Tobold's Blog
Friday, June 19, 2009
Progress systems

Via The Ancient Gaming Noob I found this article in Wired pointing out that in the 80's video games often forced you to restart from the beginning when you lost, and jokingly proposing to define "hardcore" games as those that work in a similar way. Your raid wipes in World of Warcraft? All restart at level 1! The exaggeration shows that it is obviously easy by creating something recognizably "hardcore" by simply increasing the punishment for failure. And obviously nobody would want to play such a game nowadays.

The other extreme of that scale is Progress Quest, a MMO which allows you constant progress without even requiring any input from you. Or Jade Dynasty, where you can pay the company to not play the game. But even World of Warcraft is extremely close to the Progress Quest side of the scale: Progress from level 1 to 80 (or whatever the level cap will be in the future) is virtually guaranteed, regardless of skill. And in the end game skill is only measured for large groups, enabling some people to be carried.

So Jormundgard poses the obvious question of whether having a higher punishment for failure really makes a game hardcore. Well, it can, as long as you define hardcoreness by being able to reach exclusive content. If every failure sets you back, and you make that setback big enough, then only the most dedicated will ever reach the end. The reason why in World of Warcraft raid content is considered hardcore is because there is a punishment for failure: The raid dungeon resets, so if you didn't manage to get to the end before the reset, you have to start over. If there was no reset, any guild would be able to reach the last boss eventually, just like anyone can reach the level cap eventually.

The disadvantage is that it turns out that players are very risk averse. For example in Everquest going to a level-adequate dungeon was potentially rewarding, but also very risky. You could die somewhere at the end, with monsters having respawned at the start preventing you from getting back to your corpse, so potentially you could lose all your gear. In consequence EQ dungeons were mostly populated by higher level characters farming certain specific mobs for their rare drops, and players of the appropriate level rarely ventured there, in spite of zone bonuses later being added. Not working as intended. If dying really hurts, players don't risk anything, and end up farming boring green mobs instead of something that actually poses a challenge. Even in modern games like World of Warcraft, where the penalty for dying is minimal, players rarely try to push the envelope and go for orange or red quests and monsters. There isn't anything in the game that would prevent them from challenging themselves, but the majority goes for the safest option with the best rewards for the lowest risk of failure.

A completely different progress system is used in chess or WoW arenas, although there are numerous exploits in the latter case. The basic idea of an ELO system is that you have a score which is equivalent to your skill level, and only by improving your skill can you improve your score. In such a system playing it safe isn't an option, because winning against a much weaker opponent gains you very little, and the occasional loss sets you back by a lot. Even playing more than somebody else doesn't improve your score, unless by playing more you actually get better at it. Of course somebody playing a lot of World of Warcraft will equally claim that by doing so he now plays better than somebody who plays less; but of course that isn't necessarily the case, and the Progress Quest system isn't designed to really show the difference between skill and time spent, or a bunch of other factors. Wielding some epic from some raid boss means you were there when he died, but doesn't say anything about whether you were the MVP of that raid, or just trundled along and leeched. Being at the level cap doesn't show how many hours it took you to get there, how often you died along the way, or how well you played.

So to make an MMO in PvE work more like an ELO system and show skill, not only would you need to lose xp on dying, you also would need to lose a lot more xp from dying against easy opponents than against a hard challenge. And easier mobs would have to give significantly less xp than harder challenges. In theory a system could be designed in which everybody's level reflected their actual skill at playing that particular MMO. Why is nobody designing such a system? Probably because players don't really want an accurate measurement of their skills.

People on average, especially men, famously estimate their IQ to be much higher than it really is. That is if you asked everybody for an estimate of their IQ, the average of the estimates would come out significantly higher than 100, although by definition the average must be 100. The same thing applies to video game skills: People, especially men, estimate their video game skills to be much higher than they really are. A progress system which assures them of their superior skills, by making everyone a winner, is more popular than a system which shows who the winners and losers really are. MMORPGs are designed to be positive sum, and extremely hard to lose, so that everybody can feel good about themselves. But in reality the system simply fails to measure skill, and those who boast the loudest about how leet they are, are often simply overestimating themselves the most.
I've had experience with several games with ELO based systems, and I'm afraid farming low ranked players IS the best way to gain points, regardless of the "risk" of losing a lot of points. It only works in chess, and somewhat in WoW arena, because you are forced to play players your level and are limited in how many games you can effectively play. Imagine if a chess grand master played against players ranked 1200 in chess. Sure he only gets 1 point per win, but he NEVER loses. He plays 500 games in a row against low ranked players, his score goes up by 500 points because he wins every one of those games.

