Thursday, April 27, 2006

Wishlist Loot System

In a post on the WoW Raid and Dungeon forums, Coracus of Infliction (Dragonmaw server) outlines his guild's rather unique loot system:
Wish list system:

Each person gets to put their most wanted 6 items on their wish list. You cannot change the wish list once you have made it. Items on the wish list can only be changed once either A) The item drops and you don't get it, or B) the item drops and you do get it.

The order of roll allowance goes as follows:

1. Wishlist (X class/DPS/Healers) please roll:
2. Need Guild Members please roll:
3. If no one Needs, then Shard

If the person on the wish list decides to pass it, then can give up their roll to another person on the wish list. They cannot just pass it to any person, you can only give up a roll. Meaning it goes to open roll if no one has it on their wish list.

I find this system works very well with a close group of raiders. It's been much more enjoyable than DKP, and we've used it right through Nefarian.

Why I dislike DKP, is it adds to that feeling we constantly have about "Grinding". You have to grind dkp just to bid on items. In our guild, we all know the loot will come as long as everyone shows up on time and raids. The wishlist system lets you concentrate on having a good time and knowing you always have a chance to roll on the items you want,instead of thinking "well every other person has X dkp, so ill never get anything".

I find this is a very interesting endgame loot system. For one thing, I believe it is the first endgame system I've seen which is memory-less. I didn't think that this was possible or even desirable.

Secondly, it forces you to think about loot differently. Narrowing down to six items makes you think ahead, and think in terms of the larger picture. Also, you are no longer competing against everyone who can use an item, but only those who value it enough to be in the wishlist.

As well, everyone's wishlists are public, which means that you see the larger picture and goals of the entire guild. People's lists being public also allow you to meta-game, and choose items which are in less demand.

The overhead of this system is also extremely low. Tracking DKP for an entire guild can be time-consuming. This system is very easy on the officers. Additionally, wishlists are posted in advance, which allows officers some measure of time to detect and head off problems.

Of course, the system does have its flaws. For one thing, you cannot guarantee winning an item. It is entirely possible that someone will constantly lose rolls on a specific item, which may lead to feelings of unfairness. And it is probable that a raider with less time spent will occasionally beat a raider who raids a lot.

However, I think that overall this system is a very good one. For one thing, I think it promotes a healthier attitude towards loot and progression, and forces players to decide what is really valuable to them. The low overhead is a serious benefit, allowing officers to concentrate on more important things. I commend Infliction on their innovative system.

Wednesday, April 26, 2006

People Afraid To Try

There's one type of person in Wow who really get on my nerves. Basically, these are the types of people who actively speak out against trying something if the raid or group doesn't have the 'optimum' makeup.

For example, our guild was at Garr in Molten Core the other night. Garr has 8 adds. The 'normal' strategy is to tank four of the adds and Banish the other four. However, we only had 2 warlocks to banish. We did have 6 or 7 warriors and a strong amount of healing, so the fight was possible with a change of strategy. However, one player spoke up and declared that the fight was 'impossible' without the extra warlocks, and we shouldn't even try.

This timidity is the one thing that really bugs me in WoW. Why not try? The worst that happens is we wipe, and we count that as a lesson learned.

The next week, we had enough warlocks to do the fight normally, and we killed Garr. Thinking about continuing on to Baron Geddon, that same player says that we don't have enough Fire Resistance in the raid to continue. Argh! [1]

If we tried, wiped, and then came to the conclusion that we needed more Fire Resistance, that would one thing. But to be unwilling to even try?

You see this even in lower instances. People who won't do anything if the group makeup is not exactly correct according to perceived wisdom. Or maybe certain people have the wrong talents (Feral vs Resto druids, Shadow vs Holy priests, etc.).

It's odd, but whenever I've grouped with people from the very elite, cutting-edge guilds, they don't seem to adhere to this 'common wisdom'. Maybe it's because what I'm working on is trivial to them, but they seem to be more willing to try less conventional solutions.

[1] I do have to say that the player in question is a very good, solid player. I just wish he'd be a bit more willing or enthusiastic about taking risks.

