EPTThoughts on Mapping and TLMC #2 - Page 2
| Forum Index > SC2 Maps & Custom Games |
|
NewSunshine
United States5938 Posts
| ||
|
EatThePath
United States3943 Posts
On June 22 2013 13:23 SiskosGoatee wrote: I disagree with making posts simply to say you agree with something. Isn't that against the TL rules anyway? ^But making posts that state disagreement are okay.  Thanks for the attentive and compendious post, monk. I agree with much of it and can't find fault with any of it. I'll just comment on your closing thoughts. I would caution not to paint all mappers with the same broad brush, but more than that, I think our irritation gets mixed up sometimes, even by us. I think most or all of us are primarily happy with tlmc and what it tries to achieve. However we are constantly upset that SC2 has -- honestly -- such a shitty ethic of map rotation and experimentation. This is probably the one thing that all of us share, and personally it's the only thing I care about changing, much more than seeing my maps or the maps I like get used. In that sense tlmc is nothing but win, even though we still complain that more needs to be done. And tlmc2 seems better than tlmc1 in all ways (assuming some of these make it to ladder) so I have nothing but good expectations. [edit] By way of clarification, I'm saying that tlmc is a great platform for developing a culture of map rotation using community maps. The more regular and ingrained tlmc becomes, the more opportunity it affords experimental maps to make it in the future. Compared to the status quo without tlmc, that is great. | ||
|
Big J
Austria16289 Posts
On June 22 2013 05:21 monk wrote: [*]Ravage This map recieved a few changes after it won 2nd place in the TLMC. Here's the most updated version. For many, especially mappers, this was a surprise win. After all, it's very similar to both Ohana and Belshir Vestige. It's also very similar to Insidious. So why did it do so much better than its darker counterpart? I believe Ravage is the perfect example of how players see maps differently from mappers. Mappers will generally just see a an Ohana clone. Here's what I see: I love that killing the middle rocks opens up more aggressive options. I love that the watch tower is initially useless, but suddenly gains value when the rocks are down. I love that you have a lot of choices about which expansions to take. I love that all the expansions are well-balanced in their difficulty to secure. I love that it also has a lot of interesting paths. There are no revolutionary features on this map, but there's so many small features that I feel mappers don't appreciate enough, features that promote interesting gameplay. Out of all mappers, Ironman worked the closest with me after being selected as a finalist and often asked for advice and feedback. My feedback mostly focused on making sure all the map features had strategical relevance and that they would be used in an interesting way. [I take responsibility for those new collapsable rocks (unless you guys hate it, then it's all Ironman's fault)].! I think this is one of the bigger problems with the mapping community. Stuff gets way too often judged to be a Clone - while there is not as much you can do if you try to make a standard map. And certain themes will simply repeat themselves. That's really why I love the choice to bring professional players into the jury. When playing, there are simply huge differences between features that are only similar but not the same. | ||
|
Ragoo
Germany2773 Posts
Also mappers simply have a much bigger imagination when it comes to different types of maps that could exist. I believe that's why their category for standard maps that play out the same is relatively big. While for the viewers/players who are only used to these maps they all feel more different. It's like the only food a person knows is icecream. There is certainly lots of different tastes in icecream and he will carefully choose which one he likes and dislikes and for what reason. But when a person who knows all different types of food comes in, he will certainly still know the different tastes of icecream and appreciate them to a certain extend, but since he knows aaaall the other food he wouldn't just always eat only icecream or appreciate the difference between chocolate icecream and vanilla icecream quite as much as between icecream and steak. | ||
|
Samro225am
Germany982 Posts
