Last year, we wrote about how we almost submitted Okhlos to the IGF. We were about to do it, but we decided to back off because we thought that Okhlos wasn’t in proper shape for something like the IGF.

This year, we are much more confident. Okhlos is far from complete, but in its current state, it has lots of crazy things happening that we think are cool. Also, the visuals are on more solid ground. And that’s without mentioning that the game has changed a lot from those days, when things weren’t procedural, and we didn’t even have a roguelike foundation.

That’s why, this week, we are starting the “ROAD TO IGF” updates. They will all be very brief and straight forward, in order to make the most of our time from now on until the 22nd of October. The final update in this saga will be just two days before the deadline, and that will include a changelog from this version to whatever we have in a month.

Now, for this week:

The Displacement marker

There are certain GUI elements that are crucial in a game. In Okhlos, one of the most important elements is what we call the displacement marker, that little flag that shows you where are you directing your mob.

At first, this was a very difficult concept to explain. People didn’t seem to grasp the idea of moving the mob alongside their character, or using the stick in order to attack.

So we solved this (at least for now) by changing the graphics according to the context.

So, if you need to move the mob, your displacement marker will show you where they  will move (approximately). But if you are near an attackable thing, the displacement marker will change. We think this will help tell what’s interactive and what’s not.

Right now, we drafted some ideas about how it should look, but we aren’t quite sure about which direction should we go. They are very showy, that’s for sure, but when you have 50+ people around, maybe you need showy. We still have to try something more subtle though.

Also, we prepared an alternative, which is the arrows rotating, and the swords pointing and scaling. Exactly the opposite of the previous one.

We’ve just arrived at these options, so any comment or feedback about this is super welcome!

Finally, another thing we had to decide is to make the displacement marker a billboard or not. All the units in Okhlos are billboards, which means they always face the camera. Opposed to that, not being billboards means that they will be just stand there, and that they will not rotate according to the camera’s position.

I think, in this case, the best solution is to not use a billboard. This will give the player a better sense of depth, and a better idea of the position of the marker.

 

Anyway, that’s it for this week! We will be working hard in order to have a cool build for the IGF. Wish us luck!

 

Read More

I’m a big fan of iteration. I love letting some time pass and then revisiting something. The problem with iterating, is that takes too much time. Sometimes you have to do the same thing two or three times.

In this sense,  ideally you want to do it right the first time. Of course, you can’t plan on doing it right the first time. You can try, but it’s very unlikely that you will have it right at first.

Time is another huge factor. Sometimes, you just want to see your work in game. Seeing things implemented, even when it’s a draft, can help a lot to give you an idea of how it will look and feel in the game. Context is important.

This week, I will show you the difference between the first and second versions of a character. I think that a few of these tips are useful not just in pixel art, but in almost every creative field.

The important part: If you are not extremely happy with your result, do it again!

 

Now, the example:

The last weeks, we’ve been showing you the Shield Bearer, an Ares warrior who takes cover behind a giant shield.

This is the original enemy design. If the character is looking forward, the only way to attack him is from behind, and vice versa if he is looking back. He always moves and attacks behind the shield, so you have to sneak from behind. We can start from here.

 

Idle Animation:

 

As you can see, the base gesture is not horrible, but it doesn’t suggest action, and the stance is kind of boring because he is just standing there. The new version is much more dynamic. It has the same number of frames, but you can see that he is waiting for action! Also, you can see some shoulder deformation, almost like a hunchback. This was unintentional, but I thought that it added personality, so I left it.

 

Attacking Animation:

The new Attacking Animation preserves the same stance from the new idle animation. This version adds an extra frame, and that gives a little more momentum prior to the attack. Also, I reworked the cape, because the previous one was awful.

 

Running Animation:

This animation also derives from the idle stance.  Like in the other cases, there wasn’t much I could use from the former animation, and I had to redraw a lot.

 

Turn Animation:

The turn animation had the same problem as the idle one. It lacked action. Using the idle animation as a base, we could do something much more dynamic. Also, we added a subtle movement in the foots, that helped.

