an aside, the pie-in-the-sky thing I've always wanted for this game was cross server, everything: arenas, pvp, pve (at least tfs). They once mentioned a serverless environment as something they MIGHT look into. Haven't heard anything about it since it was mentioned during the time that they announced server transfers. (And if memory serves, they mentioned it as a "just in case" for people who later complained that they "wanted a refund for my server transferring since now there are no servers waaaaahhhh" )

they do.

Wait for DR 2.0 on that.

Edit: That was unduly pointed.

Try and look at things from the point of view of NcSoft and the PvEer. The PvEer is playing his game and thinks he might like to PvP, he goes in the zones and the first thing he notices is the NPCs are killing him and when he talks to anyone about what he needs to do, he is told the ante is a new build. From NcSoft's view they have something that they have sunk resources into that nobody is using.

What do you get ? The obvious advantages got nerfed. If you recall they even tried to nerf teaming.

The PvEer is just behaving exactly the way the PvPer does. If you know you are going to lose a fight, you don't fight or you evade.

wrong



I was a pver. Long before I was into pvp i was one of the top badgers in the game. I soloed and teamed many toons to 50. Please do not act like we all came to this game as pvper. In truth everyone started as a pver and we call had to go through the learn curve. As far as seeing it through NCNC's eyes those genius actualy figured out a way to make pvp harder to learn for new pvper so I don't think I will go at it from their angle.


Obvious answer avoid the npcs.



Guess what, you get the same answer for tweaked farm builds and av soloing. Why should pvp be given a different way to work over these. If you say you are asking for pve builds to given open access to all enhancements as well as pvp you may have a point, otherwise not so much.




Oddly enough more people used it before they 'sank resources into it'.




Like I have said before and will say again in hopes you get it. The problem with lack of pvpers in this game has more to do with the pve centric nature of the game. It has 0 or close to it to do with game mechanic or loot disparity. If they want more pvpers the only really way to do it is to turn the entire game's direction and recruitment of players to pvp. I honestly don't ever se that happening.

The attempt at anti spike code was an obvious attack on teaming, or at the very least an attack on coordination and communication. Recharge, plus defense, plus to hit all got nerfed.

Anyway as long as the buy in for PvPing is in the same category as good farm builds and GM soloing, PvP will be a dead letter in this game. I know people that have been playing this game for years and have never soloed an AV or a GM and dont particularly care for farming. I don't have hard numbers but it really wouldn't surprise if they were the overwhelming majority. For you it may be nothing to do this for most people its considerably more difficult.

Now if PvP were overwhelmingly fun to the point, that it became the prime attraction of the game, yes you would find more people willing to slog through the hoops to play. At that point the people at NcSoft would be asking themselves why they have barricades on the road to their big attraction.

Your entire last paragraph is why you're never going to see what you're asking for. PVP will NEVER be THAT attractive in this game.

As I stated before, the more REALISTIC suggestion (always has been) is for cross server pvp. Or at least cross server arenas, if not zones.

This argument has been going for years. The number of folks who started playing this game cause it was the only super hero mmo on the market, was divergent from Evercrack, was divergent from typical fantasy has always been > than those who came for an excellent pvp system.

Cause the arenas most certainly were NOT an excellent pvp system when they first came out. No one can possibly tell me the original arena system was what drew them to this game. And I'm talking about the original one pre-ED, pre-COV.

The words balance can't even be spoken in the same sentence as the original pvp arena system. Nor can the words fun be spoken with it in the same sentence, since this was the original arena system that used to CRASH every other match, and there was no control over what map you got. And (ofcourse) there were no villains.

Bottom line (to get back on point) is the best you can ask for is a cross server pvp, cross server arenas.

The most that we can assume from the anti-spike code is that it was meant to stop spikes. You were not punished for being on a team, however if you all targeted the same person, then the anti-spike code came into play. A Blaster could team with 7 Empaths (All targeting and buffing the Blaster) and the Blaster would not be at a disadvantage. Perhaps a case could be made if you're talking about discouraging team set-ups like a team of 8 Blasters, however they are not at a disadvantage by the specific act of teaming.