Saturday, April 22, 2006

The 'Men In Tights' Problem

If you wander across the blogs of MMO game devs, you see a lot of reference to the 'men in tights' problem. The vast majority of MMOs are fantasy games. SirBruce, of www.mmogchart.com, estimates that 89% of all MMOs are fantasy-based. So it's a pressing question (for game devs, anyway) of why this is so, and does a non-fantasy game have a chance of succeeding.

There are a lot of explanations proposed, from the fact that our culture is steeped in the past, or that the first MMOs were fantasy, so that's the way it is. I think there is a different reason that most games are fantasy. The fundamental conceit of rpgs is that:

As I do stuff, I become harder to defeat.

In fantasy games, you have all sorts of explanations for this, from magical healing, to better armor, to better skill with weapons, to just being physically tougher than the other guy. Levelling up represents becoming tougher and better skilled. In a swordfight, we expect the more skilled fighter to come out alive. And a lot of the attraction of RPGs relies on this: our working on a character and slowly making them more and more powerful.

However, as you introduce technology, it becomes a lot harder to justify this basic idea. The entire point of technology, in a lot of ways, is to replace skill. There is an old saying, popular in the eighteenth century, that "God made all men, but Sam Colt made them equal." This saying perfectly encapsulates the point that technology makes the unskilled deadlier. Non-fantasy games need to be able to account for this. If you have a game where a character can get shot with a gun at point-blank range and almost always survive, that's really hard to take.

Now, that's not to say that non-fantasy games can't do this. Take Eve Online. Eve Online is pretty much as non-fantasy as you get. Eve Online, however, replaces "I" with "my ship". It makes sense that a better, more expensive and technologically advanced ship is harder to defeat. That fits in with our worldview. And thus Eve works as an MMO, where "levelling" involves improving your ship.

However, it's easier for a fantasy game to account for this. And this is the main reason, I think, that fantasy is the most popular genre. The levelling idea, of character improvement making survival more likely, fits fantasy better than any other genre.

Team and Individual Skill in PvP

In the comments to my last post on PvP, Thoma writes:
Again, it all comes down to you assuming that team skill/number of players = player skill.

This isn't quite what I'm assuming (though if you amend it to *average* player skill, it is correct). I'm assuming that team skill is equal to the sum of the individual players skills.

Is this a good assumption for WoW Battlegrounds? I think it is. For example, if Alice and Beth beat Joe and Harry, but Beth and Carol lose to Joe and Harry, we can infer that Alice is more skilled than Carol. That seems like a reasonable result to me.

Now, there are other games where this assumption does not hold. For example, you could have a game where team skill is equal to the skill of the worst player. Is this a good model for BGs? I do not think so, because it implies that a team with 14 people will always lose to a team of 15 people. (The missing person is essentially a player with zero skill.) And this is not true for BGs. A team with 14 people is at a disadvantage, but the skill of the remaining players can make up for it. So this model is not good.

How about the opposite model? Is team skill equal to the skill of the best player in the Battleground? Again, I don't think this model holds, because it would imply that in a 1v15 match, if the one person was slightly better than each player on the 15-man team, she would win. And I don't think this is true in BGs. The most likely result is that the one person is going to lose, regardless of her skill.

So I think it's pretty clear that team skill in BGs is some combination of the skills of the individual players. And that means that we can infer individual skill by looking at how different teams with that individual perform. Which is what a system like TrueSkill™ does.

Is this model 100% accurate? No. For one thing we occasionally see teams which are greater or less than the sum of their individual parts. A classic example is the 2004 USA Olympic Men's Basketball Team. But for the most part, the model is pretty accurate, especially when players change teams reasonably often, as is the case with WoW Battlegrounds.

As an aside, I think a more accurate model of team skill would be to say that each pair of players on a team has a bond, and that bond may have some skill, positive or negative, associated with it. Two players may play especially well with each other, or two players may play poorly with each other. Then team skill becomes the sum of the skill of the individual players plus the sum of the skill of all bonds formed. In graph theory terms, under the first model only the vertices in a team have weights, but in the second model, some of the weight of the vertices shifts to the edges in the graph.



And individual skill in my first model is not just individual skill, but individual skill plus the sum of the skills in all bonds with that player multiplied by the probability of each bond occurring. But the first model, where only the vertices have weight, is a *lot* easier to deal with, and is probably accurate enough for the purpose of PvP ranks.

Wednesday, April 19, 2006

Don't Anger the Kodo!