Maps which encourage the metagame to develop in interesting ways and the finalists reflected this quite a bit, but most of them did not end high up in the ranking. The problem is - as already pointed out quite nicely by Ragoo - that many people shy away from the new, because it would challenge their game and them to a point where they are not comfortable.Another problem is that people judge maps by looks quite often. What i mean is not the quality of the visual design per se, but the overview. One should be able to read the overview, but one should also take the time to read an overview. I do not go so far to say many people cannot read overviews, but just too often judgement is made with too little insight in the map's design. A minor aspect - yet something to consider: some features stand out and people understand what you want easily. On the flip side that also means that a map where not one feature stands out but has some small tweaks here and there that make it play different will often be overlooked. - These points basically ask map makers for several things, which I think are restricting and will not help the game evolve: (1) restrict the map's complexity that your grandmother could read it, make one feature distinctive enough so that it stands out , for example the outsider thirds. (2) never use a non-bright tileset, no matter how well you made it so that it is readable in the in-game overview, minimap or during play. (3) The Overview needs to be readable in a size of 200 x 200 pixels. - Analysis why Strangewood did not succeed over other maps, but ended up as one of the sad finalists that did not even receive a placement behind the three winners and why it is a good map - as it is currently being proved in the RSL Season 5 (well, hopefully at least): First I need to mention that monk send me the link to this thread and I was almost killing him on Skype for portraying my map in the way he did. After some back and fourth we agreed on several points, but I guess that is not so interesting to you guys. Basically you can read Ragoo's comments and you have our discussion. Yet I guess looking at his summary here is a good starting point to see why the map did not succeed in the TLMC#2. On June 22 2013 05:21 monk wrote: [*]Strangewood I thought this was a very well-made map and the only real problem we saw was that close position rush distances were too short (Think less than Antiga close pos). Samro, the mapmaker, intended this to be a feature, that you would have vastly different types of games depending on rush distance. To help alleviate some of these close rush distance problems, rocks were added in but many of them felt unnatural and just "a fix" instead of a natural feature of the map. In later versions, these were smoothed out more. Overall, I don't think the concept of such varying rush distances worked out as many voters were turned off by both this and the shorter distances afforded by close positions. Problem detected: close position rush distances Feature: All spawn distances are different Analysis: The map's design to have one short, one mid and one long distance might not be the best decision. As monk writes he think people were put of by it. When you think about it, the issue is that people want to win games. Hence they veto maps that do not match their gameplan and choose maps were they think they know what to do in a certain situation. What Strangewood does is that you can not count on the map being a macro or an agressive map, this map provides the playground for everything. If monk's analysis is right, then i will refrain from such concepts in the future. Right now the map is used in RSL Season 5 and I hope to see more good games on it. From what I have seen before the close position is not abusable to the extent where if forces a specific playstyle, because of the push you have to expect. Verdict: Players prefer all distances fairly the same (think: Frost). It might not be overly entertaining to watch, because games are rather similar, because you can have the same game plan for each spawn setup pretty much. Problem detected: Antiga Feature: a close rush distance around 40-42 seconds Analysis: I have to say I totally disagree with monk. Not only is the Antiga comparison missportraying Strangewood, also it is just wrong in the sense that the actual Antiga problem is not the distance in close position, but the positionally imbalance we experienced there for known reasons. Strangewood unlike Antiga provides positionally balanced thirds (unlike Antiga) and it would be intersting to see games on close distance hat actually play out aggressively. Verdict: Bad comparison, I think that portrays the map really really bad. Problem detected: just "a fix" instead of a natural feature Feature: Author reacted to judges inquiry, added rocks, changes terrain and later took the rocks out again. Analysis: I feel a bit bad about this. Being allowed to react on very short notice to give your map a shot you change a map and afterwards you have to read that the back and forth hurt the image you map has build up. Interestingly the set up of rocks added were tested before during the initial design of the map and rocks actually play a big role in the map's deisgn. This is totally not mentioned here, which is a pitty. The map as it was played in TLMC and was up for vote was very streamlined and had a very clear use of rocks. Verdict: Do not change your map in fruitful discussion with the judges during the testing phase that TLMC provides, better just hope you get it through and change your map to add finally something interesting after the