You can try and manage this, but ELO and other systems are no different in the respect that players will always "game" the system at hand. Players would complain that the only way to reach the highest levels is to farm a billion super-weak creatures, because you know they can't kill you. Fighting stuff that's a challenge seems faster, but sometimes you die and that progress is erased.
This is really just a general game design philosophy, not specific to MMO's. People want to be challenged, but not to the point where they can not progress through the game any further.
1) WoW is already using an ELO system for their arena ladders. So WoW already measures your skill.

2) ProgressQuest. Crazy how I'm having fun just looking at the little program. Gogo, my lovely Talking Pony MoFo Monk.

3) People tend to think of themseles better then they are in general. Not just men and not just for their IQ or video game playing skills:
Consider eating disorders, for instance. It’s generally been believed that an unrealistically negative body image is an important factor in the self-abuse that characterizes anorexia and bulimia. A 2006 study at the University of Maastricht in the Netherlands, however, came to a very different conclusion. Here, groups of normal and eating disordered women were asked to rate the attractiveness of their own bodies. They were then photographed from the neck down, and panels of volunteers were brought in to view the photos and rate the women’s appearances objectively. The normal women, as it turned out, evaluated themselves much more positively than the panels did, while the self-ratings of the eating disordered women were in close agreement with the objective ratings. The eating disordered subjects, in other words, had a more realistic body image than the normal women. However, it is important to note that the study was based upon the broad concept of "attractiveness" rather than body weight specifically—while the eating disordered women may have rated themselves poorly because they felt "fat," their weight was a controlled variable and not the basis of the volunteers' assessments.

It leads to the weird theory of "depressive realism". People who are depressed have a self image more in line with what other people think. While normal people per definition think of themselves as better then the rest.
I was a big fan of permadeath in Diablo 2 and have been one of the handful of players lobbying for a permadeath WoW server.

I wonder if Diablo 3 will have permadeath? It's distinctly possible considering it adds replayability and (in the case of the previous Diablo games) has been something they simply implemented without doing any extra work.
I like the idea of gear loss on death. It adds risk and excitement. Asl ong as your character retains their skills, they can earn back the gear and try again.

However, the majority of players want easy victory and low risk so such a game wouldn't be very popular -- or at least -- such activity within a game would be less popular than safer activities.
Funny how you fail to make appropriate conclusion out of all this due to your aversion to PvP. You said yourself arena in wow is the most accurate measure of skill , pitting players of similar skill against each other

If you want have stuff interesting -you have to make pvp work, dividng population in their respective skill brackets. Premade BGs were NOT zergs ,its pug bgs which are.

You dont even have to divide people physically - there are other , much more clever ways to level the playing field

p.s. and btw Grossmeister farming does not work exactly for the reasons it does not work in wow- you lose one match and you get penalized 3:1,5:1 and even 10;1- there is RNG, there are bad comps, and sometimes you jsut play bad and other side plays well.

In your analogy he might win 500 games but one lost game would cost him 250 points lost
"Hardcore" characters were quite popular in Diablo 2, and I'd be surprised if Blizzard didn't include the option in Diablo 3.

The big problem with permadeath in online games, however, is that lag can often be the cause of death, and that's difficult to manage. I distinctly recall losing multiple level 50+ characters in Diablo 2 specifically because of network lag.
In game that don't actually reward or necessitate skill, players who have invested a lot of time in their characters and HAVE skill will be pissed to see the scrubs not far behind them. Especially if you attach serious rewards and penalties to game actions. If you implemented permadeath in such a game, it would be absolutely terrible. Some crappy mage crits you and you die--nothing you could have done to prevent that.

WoW-like MMOs are not built for permadeath. If you want permadeath to work, players need to have tight control over their characters to the extent that if you died you can narrow down why and take steps to prevent it in the future. If the RNG screws you, losing 100 hours of work is ridiculous. If you used the wrong spell at the wrong time, if you're a skilled player you'll understand that and get better at the game.

WoW-style MMOs and serious penalties and rewards do not mix because they are fundamentally at odds. If everyone wins, penalties and rewards cannot be serious.
A couple of points, if you lost all your gear in WoW, getting it back would require you gaining access to some sort of gear.

WoW is NOT designed for you to lose all your gear and get it back easily.

Diablo is. Diablo has random super loot drop off of random mobs. You also, can play solo, even online, kill a bunch of mobs, level up, gain a bunch of nice gear, then jump into another server doing a boss, kill her together with some peeps and there you go. Since WoW doesn't work that way, a hardcore version like in Diablo just wouldn't work.