I'm thinking about some of Thoma's points in the PvP thread, and hopefully will post a response soon. But in the meantime, here's a fun World PvP story that happened today:

My Forsaken warlock alt was questing in Desolace. At one point, I run into a Nightelf hunter and a Human mage, both of them around my level. I didn't really want to fight, especially outnumbered, as I just wanted to get to the town and hand in a quest before logging off. So I wave to them, hoping they'll let me go.

No such luck. They back up and start attacking. My succubus Seduces the hunter, and I start casting spells at the mage. I'm pretty sure I'm going to lose the fight eventually, when the mage runs in and starts casting Arcane Explosions.

Unfortunately for him, his Arcane Explosion hits a nearby herd of kodo. Enraged, the kodo charge over and promptly stomp him. Then they go and crush his hunter buddy, before wandering off.

So, the moral of the story:
  1. Focus Fire.
  2. Don't Anger the Kodo!

Monday, April 17, 2006

Comments on PvP System

Thoma points out some issues with my proposed system, and I've decided to make my responses into a full post, to expand upon my idea.

1. Skill isn't rewarded, just wins/losses. Being AFK is even more rewarded under your system then the current system. Since all honor gained is win/loss rather then a mix of win/loss/HKs then you just turn AV into a bot farm.


No, because a loss means your rating goes down. That's the difference between now and then. When you go afk, you basically have a 50% chance to go up and down, which becomes less than 50% because your team is effectively outnumbered.

To go afk and gain in rating implies you were on a succession of teams that consistantly won, even though they were a man down. That seems improbable, at the very least. If it was an organized team, the leader should recognize that you aren't contributing and boot you.

The reason botting works now is that there is no risk to just sitting and collecting HKs. In the current system, if you spend the entire match fishing, you will end up better than you started. Guaranteed.

2. The stake to the heart you propose world pvp is going to really marginalize the PVP servers. It just makes it into Gank fests that don't end due to diminishing returns.


Perhaps. It won't be a gank fest though. It will return world PvP to the same state as before the honor system (and to be honest, roughly the same state it is now, at least from my latest experiences on a PvP server). As well, each goal would only be achieveable once per week, hopefully minimizing the disruption.

The real gankfest was when we had the Honor system but no BGs. That was a terrible time. I was in the high 40s on a PvP server, within range of the 60s. It was supremely unfun. It's why I have a character on a PvE server now.

My system has no rewards for killing individual players, only for achieving goals, so people will only gank for the old reasons of boredom, opportunity, malice, and fun.

3. You assume that all servers have enough PVP to make seperating out "skill groups" worth while. And what happens to the Grand Marshals? Do they only face each other? Or do you have to wait for 40 of each groups high ranks get together for an AV?


Yeah, to be able to separate in skill implies sufficient numbers. This is probably not possible for most servers at the moment. However, Blizzard is working on cross-server BGs, and it would be a good solution then.

As to high ranks, the players closest in rating play each other. If there were only two teams, they would play each other, even if there was a great difference in average rating. However, the better team would only increase rating by a tiny amount if they won, and the losing team would only lose a tiny amount. However, if the weaker team won, there would be a larger gain/loss.

It's basically the same system--in principle--as the ELO chess rating system, or Microsoft's Trueskill™ system.

Sunday, April 16, 2006

PvP, Skill, and the Honor System, Part II

Before we get to my idea for the honor system that rewards skill instead of time, we should define what skill is. In my view:

Skill is the ability to defeat an opponent in a fair fight.

Seems logical. If two characters fight each other, the more skilled player should win most of the time (all other things being equal).

The hard part, in World of Warcraft, comes with finding a fair fight. If a level 50 is attacked by a level 60, is that a fair fight? If a level 60 is attacked by 2 level 55s, is that a fair fight? What about if a character is fighting a mob, and another character jumps him? What if the character fighting the mob is higher level than the attacker?

Also, when does a fight end? In the zerg at Tarren Mill, how do you tell when one side wins? Is each individual death the end of a fight, or does it matter if one side pushes the other side back?

If you look at all the possibilities, it seems very hard to actually determine if a fight was fair or not, especially in world PvP. However, if we can see that Battlegrounds are reasonably fair. Both sides start with roughly equal numbers, level, and position. There is no interference from outside. There is a defined starting point and ending point. For the most part, the team that wins the Battleground is more skilled than the side that loses.