map won second place (...) Problem detected: many voters were turned off Feature: building a map that is different in many details Analysis: Voters did not veto maps, voters picked ONE map that liked the best! So they picked maps that were either on the secure side (ohana2.0) or had a distinctive cool feature. Also they were clearly put off by the dark tilesets as it was at least represented by the overview. Biggest feature why they were put off is the insecurity about if a map is a close spawn or a long spawn map as discussed before. Verdict: See second paragraph in this post. - Dear judges, you did a good job, but you were not able to sell your idea of challenging the meta game to the crowd. I just hope my maps shows some good games in RSL and aI can prove Monk wrong in his analysis (; I remember when Ironman said on skype that his map should not have a chance really, but everybody knew he was wrong, because it was so similar and solid just too many people would like it. The frustration for many map makers is not that they are not able to build super solid maps, but it comes from seeing that even in such a contest you do not get too far with something that actually is challenging the meta game. Monk, I respect you a lot, but talking about ohana2.0 central rocks as if they were the newest shit and oh so interesting is really really putting me off. We already had that in early WoL (with smaller maps overall for sure) and obviously in bw before. On that map they do not actually do something interesting i would say. | ||
|
Semmo
Korea (South)627 Posts
| ||
|
Samro225am
Germany982 Posts
| ||
|
eTcetRa
Australia822 Posts
| ||
|
ScorpSCII
Denmark499 Posts
On June 22 2013 23:05 eTcetRa wrote: Thanks for the write-up Monk. Some people do appreciate it, even if they don't show it. | ||
|
SiskosGoatee
Albania1482 Posts
On June 22 2013 22:59 Samro225am wrote: What's this?:statement taken back: If you're going to publicly make a big and damaging line like this at least provide some context to it I'd say. | ||
|
FlyingBeer
United States262 Posts
A truly great map is one that: Maximizes the number of ways a player can play a match, and minimizes the number of ways any one race has an advantage over another. Simply meeting this criterion is generally going to result in quite standard maps. Being able to deviate from standard while continuing to meet this criterion requires a great deal of skill, and from what I've seen, I'd say less than a dozen mapmakers understand the game well enough to accomplish it on a consistent basis. I thought about half of the finalists had glaring balance issues that should have been obvious to the judges. If Ragoo really wants more interesting maps in the pool, there's a way to get it. Put 8 really standard maps up as finalists and 1 interesting one. The 8 standard maps will eat into each others' votes while the interesting map will get all the votes from players who want crazy maps. By pushing for more out there maps, you may have actually hurt the chances of one getting into the top 3. It's a known flaw of the winner-take-all plurality voting system. You could fix this by switching to, say, the Schulze method. I just went and looked at the original version of Strangewood Mire from the original TLMC 2 thread. I didn't like that map for a long time was because I thought the third was too safe because of the rocks. I had thought that was the original version of the map because that's what was displayed when the winners were announced. But you're saying the judges are the ones who talked you into putting those there? Wow! The judges don't understand balance.... | ||
|
chuky500
France473 Posts
![[image loading]](http://i.imgur.com/Zydsqos.jpg) ![[image loading]](http://i.imgur.com/AhU1zY5.jpg) Mappers are too lazy to make good lightings in my opinion, and texture too safely by choosing textures with only 1 or 2 colors. Also I have no idea why the majority of mappers lock their maps, there's no way to improve them unless they get picked by Blizzard. The objectives of TLMC #2 were to find maps that don't play the same and to build the hype for those maps. The 2 objectives were missed, maps are standard and there were only 6 full pages of comments on the results announcement though it's been twice on the front page. Blizzard and Ritmix don't need TLMC to find new maps, they've considered and added maps that were not finalists and they problaby did a better job at picking and promoting new maps. The limited hype isn't a really big deal and it probably has to do with the lack of originality of the winners, maybe with the finalists. That's why TLMC should have been organized differently and aimed more strongly at its objectives. It should have had a very big map pool of every uncommon map submitted, even maps that are obviously imbalanced like Koprulu or the 1st Keru. Then you remove maps that don't strictly follow the rules, this means if your rules say don't submit other people's work you can't port a Kespa map, and if your rules say