Metacritic calculates numerical averages from categorical data. My inner statistician does not approve.

'Pocket mango' was released in Sweden two days ago. I do not get why Niantic keeps pushing this game to new territories when they obviously do not have the server capacity for it.

Me and the spouse walked around for almost an hour yesterday without finding a single creature (not counting an IRL cat). It seems like the local client can keep running despite loosing the server connection. Pidgeottos and Krabbies did however start showing up after a restart.

(Also. The local daycare centre has a horse statue that lays eggs. I find this amusing)

... lots of cool places

The real world is pretty awesome at times.

I moved back in everything except graphics. I am not suprised though. This became mostly a "learn toe tools" jam. Next time I'll use the same libs and tools, to that I can perhaps create something more interesting.


(Last time I was in the Innovation Top 100. This time is just..  ouch... but I do deserve it this time)

That's a really neat idea. ... If the AIs are to improve their decks in the same fashion as the player then I'll have to figure out a way to make them understand how to build a good deck. A step beyond having them just play well when given a preconstructed deck.

I think you work around this by givnging the AI predefined wish list. I.e. a list of predefined cards that works well with the AI:s predefined starting cards. The AI would then challenge opponents based  this list and opponents known cards.

[Drafting Roguelite]

Drafting Roguelite
The game is played in campaigns, a la XCOM or Invisible Inc. You begin the game by picking a character with a special ability and drafting a deck of 30 cards. Your goal is to take down a certain super powerful opponent. You can assault that opponent from the beginning of the game but that is not advisable as their deck is ridiculously powerful and yours is not. Instead you are given the option of playing against one of several other opponents. The opponent selection screen gives you some information about each opponent: their special ability and half of their deck list. Beating an opponent allows you to steal up to 5 cards from their deck to use in future matches. Losing gives you a Strike. Three strikes and you're out.

You have only so and so rounds to get to beating the final boss. Non-boss opponents get more powerful each round. Opponents are procedurally generated. The game autosaves after every action. I'm not sure how long each campaign should be but currently I'm leaning toward 2 hours.

What if the opponents fight each other while you fight one of them, applying the same rules (win = take 5 cards, lose 3 times = out)? Some opponents start of with cards that are weak on their own, but become ridiculously OP if combined with other opponents decks. This adds an additional layer of strategy to how you pick your next opponent.

The same think can probably also be said about WWI and WWII settings outside europe or off the battle field.

My entry this time was: Shifty withch


I've got mixed feelings about my contribution this time. The game play could have been so much better if I had just managed to get one 1-2 hour or so to play with enemy patterns. On the other hand I'm d really happy with how I got the game code, libs and tool chain to work together.

You're allowed to reuse code in future Ludum Dare compos right?

I've wanted to take part in one of these for a while but don't usually find the time.  Today though I've got the house to myself and no responsibilities for about the next 10 hours, minus an hour or two to watch the F1, so I'm thinking I'll give something a crack 

The say fine folk are supposed to arrive late to parties. 

Summing up the Pomodoros gives me roughly 10 hours of focused work so far, and I just got to the point where my game can be won and lost. I'm a relatively slow code, people who can really crunch impress me.

Best of luck!

Good morning,

So far I have the meager basics of a side-scrolling shooter. The player character is a witch that can shapeshift into a bird (altering movement speed, weapon strength, hitbox etc). 



I'm continuously deploying (rsynk-ing) the game here as I progress:
http://www.superattack.se/games/ld-35/latest/game.html

I have input handled and some collision detection. At the time of writing the next thing on the todo-list is to have something actually happen (and not just printed to log) when hit boxes overlap.

This is my second LD (of a total of 2) where my possibly least desired theme from the finals wins. But I figured something out that sort of fits.

But I'm gradually realising how unprepared I am this time that the theme does not really matter.

Well, okay apparently spouse strictly means someone one is married too, but since that doesn't actually make any difference, whatever. Partner.

"Marriage" is a legal formality. If you've been through five LDs together, then you are definitely allowed to use "Spouse".

I wanna try, I wanna try! I'm not a real programmer, I just know how to use Twine and Rpgmaker, but they seem too limited for Ludum Dare projects.

Real programmers use Butterflies. Twine makes an excellent jamming tool. RPGMaker will also work fine (assuming you have one of the newer versions that can export to html).

Gobble, Pixi, Howler, Ractive: Check
Weekend cleared and Spouse distracted: Check
Cheapest possible noodels for lunch and dinner: Check

This is going to be so much fun :D

(Seriously rooting for the theme "You are the Dungeon")

Edit: Added them URLs

[First]

So to test the hypothesis that there is always roughly twice as much to do than originally imagined, here is the test runner output from the first and (probably) last commit on my current (just completed, but still unmerged) feature branch.