So my proposal is to make honor dependent on performance in Battlegrounds.

The basic idea is that each character has a PvP rating, like a chess rating. When a character's team wins a Battleground, her rating goes up proportional to the combined rating of the opposing team. If she win a victory against a good team, her rating increases more. If she loses, her rating goes down.

Then rating determines how fast you progress through the ranks, and what the maximum rank you can attain is. Each week, during Tuesday maintence, your increase in rank based on rating is applied. You would need to play a certain number of matches in the week (say 3) to be eligible for a change in rank. A person who has an average rating might only be able to reach Rank 5 (and take 8 weeks to get there). A person one standard deviation higher might reach Rank 8 in the same time. Rank 14 would need a *very* high rating (three or four standard deviations, probably).

There are a lot of benefits to this proposal. First, ratings are an intuitively simple idea. Everyone understands them. If a Grand Marshal is someone who wins all the time, that just makes sense. Second, since your rating changes with each game, you don't need to play a great many games. However, you need to win in the games you do play.

Third, since ratings depends on victory rather than kills, it also rewards players who play for the victory. People who heal, who play defense, who use strategy and tactics. Leadership becomes a valuable asset, as leadership leads to victory. Right now, I find that a lot of Rank 13/14s aren't really leaders, but loners who just rack up kills.

Fourth, since you now have ratings, you could match players by skill level. The game could match players of equal skill level. Higher rated players get to play other higher rated players.

Now, there are some negatives to this proposal. People who solo and join pickup groups will be at a disadvantage compared to people who run in teams. Matching using rating could moderate this a bit.

The biggest downside though, is that it will completely kill world PvP. World PvP just does not fit into the rating system. You simply cannot guarantee the "fair fight" that the rating system depends on.

The only idea I have to encourage world PvP is to have faction-wide goals. For example, if one of the six Race leaders (Thrall, Magni, etc.) is killed, everyone in the enemy faction gains an extra quarter-rank on Tuesday, and everyone in the friendly faction loses a quarter-rank. By making the gain or loss faction-wide, there is extreme incentive on both sides to help. Otherwise, how would you know which defenders to penalize? Making it zero-sum ensures that people don't simply trade leaders. As well, it may promote larger strategies like having one raid feint at one leader while another raid goes after a different leader.

Perhaps there could be other goals. Maybe each small town (Southshore, Tarren Mill, etc.) could have a flag like the nodes in Arathi Basin. If a flag is held by the opposing side for a full 24 hours, the town is declared "defeated" for the week and everyone gains and loses 10% of a rank on Tuesday. (Nothing changes in town other than the flag, all npcs and quests would stay the same.)

However, even if something like the world goals is not implemented, I think that performance in the Battlegrounds should be the primary determiner of skill and rank. It is as close to a fair fight as you will find in WoW, and is a much better and simpler measure of skill than the current system.

Saturday, April 15, 2006

PvP, Skill, and the Honor System, Part I

The Honor System in World of Warcraft is fundamentally flawed. Most people expect the Honor system to measure skill in PvP, and it really does not. A Grand Marshal or High Warlord should be the most skilled fighters on the server, and often they are not. The flaw in the Honor System can be summed up as follows (Coriel's First Law of Skill):

If the metric used to measure skill cannot decrease, you are not measuring skill, but time.

Intuitively, this idea is pretty clear. In most other games, your skill metric can go down as easily as it goes up. Consider poker, where you can measure skill with money won. A good player wins more money than an bad player. In fact, a bad player often loses money. And this possibility to win or lose money occurs on every hand. If you could never lose money in poker, the person with the most money would be the person who could play the longest.

Another example is the batting average in baseball. You get a hit, your batting average goes up. You strike out, your batting average goes down. Or how about chess? You win a match, your rating goes up. You lose a match, your rating goes down.

Pretty much every game other than WoW gets this. Yet in WoW's PvP system, your honor total cannot decrease. If you go into a Battleground, or attack another player, you will never end up worse than when you started. And so the people who gain rank are the ones who play the most, not necessarily the most skilled players.