a base have this many gas and minerals you remove the ones who don't stick to that (so no Electric Circuit and no Koprulu). Then you remove a few maps that are really really way to imbalanced to be in the TL Open and then you have your TL open on all the maps remaining (many of them). Then you start judging. If the goal was to find original maps, standard maps shouldn't have made it to the finalists. Plus what we call standard is actually what has been in map like Shakuras Plateau, then maps like Daybreak or Cloud Kingdom. So maps judged to be standard in the TLMC corresponded to a 2 year old WoL rather than Hots. For example in Hots mutas are a lot stronger but I don't see anything regarding that gameplay change in the winner maps, nothing to help defending that increase of threat. I hope there's a TLMC #3 but if I were Blizzard I would question letting TL organize that, just like when they took away the WCS NA organisation from MLG. | ||
|
SiskosGoatee
Albania1482 Posts
Plus what we call standard is actually what has been in map like Shakuras Plateau, then maps like Daybreak or Cloud Kingdom. So maps judged to be standard in the TLMC corresponded to a 2 year old WoL rather than Hots. For example in Hots mutas are a lot stronger but I don't see anything regarding that gameplay change in the winner maps, nothing to help defending that increase of threat. Well, the reasoning behind Mutas being stronger is that Blizzard wanted them to see more play. The drawback is that infestors are weaker so you can't just establish economic dominance any more, then switch to infestor/brood and completely dominate. | ||
monk
United States8476 Posts
I just went and looked at the original version of Strangewood Mire from the original TLMC 2 thread. I didn't like that map for a long time was because I thought the third was too safe because of the rocks. I had thought that was the original version of the map because that's what was displayed when the winners were announced. But you're saying the judges are the ones who talked you into putting those there? Wow! The judges don't understand balance.... To suggest that the judges (which included two progamers) didn't understand balance is just silly and frankly a bit insulting. Also, I think there's been some confusion on this. I don't recall suggesting any changes that would make thirds easier to take and I'm pretty sure he's talking about different sets of rocks. I thought the thirds on the original map were too easy, if anything. The objectives of TLMC #2 were to find maps that don't play the same and to build the hype for those maps. The 2 objectives were missed, maps are standard and there were only 6 full pages of comments on the results announcement though it's been twice on the front page. Blizzard and Ritmix don't need TLMC to find new maps, they've considered and added maps that were not finalists and they problaby did a better job at picking and promoting new maps. The limited hype isn't a really big deal and it probably has to do with the lack of originality of the winners, maybe with the finalists. That's why TLMC should have been organized differently and aimed more strongly at its objectives. It should have had a very big map pool of every uncommon map submitted, even maps that are obviously imbalanced like Koprulu or the 1st Keru. Then you remove maps that don't strictly follow the rules, this means if your rules say don't submit other people's work you can't port a Kespa map, and if your rules say a base have this many gas and minerals you remove the ones who don't stick to that (so no Electric Circuit and no Koprulu). Then you remove a few maps that are really really way to imbalanced to be in the TL Open and then you have your TL open on all the maps remaining (many of them). Then you start judging. If the goal was to find original maps, standard maps shouldn't have made it to the finalists. The goal of TLMC#2 was not to find the most creative maps; it was just to find the BEST maps. The creative part at the end was just a footnote we received from Blizzard right before we began judging. This was why the judging was focused more towards creative maps. The contest was already planned out before we even heard of this note. And even after this minor note, the focus still remained to find the best ones, not the most creative. Plus what we call standard is actually what has been in map like Shakuras Plateau, then maps like Daybreak or Cloud Kingdom. So maps judged to be standard in the TLMC corresponded to a 2 year old WoL rather than Hots. For example in Hots mutas are a lot stronger but I don't see anything regarding that gameplay change in the winner maps, nothing to help defending that increase of threat. This statement is just ridiculous in so many ways. How are you assuming we didn't account for HotS? Just because you don't notice changes doesn't mean they're not there. For example, most of the third designs in the finalists wouldn't have been accepted in WoL when roach maxing was a huge problem. Also, the