<< Back to Warzone Classic Forum   Search

Posts 1 - 20 of 20   
Let's talk about SinglePlayer!: 6/23/2016 13:20:08


TBest 
Level 60
Report
Now that we have all enjoyed the new SingelPlayer (SP) for a little while, lets take a deeper look on the levels.
Don't hate me, but many examples will be my own levels....since I have extra stats that I can see, I know them, and I am too lazy to find other levels. Feel free to show us your levels with stats in this thread!


Questions to consider. Please refer to the number when responding to this thread.

1. What makes a good SP level?
2. What is a successful SP level?
3. How can SP be improved?
4. If TBest makes more SP levels, what kind of levels do you want him to make.

1. What makes a good SP level?

Easy, it has to be fun to play! Now what is a fun level? That is more individual, some like to battle the AI head on, while others would like the 'impossible' challenge. Thus levels of all difficulty has a place, naturally. Yet, I think there are a few principles that can be derived. First most, a SP must not be too long. In 'An Adventure' https://www.warlight.net/SinglePlayer/Level?ID=882705 someone took 250 turns to win! While the record is a massive 49 turns. The median is a staggering 79 turns.
Interestingly enough it is almost no limit to how short a SP can be for the record. My own 'Corrupted Tutorial' https://www.warlight.net/SinglePlayer/Level?ID=882158 that I will admit, I made in 2 minutes, Is only mastered by 1 player. It is possible to win in 3 turns, but, this is the current win distribution.


There are some slower wins, that have been cropped out, but this is the majority of the wins. This win distribution just looks nice ;) Point is, a level can be very short to win, and still fun. But it can't be to long. Through a level that can be won in 1 tun is not something I thing is a good level, anything from 2 and up may work.

Another thing that I noticed is that players seems to dislike having a AI teammate. In "Lunar War; Powerful Friend" https://www.warlight.net/SinglePlayer/Level?ID=883865 only 25 attempts have been made. (It is possible that more players opted for MoD's challenge which is right below this level, but this is still very few unique players who have tried it) In fact, the current record is more then the number of turns I took when publishing the level. So simple rule, do not give the player an AI teammate

One more rule I would like to add in. DO NOT USE RANDOM DISTRIBUTION. A player can simply refresh until he gets something he likes anyway. Thus, all levels should use custom scenario, in my opinion.

2. What is a successful SP level?

One way to measure success is the total number of attempts. Another way is the number of attempts the average unique players play the level. Another way is to look at the win/attempt ration to see that players are challenged. Now the point of this question, if you could sort SP levels. How would you want them sorted? One obvious way is by number of attempts, but it is clearly not the only way. I am curious to what kinds of sorting system you think would work.

3. How can SP be improved?

This is basically a wishlist. It is already expected that WL will develop some kind of search system to find good SP levels. But other things I would like to see, are the possibility to select a difficulty rating for your level. To change/add/update the description to a public level, without making a new version. Make the win frequency visible for everyone. And make it possible to view past record holders on anyone's level.

4. If TBest makes more SP levels, what kind of levels do you want him to make.

AKA, what is TBest (here: the best) SP levels? Some examples of way to categorize levels.
Very Quick, easy fun. https://www.warlight.net/SinglePlayer/Level?ID=882064
Very Quick, easy but hard to master fun. https://www.warlight.net/SinglePlayer/Level?ID=882158
Long, level, with a roleplaying theme. https://www.warlight.net/SinglePlayer/Level?ID=882705
Short, no neutrals AI skirmish. https://www.warlight.net/SinglePlayer/Level?ID=883865
Medium length, Hard difficulty (not for MoD through.) https://www.warlight.net/SinglePlayer/Level?ID=884883
Let's talk about SinglePlayer!: 6/23/2016 13:47:57


master of desaster 
Level 66
Report
I like your post and your "adventure levels" are really unique too!

I highly agree that a good singleplayer level should use a custom scenario. Ai as a teammate is also a rather bad idea imo. (My own level has too much randomness too, cause the number of spawned bosses can vary).