Note: I am not saying that I don't believe teaming was discouraged in I13, however we cannot directly infer that from the anti-spike code.

By the way, how was Recharge and Defense nerfed? Are you talking about Diminishing Returns now, because as I recall the Anti-Spike code only nerfed your To-Hit (or was it Accuracy?).

I believe the sticking point will be the way you worded the PvPer's response in the first place.

Very rarely does someone wander into a PvP zone with the best PvP sets, however the current FotM is not needed in order to do well. There are a few sets that won't do as well as the best sets in PvP, however any combination can PvP.

I have been following a thread recently about Meleer's soloing AVs. It basically went like this.

Poster: I want to make a Dual Blades/X Scrapper. I want to be able to do, X, X, and solo AVs with it. Is this possible.

Response: Sure but your fights are going to be long. You will be doing just over the AV's regeneration so be ready for 20-30 minute fights. Scrappers aren't Illusion/Radiation Controllers.

The answer from the PvPer's to the PvEer should have been "You will need to respec your build to optimize yourself for PvP. If you want the 'best-of-the-best' build, make an X/X."

Before it comes up things won't translate exactly from AV soloing to PvPing. While it could be said they are in the same category, AV soloing is basically a can or can't thing whereas PvP can have differing levels of success at different times.

The ability to coordinate attacks and take out targets is an advantage teams enjoy, the better the teamwork, the greater the advantage. Anti spike code hit directly at this.


DR is under the category of obvious advantages. More or less the devs looked at the situation and said that access to IOs and sets makes things incredibly lopsided, so lets just make them not nearly so useful.



These days you can leave the rest to just

Quote:

Before it comes up things won't translate exactly from AV soloing to PvPing. While it could be said they are in the same category, AV soloing is basically a can or can't thing whereas PvP can have differing levels of success at different times.

Misses the entire point of if you put pvp into the same category as av/gm soloing you can expect the same level of participation.

I think you are right that PVP will never be the big draw. PvP and even PnP RPGs rarely worked. Player ats are always balanced to be overwhelming to NPCs but almost never to be equally powerful against each other. What works for individuals wont work for a team etc.

Cross server is a great idea, it would at least increase the apparent userbase and allow much more play. I don't think we would get it though as nothing in the game except the market and chat is cross server currently. Just at a guess they would need to set up a dedicated pvp server to make that work and then implement code to move your instance to it.

Probably a similar server to what they have for other cross server functions, ya. I would however use it for cross server tfs so that there is a double benefit for it. Just in case the pvp pop (who are we fooling, the tf pop would just be larger than the pvp pop. ) was not large enough to justify the investment.

• Coordinating attacks is something teams can do.
• Not coordinating attacks is something teams can do.
• Coordinating attacks is something people not on a team can do.

My point is that the Spiking, multiple people attacking the same target, was the target of the Anti-Spike code. Spiking is something teams and non teams can do, and things that teams and non teams can choose not to do.

The Anti-Spike code impaired the ability of multiple people attacking the same target.

Examples:

1) On occasion I play my Stalker and a friend of mine plays his Dominator. When he gets on he is usually invited to teams right away, however I like to play without a team. I usually play near him in order to kill targets that attack him. I know how he plays and know when to jump in for an Assassin Strike to help him kill his target (ie. We have the same target). We're not on a team*, however if Anti-Spike Code was in place we would be nerfed.

2) Quite frequently I play my Empath when a friend of mine plays his Blaster. My Empath is /Energy Blast so he cannot provide nearly as much damage as my friends Psi/EM. I keep my friend targeted and from the air I buff and heal him. We're on a team*, however if the Anti-Spike Code was in place we would not be affected.

*Can be replaced by "We're working as a team" and it will mean the same thing.

Alright, I was just wondering because it looked like you were saying that the Anti-Spike code nerfed defense and recharge.


What I meant by that was people shouldn't take:

And change it into:

It had nothing to do with participation level.