Sirlin's Take on the Evolution of Skill Progression
Freemantle
Join Date: 2002-06-16 Member: 783Members
Join Date: 2002-06-16 Member: 783Members
David Sirlin is a game designer best known for his work in balancing multiplayer games and developing a few of his own. His contributions range from fighting games to table-top RPGs; designing and balancing multiplayer games are kind of his gig. He recently posted an article on his blog about the progression system Diablo III, and how it is built to address the shortfalls of both Diablo II and the three different iterations of talent trees in World of Warcraft. While it is a bit of a stretch to compare an RPG progression system to an RTS tech tree, <a href="http://www.sirlin.net/blog/2012/5/3/diablo-3s-ability-system.html" target="_blank">the article</a> makes for some good reading. I'm sure many people will read many things from this, but I see a lot of talk of "less is more" in here, and believe UWE has done an <b>exceptional </b>job at the overall design of NS2 in terms of making multiple strategies viable, and avoiding the traps that exist in many other RTS and RPG titles.
Some of my favorite bits:
<!--QuoteBegin-"David Sirlin"+--><div class='quotetop'>QUOTE ("David Sirlin")</div><div class='quotemain'><!--QuoteEBegin--><!--coloro:#00BFFF--><span style="color:#00BFFF"><!--/coloro-->... [In Diablo 2] the whole system of allocating points in the first place didn't really customize anything. It was just a giant test of if you did your web research enough to know the only reasonable way to spend those stat points...
...it's pretty hard to make talent trees that give any real choice. They sure seem to allow choice, and in theory they really could, it's just very hard to balance it all so that there's a lot of good builds. Blizzard learned this in World of Warcraft, and I think we can see a clear progression in their thinking here. The first thought was that talent trees are great (and I was on board with this). Then there were too many talents that were "required" because they were so damn good, you couldn't pass them up. Stuff like +5% damage, you just have to take that compared to various utility skills...<!--colorc--></span><!--/colorc--><!--QuoteEnd--></div><!--QuoteEEnd-->
David expands on the merits of limiting customization in one of his responses to a comment.
<!--QuoteBegin-"David Sirlin"+--><div class='quotetop'>QUOTE ("David Sirlin")</div><div class='quotemain'><!--QuoteEBegin--><!--coloro:#00BFFF--><span style="color:#00BFFF"><!--/coloro-->The deeper the tree goes, the fewer choices you have (just as a straight math statement, not a subjective opinion). While the WoW trees have a lot of leaves in them, given how you progress in the tree, there just aren't a lot of actual choices.
Consider this spectrum: 1) no customization at all, 2) WoW tech trees, 3) ???, 4) diablo3, 5) Magic: The Gathering.
If you want to go with 1), that's fine and makes sense to me. Yomi does that, and by having fixed decks it means I can build weakness into the decks on purpose that can't be customized out. Also, by having 10 (fixed) decks, it's possible to balance all those known quantities so that every one is viable. Meanwhile, if you took a game that had waaaaaaay more abilities than all of Yomi put together, and allowed super maximum customization (MTG), you could very well end up with FEWER viable decks. It took me many years of working on these games to fully come to terms with that. The promise of customization is very exciting, but the more you have, the more chance for some degenerate thing to outshine everything else and make the entire system just an illusion of choice.
The CCG I'm working on is in that #3 slot that I labelled ??? on the spectrum. Lots of customization, but way less than MTG. I'll have to write a full article someday on why that ends up being way, way better than full customization. I resisted that for a long time, but after testing it for years, it's definitely the way to go.
ANYWAY, you care more about the Blizzard stuff probably. I think 2) on the spectrum ended up being kind of a no-man's land. Lacks the advantages of "no customization" yet doesn't have the excitement you'd expect from a customization system because there ends up being too few viable choices. Diablo3 is much further along in the spectrum. My suggestion in my main post was that they might be better off being in position #3, with the limitation of one ability per category. But they are further along, in position 4 where you can build your "deck" of abilities with any 6 you want. That can still be fine.
