IRPDS: Postmortem and reply to comments

Hey everyone,

Thanks for all the feedback and congratulations to both "supers". :) Thanks to Richard for hosting this Pyweek once again.

I'd like to reply to a few of the comments (parts of it, at least) and then write a little postmortem.

Before, I'd like to say that whoever are writing verbose and thoughtful comments are awesome. I really try to make an effort to do that, but I find it very hard, not only because of lack of time, but because I often don't know what to say. This Pyweek I only managed to rate a little more than half of the games due to lack of time, often giving comments without much substance because I couldn't think of anything, and it really impresses me how they write well. If you are one of those, a special kudos to you. I should really learn how to do it.

Second, I'm wondering if anyone played the game without text due to some bug. Text was quite important in the game. If you have, please let me know.

For the sake of not making this post too long, I'll reply only to parts of comments that mention some specifics about my game. Thank you very much for all of you who complimented on my game, even though they're not mentioned below.

Reply to comments

"It works, but for some reason, it crashed suddenly while playing. Also, when starting over, the music sometimes doesn't stop itself before playing again, so you hear the same song twice at the same time."
"And it segfaulted a lot until I changed the fonts."
"Just needed some extending and fixing of crashes."

Odd. Does anyone have any clue to why these happen so I can avoid them next time? These crashes don't happen to me, I could use some help. (The music bug is indeed a programming bug - I forgot to stop it at restart.)

"Although quite short"
"a bit too long"
"the flight itself is extremely long given that there are only two different things that seem to happen."

Not so sure if it's short or long, but I think I understand what you mean. I think all three are right. The game does take little time (only 160 seconds, to be precise), but it feels long because the events that happen are the same over and over.

"Just needs more random events that can occur."
"it's too bad you didn't have enough time to add more events"
"Unfortunately it seems like all of the planned events hadn't been completed so the game was lacking substance."


"Could you have an event where good things, like yummy cakes and pies appear, that you WANT to hit? Maybe randomly change the direction of travel from horizontal to vertical?"

Yes. These are great ideas, by the way, thanks. I'll write the events I had planned further below, on my postmortem.

"This is madness."

Madness? THIS IS PYWEEK! (Too obvious? Sorry. :) )

"the balance was a bit off"
"I'm finding it impossibly hard, though. The difficulty needs to start off much lower and ramp up."
"I didn't get very far, but that's mostly because I suck at dodging things."
"I didn't find the flying pigs to be that hard"
"Actually as it stands, the pigs and the traffic jams are at about the right level (does this depend on the probability you're at though?) - you *can* get through the pigs."

I seem to hear comments like these first three pretty much every Pyweek. :) I should make a goal to have playtesters next one. However, I'm happy to hear that some people didn't find it so hard. Like I wrote in a previous post, I think it's because it strongly depends on how you play the game. By the way, to answer the last one, the difficulty of the events don't depend on the probability; the probability only affects whether they happen or not.

"Also a problem is that there's no time to take your eyes off the action to see when the event has changed and what it's changed to. Might be better if there were a sound indicating an event change, and the text appeared overlaid on the playing area."

Couldn't agree more. I noticed that problem during playtesting and I tried to fix it by coloring the numbers that trigger the event as green. It helped a little and I knew it needed something more, but by then I had no time left.

"And show some indication of the player's health (if there is one already, I couldn't find it)."
"I would also have liked more feedback when I lost a life."

There is one, lower left, above the number. I guess it was too small, sorry. There definitely could have been an effect for when you're hurt, but I had no time for polish.

"Did not follow the theme of the challenge"

Even though the words "Nine times" were on screen at all times? Unless the text wasn't showing up in your case, I really don't know how to make it more obvious.

"As you've said, it needs more events to be implemented; once you start picking up the atoms, the gameplay becomes less interesting."

That, I believe, is a problem with the core idea. The original plan was actually to have "opposite" events occur when you probabilitize them out, not completely remove them. For example, if pigs didn't fly, you'd still see them wandering around at the top of the screen, possibly a few of them would still fall. That way, the game would remain interesting even if you're extremely good at picking up atoms.

"If you're going to develop this game more, here's an idea: stick some floating obstacles in the sky (unlikely things, of course.) This will force the player to pick up atoms that they don't want to."