Singleplayer levels don't have to be super hard for me. I like it if there is a trick on how to do it in less turns and i figure it out. Giant frogs level is a good example for that

I think for most players, beating a level is not the goal. Many want to do it in as few turns as possible!

What would I improve? At the moment it seems to be fine, but if there are too many bad levels, they should be removed by a mod

Edited 6/23/2016 13:49:33
Let's talk about SinglePlayer!: 6/23/2016 15:51:49


Deadman 
Level 64
Report
I agree with all of your points. I find it quite addictive if the level is deterministic and the solution is not straightforward.

I think Giantfrog's level captures everything you say and should be the gold standard for a single player level. It is so elegant that I would recommend everyone try that. I've tried it a 115 times and still do not have the optimal solution!

https://www.warlight.net/SinglePlayer/Level?ID=886689
Let's talk about SinglePlayer!: 6/23/2016 16:35:42


Richard Sharpe 
Level 59
Report
While I agree in theory that 'bad levels' need to be removed (there's a reason I pulled my own sloppily built level), I am concerned about the implementation of it. Who determines what levels are 'bad'? Ideally you want an objective metric but that's really hard to implement given the nature of this game. For example, I think diplos are utterly pointless but others love them and will only play them. Just because you or I think a level is bad it doesn't mean everyone will feel the same.
Let's talk about SinglePlayer!: 6/23/2016 16:38:21


master of desaster 
Level 66
Report
^ maybe not one person alone decides on this, but i think we agree that we don't need 10 levels on MME, all with ffa or team ffa.
Let's talk about SinglePlayer!: 6/23/2016 16:41:22


Richard Sharpe 
Level 59
Report
That or just use a rating system similar to the current map implementation. Perhaps make it more tiered so only highly rated single player levels are presented to newer players or are found more easily.

Obviously you still risk the cronyism and alt-manipulated voting but it still seems like a more fair process than simply deleting some SP levels. I know I wouldn't be too pleased if a level I worked to create and submit were summarily deleted.
Let's talk about SinglePlayer!: 6/23/2016 16:46:15


Richard Sharpe 
Level 59
Report
Another thing regarding single player is that a better descriptor of the bosses/commanders is needed. Perhaps it's available already and I haven't seen it but it seems rather confusing/opaque.

When I tried to submit my level that used standard commanders, I ran into an error. Upon reporting it, Fizzer asked what kind of boss I was using, listing types 1-4. I had no idea what the difference was. I played with the scenario generation and saw that the four types of bosses were listed but didn't see any language explaining them. Add in MoD's issue with a random number of spawns occurring and I think more information is required somewhere.
Let's talk about SinglePlayer!: 6/23/2016 17:45:06

iamtaller
Level 52
Report
For the bosses, from my testing I do have some information on them, I believe. (When I say 'several' in this, I mean somewhere 5-10)

Boss 1: 400 health, as they lose health they also lose strength. Every several turns they will spawn a bunch of armies on their space after other deployments, starting with 11, then 36, then presumedly more.

Boss 2: Starts with 20 health. Every several turns they will increase in strength, going to 60, then 300. These cannot lose health; if they are not killed they will remain at their full strength, making them quite hard to beat once they are at 300. These transformations happen on turn 11 and 14.

Boss 3: Starts with 100 health, when they die they spawn four clones with 90 health. When those die they spawn clones with 80 health. (I forget how many 80 health clones spawn per each; though I think its four.) I have not tested it, however, I believe that they do not lose health similar to Boss 2.

Boss 4: Starts with 400 health, they also lose army-killing power as their health goes down .Every several turns a random territory will turn into armies belonging to the player that owns this.

Edit: Fixed the boss descriptions a bit. May still not be completely accurate.

Edited 6/23/2016 23:09:33
Let's talk about SinglePlayer!: 6/23/2016 18:00:17

[V.I.W]*mtanzer*
Level 47
Report
This is just my opinion on some of the points that have been made, other opinions may differ.

