The cleric class has long languished - not because it was useless, but because until recently there wasn't a lot of reason to have a really high cleric level. Prior to the 2010 christmas expansion, there weren't a lot of cleric spells or skills above level 30; and many cleric spells didn't scale in power as you gained levels, like mage and necromancer spells do.
As a result, cleric has been a useful, powerful class - but not something you want to be your primary. Our future update to cleric aims to change that.
The biggest updates to the class aren't likely to be new spells and skills. Rather, we want to fix existing spells and skills that aren't up to snuff with the rest of the class.
Take for example the 'peace' spell. This could be a very powerful, useful spell - if only it worked differently. We're currently imagining room effects, duration effects, and a handful of other options to improve this.
Another spell that's long been on the todo list is charm. The current version of charm has basically been dented into the ground to prevent charm armies. With this update, we're planning to switch it to a control point system similar to necromancers. This should make charm a lot more useful while at the same time limiting "charmy" problems.
Other things that could use similar kinds of improvement are turn undead, the ward spells, curses, and the monitor skill.
All this said, we do plan to add more spells and skills. So far these hit most of the major groups - character buffs and enhancement, a possible brew skill, some new curses, new damage spells, and a new undead spell.
The schedule for this is some time in august, about a month after the thief updates of the previous post. After this, we still have a ton of things to work on, including elemental channeling for mages, god code, and a couple of new character classes.
Showing posts with label design. Show all posts
Showing posts with label design. Show all posts
Friday, July 1, 2011
Sunday, January 9, 2011
Scaling
When we released the necromancer expansion back in November, I expected a pretty large spike of new players. I figured we'd get a bunch of new people from the announcement, but also a lot of high levels would want to try out the new class from scratch to see how it worked.
And new players we got; perhaps four times as many as usual, which again was about what I expected. What I didn't expect was that we would be nearly incapable of handling them in terms of areas and content. There wasn't really any fighting over equipment and quests, but from time to time it seemed newbies literally had to queue up to get their turn at a particular quest item.
This got me thinking about scaling. If we can't effectively handle four times the standard newbie load, we'll be limited to a player base less than four times the size of our current one. And it's not just newbie areas; we have to think about it everywhere, from areas to exp thwacking to the who list to the the size of Ralnoth. We might not have five hundred people logged in right now, but even if we did the game probably couldn't handle them all.
Here are some things I think could help with that:
- Instancing. Instancing allows you to create private copies of areas for players, so that they don't interfere with each other. This is great for quests, in that each player or group can be on their own private copy of the quest, and don't need to worry about someone else getting through it first and taking the quest item. Judicious use of instancing can help eliminate quest bottlenecks, and it makes the world seem a lot bigger than it really is.
- Gold and experience thwacking. Right now, gold and experience are thwacked globally, so that if a mob is run too much the returns on it drop for everyone. This is fundamentally dependent on the number of players; a better way would be to put thwacking on individual players, so that the actions of one player don't break the game for someone else.
- Item maxload. Having limited load on items is another thing that's fundamentally dependent on the number of players. Once the player load rises to a certain level, a maxloaded or rarity-based item will simply never load again. Maxload and rarity effectively need to be removed to fix this.
- The who listings, and other display systems. Having a five hundred person who list would be ridiculous. Beyond a certain point, you pretty much have to switch to a query system instead of a global listing system.
Some of this stuff is already in transit or is being worked on; in particular, instancing was booted in earlier this evening and is being tested as we speak. Maxload for the most part was phased out last week.
It's a decent start. One bottleneck at a time...
And new players we got; perhaps four times as many as usual, which again was about what I expected. What I didn't expect was that we would be nearly incapable of handling them in terms of areas and content. There wasn't really any fighting over equipment and quests, but from time to time it seemed newbies literally had to queue up to get their turn at a particular quest item.
This got me thinking about scaling. If we can't effectively handle four times the standard newbie load, we'll be limited to a player base less than four times the size of our current one. And it's not just newbie areas; we have to think about it everywhere, from areas to exp thwacking to the who list to the the size of Ralnoth. We might not have five hundred people logged in right now, but even if we did the game probably couldn't handle them all.
Here are some things I think could help with that:
- Instancing. Instancing allows you to create private copies of areas for players, so that they don't interfere with each other. This is great for quests, in that each player or group can be on their own private copy of the quest, and don't need to worry about someone else getting through it first and taking the quest item. Judicious use of instancing can help eliminate quest bottlenecks, and it makes the world seem a lot bigger than it really is.