That's an interesting suggestion and I did think of this initially, but I worry that this might make the game even harder. Maybe there's a better solution to make atom picking more difficult. Thanks for the idea, though.

"I know you said that collecting the probability particles affects gameplay, but it wasn't clear to me how that was happening."
"The mechanics of what was happening in the game were difficult for me to understand at first. I thought I was trying to affect the probabilities of specific major events, but on closer examination it seemed like it was more the frequency of whatever random (I think?) event had been picked?"

To answer the question, no, you are trying to affect the probabilities of the specific major (random) events that are written at the bottom of the screen, not their frequency. I wonder why you felt it was the frequency - that's probably something that needs fixing. Do either of you have any suggestion to make it clearer? These comments, combined with the "font is buggy" comments, make me wonder if you played without text, as I think it's not so obscure with the text, but I may be wrong.

"honestly saying, the visual is a huge deception - from Tee i were expecting the excellence he
reached at Pyweek9, with Quetzal"

I'm not someone who likes to rant much, but I simply can't help it but make this one. Apologies in advance.

To this judge: based on your comment, previous knowledge and a little guesswork, I'm fairly certain who you are, but since I'm only 90% sure and I don't want to disrespect the anonymity of the judges, I'll speak in vague terms.

I understand your disappointment, but was your 1-1-1 rating really fair? According to your comment, you have a strong taste on a retro style, but not all games have to be that way. Although I don't care much about the ratings themselves, it does upset me a little when they're clearly not fair, but what particularly compelled me to rant is that I noticed, by looking at comments to other games and making reasonable guesses, that you also did this to other games, either voting very low or very high based apparently mostly on the style of the game ("i couldn't run this game [...], but the pixel art quality [...] is so far beyond average, that i really bet this game seems Exceptional in everything"? Seriously, how can you rate 5-5-5 to any game solely based on the screenshot?).

Please don't let your personal taste influence the ratings so much, particularly given that you have such strong tastes. Please consider fun, innovation and production separately, and please give the game a few minutes before rating even if you don't like the visuals. I'd really appreciate it if you made a better effort to be more fair for next Pyweek. Assuming I'm not wrong about who you are, I know you like this community and I enjoy your enthusiasm around here, but I think you should be more fair to everyone, even to those that clash with your ideas and style. I know you're disappointed with this whole Pyweek this time because most games didn't suit your style, but if you really like this community, please try to be a little less zealous about your style while rating and be constructive to everyone. Thanks. :)


I usually have more time for Pyweeks, but this Pyweek I had only Sunday and Saturday, one of which I wasted completely, so, not counting "thinking about the idea" which I can do anywhere, I had only one day.

This time, I decided to do something I usually don't do: make up concrete concepts for the five themes before the challenge started. I was lucky to have gotten a theme for which I had a nice idea, playing with the expression "Nine times out of X". That's something that went right, though I suppose it was a little based on luck.

The original concept didn't involve a flying raptor; that idea came along the week. Originally, I was going to do a Canabalt-style game, but I decided it was too much work for only one day considering all the rest I had to implement, so I opted for the much simpler "use your arrow keys to move" scheme. The original concept also involved a robotic-like newscaster telling you the probabilities of events happening.

As I had said before, I had planned more events to the game. These were:
- floor becomes lava: the floor above you becomes lava and lava balls start shooting down Mario-style;
- a train cuts across the screen: a train runs over the bottom part of the screen; I wanted to have one single event that inspired fear so the player would be desperate to "probabilitize" that event out when he sees that message;
- bugs happen: Python exceptions would cover the screen, disrupting the player's visibility;
- the screen turns upside down: self-explanatory; I thought this would be fun as the ground would actually become the ground;
- rockets boost: the player would get a boost to get to the ending faster, though this means obstacles would also be faster;
- dinosaurs shrink: self-explanatory, easier to dodge obstacles;

There aren't that many events (it particularly lacks good events - though one suggested in a comment, flying cakes, was a good one), but at one moment I stopped thinking about them because I knew I wouldn't have time to implement all of them. Maybe with time I would have thought of some crazier, more out-of-the-box events. Also, as I mentioned in my reply to comments, I also planned to always have things happening even when an event didn't happen, like pigs walking when they didn't fly.