TBest you talk about ways to measure the "success" of SP levels. When you're looking at number of attempts, this will naturally be larger for shorter levels as attempts on them take less time. In fact I think long levels can be quite fun too if made well, I just won't have time to play as many of them. But ideally there will be a good variety in length, difficulty, maps, etc. An example of a long level that I really enjoyed was "Hexagon War" (https://www.warlight.net/SinglePlayer/Level?ID=884252). Winning this level is very different from just playing a standard game of Warlight. There are some tricks you have to figure out and the way to beat it is counterintuitive, but once you figure it out it's not that hard to win. I've played tons of normal warlight games before, so I tend to prefer levels on unique maps or with unique settings over something standard like small/medium earth, base income 5, 2 neutrals, 1-5 auto distribution spots.

Also I imagine some portion of players do SP levels just to complete them, but I am definitely in the group trying to get them in the fewest number of turns. This means if a level has some sort of automatic distribution I just keep refreshing until I get one I like, so for players like me I agree custom scenario is better than random. I also agree that having an AI teammate isn't so fun, quite often the optimal strategy on these levels is to conquer your teammate's territories, and it's quite frustrating for me when getting a record number of turns depends on a friendly ai killing an enemy ai rather than me killing an enemy ai.

On the topic of what to do with "bad" SP levels, I don't see a reason to ever delete a level. If a level is unpopular/disliked then one can always choose not to play it, but there might be some people who still liked it and the person who created it may have put a lot of effort into it. What we need is a way to group the levels so that people can only work on the "good" ones if that's what they want, but so that you can still find and play any SP level that has been made. I think a rating system is a good place to start, so that the less popular levels don't clog up the levels pages and it's easier to find ones that people are likely to enjoy.

Also here's my wishlist.
1. Some sort of leaderboard that allows competition across all the levels. Right now only the record-holders for each level are shown, but maybe it could incorporate what rank someone is on each level (e.g. 10th out of 100), and maybe with a rating system, higher-rated levels could be worth more, or only levels above a certain threshold could be included. Something along these lines, obviously the details could be tweaked later.

2. A way to save replays. There are some records (e.g. Corrupted Tutorial) that I had to work hard for and want to be able to look back and see how I did it later, so right now I've just been keeping the tab open all week since I got it lol. It would also be nice to share replays with other people if someone is wondering how to do a level or wants to see someone's record-setting replay.

3. It would be cool if "Record-holder" displayed everyone who has tied a record rather than just the first person to get it. Yes this would be a lot of people for some of the 1-3 turn levels but I think it would be worth it. Right now there are a bunch of levels with records under 10 turns where it comes down to who gets the optimal number of turns first, which may just mean being online when the level is made. Kind of related, it would also be cool to see past records/record-holders on each level too as TBest suggested.

4. I know right now a lot of stats are only available to level creators, but I would love to see the distribution of number of turns for wins. Getting a level in say 24 turns when the record is 19 turns could be 2nd or not even top 10 and right now there's no way to tell. It would be cool to see other stats like number of unique players who have tried a level too.
Let's talk about SinglePlayer!: 6/23/2016 18:34:38

Hennns
Level 58
Report
I think a rating system, like for maps, would be a very good addition. Also ability to filter by ratings :)


I don't think there's any way fizzer would ever delete a single-player level- just look on the worst rated maps.. However them being rated so bad is good enough imo, no need to delete them.






Also, currently the official singe-player system have a gold star system where completing a level quick enough gives you a gold star. I think taking that over to the user-created levels somehow would be pretty cool. For example give a gold star for people within X turns of the record?