- Gold and experience thwacking. Right now, gold and experience are thwacked globally, so that if a mob is run too much the returns on it drop for everyone. This is fundamentally dependent on the number of players; a better way would be to put thwacking on individual players, so that the actions of one player don't break the game for someone else.
- Item maxload. Having limited load on items is another thing that's fundamentally dependent on the number of players. Once the player load rises to a certain level, a maxloaded or rarity-based item will simply never load again. Maxload and rarity effectively need to be removed to fix this.
- The who listings, and other display systems. Having a five hundred person who list would be ridiculous. Beyond a certain point, you pretty much have to switch to a query system instead of a global listing system.
Some of this stuff is already in transit or is being worked on; in particular, instancing was booted in earlier this evening and is being tested as we speak. Maxload for the most part was phased out last week.
It's a decent start. One bottleneck at a time...
Thursday, December 30, 2010
Spells, skills, and class design
One of the things that takes a lot of time is tweaking the spell and skill trees to make the classes more fun to play. The various factors I have to balance are a lot more restrictive than you might expect:
1) The spells and skills within each group, for example "mage fire spells", should have proper dependencies. Greater Fireball should probably be dependent at some level on 'basic fire magics'.
2) The spells and skills within each group, for example "mage fire spells", should be separated by similar level differences, to spread the group out across a level range. If there's too much clumping, there's dead zones where you don't get a new spell for a long time, then suddenly get a bunch followed by another dead zone.
3) Spells and skills for the entire class should be evenly distributed across the entire level range, again to limit dead zones and make sure players have something to learn at every level. Necromancer is a good example of how to do this; every three levels, you get 5-6 spells and skills, from level 1 up to about level 28. Mage is a good example of how not to do this; there's 6 spells between levels 2-4, 12 spells between levels 12-14, and 4 spells between level 26-28. Mage is currently very clumpy and needs more work.
4) Low level spells and skills should be useful and interesting. Having a spell like 'ventriloquate' below level 10 makes no sense at all; newbies will think it's actually useful (it's not) and waste practices on it. It's ok to put useless stuff at higher levels, because high level players are more likely to do research before wasting scarce practices on something.
5) Really high level spells have to actually be useful, even if it's a specialized use.
The above generally result in me spending a lot of time staring at various views of the spell lists for each class. You have to keep an eye on groupings across level ranges, while making sure you haven't done anything dumb to the dependency trees for each group, and making sure you don't move useless things into the low level ranges.
And even though it may not seem that way, the above are actually really important. It seems to give you more to do, and you know that you get important, useful stuff even at low levels. I've certainly noticed a difference with my characters.
1) The spells and skills within each group, for example "mage fire spells", should have proper dependencies. Greater Fireball should probably be dependent at some level on 'basic fire magics'.
2) The spells and skills within each group, for example "mage fire spells", should be separated by similar level differences, to spread the group out across a level range. If there's too much clumping, there's dead zones where you don't get a new spell for a long time, then suddenly get a bunch followed by another dead zone.
3) Spells and skills for the entire class should be evenly distributed across the entire level range, again to limit dead zones and make sure players have something to learn at every level. Necromancer is a good example of how to do this; every three levels, you get 5-6 spells and skills, from level 1 up to about level 28. Mage is a good example of how not to do this; there's 6 spells between levels 2-4, 12 spells between levels 12-14, and 4 spells between level 26-28. Mage is currently very clumpy and needs more work.
4) Low level spells and skills should be useful and interesting. Having a spell like 'ventriloquate' below level 10 makes no sense at all; newbies will think it's actually useful (it's not) and waste practices on it. It's ok to put useless stuff at higher levels, because high level players are more likely to do research before wasting scarce practices on something.
5) Really high level spells have to actually be useful, even if it's a specialized use.
The above generally result in me spending a lot of time staring at various views of the spell lists for each class. You have to keep an eye on groupings across level ranges, while making sure you haven't done anything dumb to the dependency trees for each group, and making sure you don't move useless things into the low level ranges.
And even though it may not seem that way, the above are actually really important. It seems to give you more to do, and you know that you get important, useful stuff even at low levels. I've certainly noticed a difference with my characters.
Subscribe to:
Posts (Atom)