First

Code:

 579 -_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_,------,
 0   -_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_|   /\_/\
 17  -_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-^|__( o .o)
     -_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-  ""  ""

  579 passing (863ms)
  17 pending

Final

Code:

 611 _-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-__,------,
 0   _-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-__|  /\_/\ 
 0   _-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_~|_( ^ .^) 
     _-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_ ""  "" 

  611 passing (1s)

So where I initially imagined there would be 17 new tests there where actually 32 new tests. That is a scope creep of almost 90%. I have not falsified my hypothesis yet.

I'm starting to think there is a mathematical pattern to all creative work. Whatever one attempts, completing some work will always find 50% more work that need to be done. And completing this additional work will uncover more work roughly equal to 50% first found batch of extra work. To the total work amount goes from 100% to 150% to 175% to 187.5 % and so forth. It never reaches 200% but it gets infinitely close.

Since this project is as much about process as about outcome I will attempt to prove this with my current feature branch.

(I'm not completely certain that the constant is always 0.5, it' may vary with project and environment. But the general pattern is most definitely there there)

The whole process of where and when you click just seems to need some optimization. 

Would you mind elaborating a bit of where the user interface is causing you friction? The UX certainly needs more work, but to fix it I must first understand what the problem is.

[<title of my game>]

but since you mentioned gender twice all i can say is pls don't cater to boys exclusively like everyone else who still doesn't get it enough already

One of the tricky parts of a persona is that they are supposed to be like real people. With a gender and a profile picture it's sort of hard not to ascribe some sort of gender identity. But I don't plan on making a pre-emptive choice about this. The basis of a persona comes from interviews with potential players, not from my own biased opinions.

And just to be clear: If research shows that my potential player base is best represented by a Bissu then I am completely fine with that.

yeah i don't get it either. can you give us a concrete example of how you would use this in a game?

do you mean want to tailor the protagonist to your target audience or sth?

This is Sue. She is a 23 year old university student aiming for a masters degree in contemporary philosophy. Sue still lives with her parents (a nurse and a truck driver), but still has an extra job (10-15 hours/week) avoid beeing to much of a financial burden on her lower middle class family.

Even thou she rarely has time to finish that type of game any more she still byes every Final Fantasy game at launch day (but she does not preorder). She also plays a lot of TCG:s and has competed in national tournaments for Magic the gathering several times.

Sue got in to <title of my game> when a friend of hears shared a screen shot of one of the more comedic moments from the game on tumblr.

It sounds from how you're talking that you might be thinking of creating a single persona, but my understanding is that you need several for them to be useful.  What's helpful is getting the composite view.  Something that's no big deal to your first persona might be a point of failure to your second.

Thanks. This is a good lead.

Really good to get feedback from someone who has done something like this. How do you find people to interview?

What I've been thinking about is selecting 3 games on steam that resemble <title of my game> in different ways. I would then interview a few people in the their respective communities. Personas would then be created from that input. Possibly one persona per game, named after a (minor) character in the game.

This would unintentionally optimize for very talkative and community oriented players, but that might be a good thing from a marketing perspective.

As far as I understand it a Persona is a make belief person representing some subset of your applications user base. They are created by interviewing a handful of representatives and then kneading the input into something that looks a bit like a CV. They have a name, age, gender and most often a character illustration or a stock photo. They also have a distinct personality and a list of skills, needs, etc.

I've even heard of at least one company that bought a bunch of mannequin, dressed them up as their Personas and placed them in the office. That way the staff creating the products would constantly be reminded who they where building for.

I'm thinking of creating a persona for the player of my current game. What I want to optimize for is to better understand and thereby cater to the wants (and needs) of my potential audience. I'd like to get deeper than the standard gender/age/social-class/job, but without turning it into a self insert (because in the end I'm making the game I want to make).

It would be helpful and interesting to hear if any one else here has attempted this and what your experiences where.

I've seen some pretty good and some really bad personas in software development outside games. Getting this right appears to be harder than one might first imagine.

[Seaman]

Seaman always deserves a mention when talking about conversations in games.

Since it's not a very famous game, here's the gist of it: You have a fish in an aquarium and you talk to it for a few minutes every day using a microphone (that I assume is included with the game). It starts mostly like a tamagochi with fartjokes, but then gradually evolves in to a light daily therapy session, with mind screws. Seaman is not very mechanically deep, it does not have to be. Seaman just has a very large predefined list of responses to different voice commands, combined with a calendar and memories of previous conversations.




(The example above is an example of seaman game play in general, not necessarily an exchange of meaningful conversation)