existence of booster medivacs was a big factor in deciding to change Khalim, as I mentioned. Other things considered include oracle harass, proxy oracles, reapers, and lots of stuff involving the mothership. Also, the mutalisk buff from WoL to HotS shouldn't impact mapmaking much in that maps need to be better suited to defend mutalisks. Mutalisks are used completely different in HotS TvZ from WoL. They're much more rarely used as harass and Terrans will more often use aggression to fend them off instead of defenses. As for PvZ, phoenix instead of templar are used more commonly in HotS and phoenix don't care about the map in general compared templar based defenses. I hope there's a TLMC #3 but if I were Blizzard I would question letting TL organize that, just like when they took away the WCS NA organisation from MLG. Blizzard didn't take away WCS NA from MLG. WCS NA was forced on MLG, something that they didn't want. Also from experience, I'm almost certain that no other known organization wants to invest the resources to host a similar contest. | ||
|
NewSunshine
United States5938 Posts
On June 23 2013 06:22 chuky500 wrote: A dark tileset doesn't really make the map dark as much as a dark lighting. Here's an example of the hots char marsh tileset with default lighting and a tweaked one + Show Spoiler + Mappers are too lazy to make good lightings in my opinion, and texture too safely by choosing textures with only 1 or 2 colors. Also I have no idea why the majority of mappers lock their maps, there's no way to improve them unless they get picked by Blizzard. I'm pretty sure you're just talking out your arse on this one. How do you know what we do or do not play around with? Lighting, ambience, tilesets, data, and doodads are all atmospheric elements that I commonly play around with, and I'm sure other mappers do as well. Hard to know that without actually being a mapmaker, though. And we lock our maps to control how the map gets distributed, if it is at all. Blizzard has nothing to do with it, if I wanted to improve the map, I could do this thing called posting, and gathering feedback. Skype's fun too. On June 23 2013 06:22 chuky500 wrote: Plus what we call standard is actually what has been in map like Shakuras Plateau, then maps like Daybreak or Cloud Kingdom. So maps judged to be standard in the TLMC corresponded to a 2 year old WoL rather than Hots. For example in Hots mutas are a lot stronger but I don't see anything regarding that gameplay change in the winner maps, nothing to help defending that increase of threat. If standard to you is still what you see on Shakuras Plateau, I'm soundly convinced that you're still out of touch with maps, in a disturbingly comprehensive way. I guess that's why it's surprising to me that you still act like you know everything, when I've seen nothing to indicate it. You like to talk shit about how we don't know what we're doing, and how you somehow know something none of us do, but I've seen absolutely nothing substantial to back it up. You've just recently made your second map, after spending 3 years sitting on your first. You are the equivalent of a player who got Diamond at the start of WoL, then proceeded to not play, and talks shit about everyone playing the game at a sub-GM level ever since. Where's the action, where's the experience and the effort? | ||
|
SidianTheBard
United States2474 Posts
On June 23 2013 09:06 monk wrote: To suggest that the judges (which included two progamers) didn't understand balance is just silly and frankly a bit insulting. To be fair though, when you get a map like Keru in the finals it really does make you question. Lowground Main, super hard 3rd, impossible forth? How can you even look at that first draft of it and say it's balanced? Koprulu - Air units and drops are extremely strong in HotS. Let's add mineral walls with extremely long ground distances to defend the opposite side! Khalims - Backdoor? Great idea! No way to expand away from your opponent? Even better! Longer ground distance for the defender to bounce between 3 bases then the attacker? Balanced! | ||
|
monitor
United States2402 Posts
| ||
|
monk
United States8476 Posts
On June 23 2013 11:49 SidianTheBard wrote: To be fair though, when you get a map like Keru in the finals it really does make you question. Lowground Main, super hard 3rd, impossible forth? How can you even look at that first draft of it and say it's balanced? Koprulu - Air units and drops are extremely strong in HotS. Let's add mineral walls with extremely long ground distances to defend the opposite side! Khalims - Backdoor? Great idea! No way to expand away from your opponent? Even better! Longer ground distance for the defender to bounce between 3 bases then the attacker? Balanced! No one is suggesting that the original versions are completely balanced. (To be honest, if you're looking for imbalanced maps, Electric Circuit might have been the most imbalanced out of the bunch.) Anyways, the way I personally judged, a balanced map