Another thing planned but unfortunately cut out was a sense of progression - the way it is now, there's no sense of progression except the progress bar. That is, the game would become harder and new events would show up as you progressed during the game. That was something important that I wish I had time to implement, but I barely had time to implement the small amount of events in the game.

Difficulty was an issue in my game, but those who know my history would guess correctly that that doesn't surprise me. I have a tendency for hard games and that reflects on my games even when I actively try not to make them too hard. I should really make time for finding a playtester next time.

Overall, I'm reasonably satisfied to how the game turned out even though I wish I had time to implement some more events. I like the concept and the execution turned out reasonable, particularly given it was made in a hurry.

Anyway, thanks to everyone for all your feedback and all your games! See you next Pyweek!

(log in to comment)


Actually an idea I had but forgot to add it to my comment was to have a positive and negative probability thing so getting white ones gets you good stuff whereas the black ones makes the pigs fly and other obstacles. Of course you'd have to balance this so it gets harder to get positive luck I don't know, but I rekon it could work. Then winning could be getting a full positive luck bar? I didn't read your whole postmortem so hopefully you didn't mention this.
"I know you said that collecting the probability particles affects gameplay, but it wasn't clear to me how that was happening."

Do you have any suggestion to make it clearer? These comments, combined with the "font is buggy" comments, make me wonder if you played without text, as I think it's not so obscure with the text, but I may be wrong.

That was my comment. Missing font was not the issue. I played it again now until I figured it out, and I see why I never figured it out before. When I was playing it trying to win, I couldn't take my eyes off the action long enough to see the pattern. I didn't concentrate on the HUD long enough to realize, for instance, that the type of event would reset after a random number was selected, or that the white ball had anything to do with the time before the number was selected. It also took me a while to realize that these were actual events in the game, and not philosophical flavor text. "Nothing bad happens" isn't obviously game-related.

My suggestion for how to clarify things would be to have a splash screen at the beginning of the level that explains the HUD. It might not be clear the first time you play through, but after playing the level once and then reading it, I think it would work.

It's probably also not clear to most people how the number works. I know that you can choose an event with a probability of 9/16 by generating a random number 1-16 and choosing the event if the number is 1-9, but that's not an obvious fact. My thought of how to do it - which might not be any better - is graphically. Have 9 "yes" balls and 7 "no" balls in a row, and a box that bounces around between them before selecting one at random. Or a roulette wheel with 9 and 7 wedges of different colors.

Sorry that's such a long response. It really wasn't that bad, I'm just going into detail since you asked. Anyway, good job! :)
Thank you both for your feedback. :)

@Trub: Thanks for the idea, but I'm not sure I understand it. The game, as it is now, somewhat does that, but it depends whether the event is a good one or not. Or do you mean to have particles that change whether the event you get is good or bad?

@Cosmologicon: That's exactly the key I was looking for, thanks. :)

I had realized that the action detracted the player's attention too much from the HUD, but it was impossible to me to realize through testing that a player might not realize what was going on, because I always knew where to look. Based on the comments, you're not the only one who had that problem, it was a faulty design. Someone up there suggested a sound and overlaid text for events - that should help.

Explaining the HUD would have certainly helped, and it'd be very easy to do had I known about this problem. I assumed that the player would see the text and recognize it as the event, but it wasn't as obvious as I assumed.

I admit that the number solution was thought not with "best solution" in mind, but with "fastest solution to implement" in mind. It was very easy to have a changing text, and I thought the "randomly spinning" number looked nice. But the roulette wheel is a great idea, that's probably what I would go with if I were to extend this game. The roulette would also make it clear how the probability change affects things, since the roulette wedges would be merged or subdivided. When I first thought of this idea, I wanted it to have some feeling of a slot machine. I wanted to have the player worried or hopeful with expectation to what would come up and force him to react quickly, but I ended up not aiming for that due to the rush.

By the way, when it comes to constructive feedback, long responses are much better than short ones. :)
Yeah you could have a luck meter that starts in the middle and as you collect the white balls, your positive luck increases and vica versa. But it mightn't work then because if your getting positive luck it would make the game easier and easier, maybe to win you have to get full bad luck? Just a couple of ideas. Otherwise I don't think you should have the luck reducing particles, so everything goes nuts haha.

On another note when I first played I thought the white and black balls were obstacles not 'collectables' and I was surprised I wasn't dieing so often. But if your planning on improving the HUD and stuff you should make this clear.