@mtanzer

2. The replays are saved, everyone who have held a record on a map (not tie) have their replay saved. In addition the 'publish' game is saved. Currently only the mapmaker gets links to those replays though (which makes sense so you can't peak on how others took the record and just copy them). I don't see why you shouldn't be able to access your own record-replays though, especially as they're already saved.
Let's talk about SinglePlayer!: 6/24/2016 03:27:43


Lord of Turnips
Level 60
Report
It would also be good if as well as a filter for rating, if that was implemented, also a way to search for most played levels, or levels with the most territories or AIs.
Let's talk about SinglePlayer!: 6/24/2016 04:55:03


Wally Balls 
Level 59
Report
I would like an Oculus Rift SP level, with naked supermodels and blackjack.
Let's talk about SinglePlayer!: 6/24/2016 06:24:43


Doushibag 
Level 17
Report
Might consider a duel rating system, standard one for how much you like it and another one for difficulty, if this isn't already represented in some way. Maybe separate the standard rating into two groups, those who beat the level and those who didn't (as well as combined). Not sure how people tend to rate hard levels and how well they distinguish between good difficult and bad difficult. Would also be cool if you could filter by difficulty and undefeated, to try to conquer unbeaten levels (trying to beat the perceived easiest of the unbeaten).

Curious why creator isn't allowed to hold the record. Would be interesting to see someone try to make a map only they could beat (which wasn't merely tedious as hell). Does the creator have access to some kind of extra information in a way that makes it unfair?

Cool to see Warlight still chugging along with updates after all these years. Congrats to Fizzer for keeping it going.

Edited 6/24/2016 06:28:26
Let's talk about SinglePlayer!: 6/24/2016 06:36:33


Deadman 
Level 64
Report
Curious why creator isn't allowed to hold the record. Would be interesting to see someone try to make a map only they could beat (which wasn't merely tedious as hell). Does the creator have access to some kind of extra information in a way that makes it unfair?


Most levels have an optimal solution. Often, the record holder is whoever manages to get there first. Since the creator has access to the level before it is even public, most levels would list their creators as the record holder if this was permitted.
Let's talk about SinglePlayer!: 6/24/2016 19:49:11


Doushibag 
Level 17
Report
Ah, that is a good point Master of the Dead. Perhaps a nuanced approach is possible where the creator's best score is artificially timed after everyone else's so they can only hold the record if no others have the same score? Might also be interesting to see how many people tie for the record. As a stat whore of sorts, those win record graphs creators get I'd want to be able to see for levels I didn't create.
Let's talk about SinglePlayer!: 6/24/2016 19:51:04


Benjamin628 
Level 60
Report
Wow, joined during alpha.
Let's talk about SinglePlayer!: 6/28/2016 10:01:58


Tarhani 
Level 56
Report
There are different types of players.

I play a bit after work to relax and to stop thinking of my work. I like fantasy. I like to win a moderate challenge. So a good aingle player level needs

a) a story. If the author did not write even a few sentences, the level is most probably not worth playing it. A unique story is a good start for a good designed level.

b) a nice and interesting map. Not one of those standard bonus=countries-1 map. Not one of the grey with borders realworld maps.

c) a moderate difficulty. The computer players should have a chance to win, if I make serious mistakes. But if I play better than a computer player, I should win.

d) a special. Some settings that you don't try in everyday custom games, and that special rules should fit to this map and this story.


edit: Something like this:
https://www.warlight.net/SinglePlayer/Level?ID=889473

Edited 6/29/2016 13:11:50
Let's talk about SinglePlayer!: 6/29/2016 03:26:33

PopsingCandy
Level 47
Report
Can anyone give me some tips on how to do the single player level "Getting Strategic" because I am having great difficulty beating it. Please Help! :(
Let's talk about SinglePlayer!: 6/29/2016 19:51:59


Doushibag 
Level 17
Report
Let's talk about SinglePlayer!: 7/2/2016 20:55:42


Doushibag 
Level 17
Report
Levels weren't added until mid 2013: https://www.warlight.net/blog/index.php/2013/07/coming-soon-points-and-levels/
How much Warlight do you think you'll be playing in 2020?

Also this stuff has nothing to do with the topic.
Posts 1 - 20 of 20