got only a slight boost to the its rating. As long as I thought the map was balance-able, it was fine and acceptable for a finalist. The OP of this post even describes our process that TL Strategy used to help test and make the maps as balanced as possible in time for the TLOpen. The low ground main was actually very controversial and definitely not as clearly imbalanced as you make it out to be. I asked a lot of progamer and top GM Protoss players for feedback about the low-ground mains and the response I got from them was very mixed. The third is hard, but definitely not unplayable. The fourth is nowhere near impossible either. Koprulu was a concept we were willing to test. We knew it was probably going to be imbalanced in some way. Backdoors are not a balance problem at all. Icarus is in fact a map used in the GSTL and I would argue the imbalance from that map comes from the hard third than the backdoor rocks. We also identified the other two problems you mentioned and believed they could be fixed (which was done). One of the reasons I made this post was to elucidate people on how the maps are judged and I think it might be a misconception for some that the original submitted maps have to be completely balanced. | ||
|
SidianTheBard
United States2474 Posts
Heck, in the TLMC Results thread: We were immediately blown away by the quality of the submissions and deciding between the top maps in both categories was incredibly difficult. The judges this season were instructed to look for variety and new ideas in their maps, while remaining balanced. I'd guarantee any of Plexa's top 40 maps with tweaks are balance-able, yet Keru's gets the extra attention because of the low ground main, or koprulu because of the mineral walls. I look at your sample list of balance concerns you put in the OP and I look at Keru and go through that list and I swear 75% of that list, when looking at Keru, makes Keru seem like a piece of shit map, yet to me it seems what you're telling me is that doesn't matter because we can fix that up afterwards and because the first draft had "cool features" it ended up getting selected because it can "develop the meta-game!" | ||
|
monk
United States8476 Posts
On June 23 2013 12:57 SidianTheBard wrote: I just feel that way of rating maps seems kind of flawed then. If balance doesn't really matter because you can balance it later, so instead you just purely look for new unique ideas, no matter how imbalanced because you can just go back and standardize the map afterwards. Heck, in the TLMC Results thread: I'd guarantee any of Plexa's top 40 maps with tweaks are balance-able, yet Keru's gets the extra attention because of the low ground main, or koprulu because of the mineral walls. If you disregard all maps that are potentially imbalanced, there would never be any new ideas. If we picked the most obviously balanced map, they would all be super boring standard maps. We don't expect the mappers to be complete experts on balance, especially for new map designs. Thus, we test the maps ourselves within TL Strategy and suggest balance changes. We eventually suggested changes for seven out of the nine maps. Should those first seven maps not have been selected for finalists? Overall, we picked the maps that we thought would be the best and most interesting after small balance tweaks. Sure, most of the top 40 maps are balance-able with tweaks, but after those tweaks, would those as good as the maps we did eventually select? The judges certainly didn't think so. However, one thing I will say is that I will consider balance slightly more next time if I judge future contests. Simply put, testing, identifying problems, and coming up with solutions took a lot of work and was quite difficult. I was very happy with some of the end products that turned out very elegantly after these changes (Khalim). However, I felt that in some instances, the changes weren't ideal (Kropolu) or that the maps weren't even balanced in the end (Electric Circuit). I look at your sample list of balance concerns you put in the OP and I look at Keru and go through that list and I swear 75% of that list, when looking at Keru, makes Keru seem like a piece of shit map, yet to me it seems what you're telling me is that doesn't matter because we can fix that up afterwards and because the first draft had "cool features" it ended up getting selected because it can "develop the meta-game!" This is just a difference in opinion about Keru and has very little to do with what we were arguing. The only two major things that were changed were the low ground main(which was only a borderline problem) and main size(which wasn't even that big of a deal). During judging, Keru was one of the most popular maps and a recurring theme that was discussed was that it didn't even need the low ground feature to be that good. I think Keru has great features and the other judges all seemed to agree with me. We didn't even "fix" or "change" any of the cool features I talked about in the OP. | ||
| ||