Now, there are all sorts of potential ways to change the metric so it more truely measures skill. For example, you could do Kills minus Deaths. Or Kills per Death. Or Kills per unit time played. Or some sort of rating system, where if you are defeated, your rating goes down proportionally to the rating of the one who killed you. The important thing is that each PvP encounter must be a risk, that your honor should have the potential to decrease as well as simply increase.

Of course, being a random anonymous internet pundit, I have a preferred alternative to the current Honor System. That will be the subject of Part II.

Wednesday, April 12, 2006

Epic Charger!



At long last, I finished the paladin epic charger quest. I've had this quest for so long, and failed it so many times. After a while, I just stopped attempting it.

Today I joined a guild group trying the quest and we rocked Scholomance. We had one heart-stopping moment during the 3rd wave of spirits when three people died. Luckily we got some space and were able to resurrect them.

Then we faced Death Knight Darkreaver. I used my Holy Mightstone, which I had been saving for this day for months, and we just destroyed him!

Awesome, awesome day!

Tuesday, April 11, 2006

Guild Philosophy

I joined another raiding guild in the end. This guild is less hardcore than my previous guild, and I find it suits me a bit better. For one thing they let the paladins melee in Molten Core!

(Honestly, I think the leadership is new to raiding in general, and don't realize that most guilds believe paladins are supposed to stay at the back. And we paladins are not enlightening them. :) )

Getting to melee makes Molten Core so much more fun! I find healing and cleansing can easily be done from the front lines. There are times when it's better to stay back (for example, when I need to help keep the main tank up), but the guild leaves the choice up to the individual paladin. This is actually quite liberating. I find that having the option to melee makes the game more fun, even if I choose not to take it.

Now, I'm not sure that paladins meleeing is the absolute optimal strategy. We are only on Gehennas. However, we have problems getting more than 30 people in the raid, and we did take down Magmadar for the first time with only 33 people. That seems like a pretty good accomplishment, so I'm willing to reserve judgement. And I'm having a lot more fun, so that has to count for something.

In all honesty, I think a lot of the problems at the raid level, from loot issues to class behaviour stem from the arrogance of some raiders. A lot of them believe that they know how to play your character better than you do. That's why they impose restrictions on you and on themselves. They do not trust you to make the correct decisions for your character, and so they try to make the decisions for you.

A good guild, in my opinion, needs to adopt the view that: The person who can best play your character is you. In the end, it all comes down to trust. If you trust that your people know how to play, you trust that they will only take loot they will use well.

There's a quote I've always liked. It's possibly a bit extreme for something as mundane as guilds in an MMO, but for what it's worth:
"You? I know you! You trust beyond reason."
She met his eyes steadily. "Yes. It's how I get results beyond hope.'"
- A Civil Campaign, Lois McMaster Bujold

Monday, April 03, 2006

Paladins in PvP

Paladins in PvP is a very complicated topic. Paladins are deeply flawed in PvP combat, but in many ways the flaw is not with the paladin, but rather with the surrounding structure.

I believe that a skilled paladin has a good chance against every other class in one-on-one combat (except warlocks, no one can beat warlocks). This is especially true if the paladin plays to her strengths. The best strategy for a paladin is to be a rock, and let your opponents break themselves on you.

Let me illustrate this with my strategy for fighting mages:
  1. I generally turn on Fire Resistance Aura and try to close to melee range using a 2H weapon.
  2. The Seal I use for most of the fight is Seal of Justice.
  3. I usually put up Blessing of Light, and switch to Blessing of Wisdom if my mana gets low.
  4. When I take damage, I start healing very early (around the 80% mark) with Flash of Light. The idea is to heal through the damage and run the mage out of mana. When he's out of mana, I can kill him at leisure.
  5. Save Hammer of Justice until after the mage Blinks out of a Seal of Justice proc.
  6. Save Divine Shield (bubble) for when the mage sheeps you, but don't use it right away. Stay in sheep and let your health regenerate. Most PvP mages will start a Pyroblast when you are sheeped. You need to wait until the Pyroblast is almost finished casting, then bubble out of the sheep and take the Pyroblast on your bubble.

Anyways, following this strategy I can usually beat an equally-geared mage more than half the time. Occasionally the mage will get a string of crits and blow you away, but it's a reasonable fight.