I think the takeaway (and it requires a ton more than just my comment here to fully understand) is that there are tradeoffs for every possible point in that spectrum. And they are not necessarily obvious tradeoffs. It took me years and years of dealing with these systems to understand the value no-customization, the dangers of too much, and the sweet spot of something in between.<!--colorc--></span><!--/colorc--><!--QuoteEnd--></div><!--QuoteEEnd-->
Hopefully you guys will find this as interesting as I did. No comparisons or criticisms for UWE, but respect for the kalidescope of perspectives you have to consider even at a very high level when designing a multiplayer game to be varied, enjoyable, and balanced.
Some of my favorite bits:
<!--QuoteBegin-"David Sirlin"+--><div class='quotetop'>QUOTE ("David Sirlin")</div><div class='quotemain'><!--QuoteEBegin--><!--coloro:#00BFFF--><span style="color:#00BFFF"><!--/coloro-->... [In Diablo 2] the whole system of allocating points in the first place didn't really customize anything. It was just a giant test of if you did your web research enough to know the only reasonable way to spend those stat points...
...it's pretty hard to make talent trees that give any real choice. They sure seem to allow choice, and in theory they really could, it's just very hard to balance it all so that there's a lot of good builds. Blizzard learned this in World of Warcraft, and I think we can see a clear progression in their thinking here. The first thought was that talent trees are great (and I was on board with this). Then there were too many talents that were "required" because they were so damn good, you couldn't pass them up. Stuff like +5% damage, you just have to take that compared to various utility skills...<!--colorc--></span><!--/colorc--><!--QuoteEnd--></div><!--QuoteEEnd-->
David expands on the merits of limiting customization in one of his responses to a comment.
<!--QuoteBegin-"David Sirlin"+--><div class='quotetop'>QUOTE ("David Sirlin")</div><div class='quotemain'><!--QuoteEBegin--><!--coloro:#00BFFF--><span style="color:#00BFFF"><!--/coloro-->The deeper the tree goes, the fewer choices you have (just as a straight math statement, not a subjective opinion). While the WoW trees have a lot of leaves in them, given how you progress in the tree, there just aren't a lot of actual choices.
Consider this spectrum: 1) no customization at all, 2) WoW tech trees, 3) ???, 4) diablo3, 5) Magic: The Gathering.
If you want to go with 1), that's fine and makes sense to me. Yomi does that, and by having fixed decks it means I can build weakness into the decks on purpose that can't be customized out. Also, by having 10 (fixed) decks, it's possible to balance all those known quantities so that every one is viable. Meanwhile, if you took a game that had waaaaaaay more abilities than all of Yomi put together, and allowed super maximum customization (MTG), you could very well end up with FEWER viable decks. It took me many years of working on these games to fully come to terms with that. The promise of customization is very exciting, but the more you have, the more chance for some degenerate thing to outshine everything else and make the entire system just an illusion of choice.
The CCG I'm working on is in that #3 slot that I labelled ??? on the spectrum. Lots of customization, but way less than MTG. I'll have to write a full article someday on why that ends up being way, way better than full customization. I resisted that for a long time, but after testing it for years, it's definitely the way to go.
ANYWAY, you care more about the Blizzard stuff probably. I think 2) on the spectrum ended up being kind of a no-man's land. Lacks the advantages of "no customization" yet doesn't have the excitement you'd expect from a customization system because there ends up being too few viable choices. Diablo3 is much further along in the spectrum. My suggestion in my main post was that they might be better off being in position #3, with the limitation of one ability per category. But they are further along, in position 4 where you can build your "deck" of abilities with any 6 you want. That can still be fine.
I think the takeaway (and it requires a ton more than just my comment here to fully understand) is that there are tradeoffs for every possible point in that spectrum. And they are not necessarily obvious tradeoffs. It took me years and years of dealing with these systems to understand the value no-customization, the dangers of too much, and the sweet spot of something in between.<!--colorc--></span><!--/colorc--><!--QuoteEnd--></div><!--QuoteEEnd-->
Hopefully you guys will find this as interesting as I did. No comparisons or criticisms for UWE, but respect for the kalidescope of perspectives you have to consider even at a very high level when designing a multiplayer game to be varied, enjoyable, and balanced.