As for Samus with the 1200 ranked players, that Grandmaster wouldn't PLAY against 500 noobs like that, especially not in a tournament. We saw this kind of "gaming" in WoW, and they've been trying to fix it. (Have they? I haven't arena'd in a while, but I haven't been hearing the complaint...).

More than likely, and 1800 ranked player would face a 1600+ player, so losing to them, is both a bigger possibility, and would take a big chunk of their points. That's why that system has worked for aslong as it has no?
EVE has a system like this when it comes to running missions.

When you first start running missions for an agent, you gain reputation quickly with every success. As your standing grows, the reputation gained from a success decreases, and that lost from a failure becomes larger. At the high end, it may be that you need ten successes to recover from a single failure.

This helps to rank players in PVE content, but it has its downsides. One of the reasons I cancelled my EVE subscription, many years ago, is that in a single evening I had some bad luck and failed at two missions. As a result, my reputation fell so far that the high tier missions were locked out, and I had to run the lower tier missions.

As an example, suppose that in WoW you wipe in a raid, and find yourself still at level 80, but forced to run level 60 content. Penalties are useful, but if they're too harsh, your game suffers.
Isnt it strange that we can discuss "progress" as a -given-, in that progress -can- occur despite that little something called "death"?

If I complete a quest of hand delivering a letter for an NPC, or I deliver 10 dead rats(that I must kill- Oh the danger!!) to an NPC, I am still rewarded with either gold, gear, or experience that allows me to level my character...all without the fear of death whatsoever. Even if my character dies, he/she doesnt lose anything beyond a slight gear degradation, and simply corpse runs back and tries again.

If this is Carebear'ism at its finest, I dont know what is, and it's the main reason that WoW is so popular for so many of the "tourists" that found WoW as their first MMO(myself included).

Remember the Warlock questline to acquire the mount in Dire Maul? How many players actually try this at the appropriate designed level anymore to make it an actual challenge?

The problem with defining skill and how it should relate to progression, is that players will always take the easiest path of resistance to the reward.
That article is silly though, most 80s games took a fraction of the time a modern one does, and were far less complex. You could beat most of them in 4 hours or less, but many were either endless, or incredibly hard. A game like asteroids just didn't end, you kept going till fatigue or boredom overwhelmed you. On the other side, later on games in the arcade like smash TV, or any of the neo-geo games were just punishingly hard to force you to spend quarters. You ran out of money or not before beating it, but with unlimited continues, again 4 hours or less tops.

The problem with skill though is that if you don't have it, you probably won't like the game enough to stay. Or worse, people may not want to play with you if too great a disparity exists.

A 900 ranked chess player offers nothing to a 1600 one, so the 1600 oesn't play him. In an MMO, if skill is based on defeats, no one would group with the 900 player because he sucks. You wind up giving a tool to aid players to be even more self segregating.

It's bad enough as it is, where you often have to "apply" to endgame guilds, who do things like run background checks, have attendance policies, and minimum "Education" needed (like specific classes leveled and achievements had.)
1) My guess is it would take a very compelling game to make a mass-market ELO system work. If WoW were able to tell me I was #11236564 of 11500000 I might be less motivated to play; I have coworkers and family to point out my shortcomings without hiring a game company to help. Or it would at least be a significant departure for WoW and clones: getting sustained revenue stream from customers by focusing on the journey. It could work for PvP-oriented ventures; but perhaps I am cynical but I see a lot the bottom quarter of the ELO drifting away each month.

2) I know it is fashionable to complain about how all [the other] players are lazy and risk adverse. And it's basically true, Yet I want to make the case they are doing the optimal thing. If you can do a harder zone or mob and get 20% better rewards but it takes 40% longer then the latter is not as efficient. Remember that the lower level grinder can kill accomplish a lot during the time it takes for a typical "corpse run".

E.G: a game where a lazy can do 1 pig per minute and get 1 xp per pig or you can do a boar in 5 minutes and half your time will be spent in corpse runs so 6 boar per hour. If the game designer had made boars pay 25xp, people should do them. If boars paid 10xp, then risk adverse would choose pigs. But I suggest that in a lot of pre-raid WoW, the rewards scale more linearly than the efforts (twice-as-hard of tasks pay < 200% and half-as-hard pay > 50%) The fact that people find the "sweet spot" for them should be expected. If the designer don't want that behavior, they can disincentiveze it.
Post a Comment

Links to this post:

Create a Link

<< Home
Newer›  ‹Older

  Powered by Blogger   Free Page Rank Tool