The problem, however, is that the fight takes three to four minutes. All paladin fights take a long time. All the other classes kill quickly or are quickly killed. The paladin's strength is her survivability and ability to outlast her opposition. Unfortunately, the way the Honor System and Battlegrounds are implemented really hurt the paladin.

The Honor System rewards those classes which can get a lot of kills. Since there is no penalty for dying, a character who can kill someone really fast is at an advantage over someone who kills slowly. Every class can spec for high DPS except the paladin. So every class can rack up honor faster than the paladin.

The other problem is that a paladin has high survivability, but low DPS. However, the paladin's survivability drops dramatically if faced with superior numbers. For example, if I am guarding a node in Arathi Basin by myself, and one Horde player comes by, I can draw out the fight for a very long time (several minutes if need be). Long enough for reinforcements to arrive. However, if two Horde players come by, I will die *very* quickly. I will die just as fast as any other class would. The extra survivability does not help in that case.

(The exception to the above is if I have the bubble up. Then I can stall for 12 seconds, and then die immediately after.)

So the number of times that paladin survivability actually plays a role is a lot lower than it appears. In my view, it only makes a difference in equal number fights, which are fairly rare.

The implementation of Battlegrounds also hurts the paladin. Survivability means that you do not die easily. But what happens if you do die? You are kept out of the battle for an average of 15s. This is a very short time when you consider how long it takes a paladin to kill someone. The penalty for dying is very low and this devalues the effect of survivability. Dying a lot is okay if you can kill a lot before you go. I suspect that if the resurrection timers were increased from 30 seconds to 2 minutes (giving an average time dead of 1 minute), paladins would become much more powerful in PvP.

Battlegrounds are designed for the high DPS style of the other classes, and thus the low DPS style of paladins is disadvantaged.

Saturday, April 01, 2006

Reckoning

I know this verges on heresy for a paladin, but I really don't approve of the Reckoning talent. I appreciate the fact that Reckoning-bombs are very powerful. Yet it feels like a bug, one which the developers left in because paladin PvP melee is useless.

A bit of background first. Reckoning is a 5-point talent in the Protection tree:
Reckoning
Requires 20 points in Protection
Gives you a 20/40/60/80/100% chance to gain an extra attack after being the victim of a critical strike.

This talent seems like a tanking talent, or one that works when you are trading blows with an enemy. You're in melee, you get critted and you wack the mob back with your free attack. Except that's not quite how it works. You don't actually get a free attack. Instead, the time to your next weapon swing is shortened to zero. Sometimes, if you've just swung your weapon, this is close to a free attack. Other times, if you're about to swing, it's essentially nothing.

If you are not attacking someone when you get critted, the free attack happens when you attack for the first time. And for some reason, this will stack. Two crits gives you two free attacks. Three crits, three attacks, and so on. It used to be that you could store an arbitrary number of attacks. Then a paladin soloed Lord Kazzak. Within 48 hours, Reckoning was hotfixed to only store 4 charges.

Note that there is no actual UI support to determine how many Reckoning charges you have. It's completely hidden. However, delaying the attack is important, because the talent shouldn't punish the paladin for being in the middle of a heal when she gets critted.

So the actual way that Reckoning is used is that the paladin will stand around healing herself and hope people attacking her crit. After she stores 4 charges, she targets someone, usually a cloth-wearer, and unleashes 4 simultaneous attacks. This burst damage is often enough to kill someone (or at least heavily damage them).

I think it's pretty clear that people are not using Reckoning as the developers envisioned it. The lack of UI support, the talent not mentioning the charge cap, and the fact that the original version allowed for infinite attacks, all point to this conclusion. I think that it was unintended behaviour--a bug, in fact--that the developers decided was not worth the cost of fixing. And yet this bug is one of the best strategies for actually killing someone in PvP!

To me, this just seems wrong, and is somewhat emblematic of the entire development attitude towards the paladin. Reckoning, as it is meant to be used, is flawed. It does not give you the extra attack it promises! It's only because the talent is bugged, and the bug exploited, that Reckoning is in anyway useful.

I think the developers need to fix Reckoning, even if it upsets all the Reckoning-bomb paladins. Reckoning should give ONE actual instant extra attack. If this fix makes paladins even weaker in PvP, then new abilities should be added, or old ones enhanced. The paladin class shouldn't have to rely on a bugged implementation of an ability to deal damage in PvP.