 

Back Animations:

Back animations are a pain. A real pain. They are awful to do. They don’t look good, and that was a perfect excuse to stop doing them for the mob units.

In some enemies, we still have to do it. This was the case of the Shield Bearer. The back animations are as important as the front ones, so we had to add some love to them. Also, with the new animations, it’s much easier to read where he is facing at.

 

Animating these enemies is a completely different task than animating the mob units. I’ve learned a thing or two animating these giants: It really helps drawing some guidelines on top of your character, and not to think so much about reusing every part. You might need to redraw an arm, or the torso for a particular pose, but that will give some life to your animations.

 

The little blue cross you see below the character is a small mark I do to know the absolute position the frame must have in the .psd file, since each frame will be automatically exported into .png.

The blue / red thing is to have a better reading of what’s happening. You might need that when you are trying to understand a couple of strokes in a 50px canvas.

 

So that’s pretty much it for this week. It’s a bummer to have to redo some things but, in the end, it’s usually worth it.

 

Also, this week makes the debut of our new comment system with Disqus. Feel free to leave a comment to test it!

 

 

Read More

This week we are going to take the opportunity to show you the dirty secrets behind some of the mob’s favorite enemies. We are going to make a little trip inside the minds of the Hosioi, the Shield Bearers and Phobos to tell you about a couple of little twitches we have used to improve their behavior.

It is much easier to make an AI with perfect accuracy than make it have a realistic chance to miss.  This is why I hate when robots and androids in movies miss their shots as if they were space cadets. Specially when shooting laser weapons, they don’t even have to compensate for wind speed, direction and things like that! Who makes these purposely faulty robots!? Do they do this to give the humans a fighting chance in case they rebel? Okey, that could be a good argument, because they end up revolting against living beings a lot… But I digress. The point is that is easier, more even so in the case of a simulation where you control all the environmental factors too.

Back to Okhlos. Do you remember the Hosioi? Apollo’s acolytes that shoot fireballs? Well, the Hosioi’s first prototype had perfect accuracy, because they were put together very hastily. The code simply said, the target is at that point, shoot to that point. And since the fireballs were pretty fast there was little chance to avoid them. Our next step then was to make the Hosioi fallible. To do achieve this we did two things: first we added a little randomness to the shots. Not much, but just a chance that when the Hosioi shoots the fireball it comes a bit skew. The second thing we did was making the Hosioi track its target’s position with a little delay. This way the Hosioi will have less chances to hit a moving target but will be more accurate against still targets. And so, with two little changes Apollo’s acolytes went from relentless machines that never miss a shot to being fallible creatures just like you and me, having to aim their shots and sometimes experiencing the bitter feeling of wasting a perfectly good fireball.

 

Something quite similar happened with the Shield Bearers. You remember them, don’t you? From Sparta, big shields, not bears? Well, they don’t shoot anything (they trust their nasty shield bash to punish people) so the attack wasn’t the problem here. The problem here was their movement. Shield Bearers hide behind their towering shields, making the only way to beat them to attack them from their backs, and if they see the enemies are moving to flank them they will turn and face their shields towards them. So if they can follow the enemies’ movements accurately and respond immediately to them,  it becomes nearly impossible to hit them.

We knew this from the start, so the Shield Bearers are one of the few characters in the game that have a “turning” animation. Everything happens so fast in the game, people are running from here to there all the time, that simply flipping or switching sprites is all that is needed most of the times, but in this case we wanted the Shield Bearer to take some time to turn around, hence the animation. But it wasn’t enough. It takes time to maneuver the mob so the few moments the animation gave you were not enough. This is when we turned to one of the tricks we had learned with the Hosioi, tracking the enemies position with a little delay. This way you may get enough to get them before they face you again, and the idea of these giants hiding behind their massive shields being a little slow makes perfect sense too.

 

 

Last comes the story of Phobos. Not long ago we decided that Phobos would move around by leaping from place to place, kind of in a hulk-esque way. He is a big guy, bulky and mean, so it is pretty cool to see him land next to you and take a swing with his mighty axe. But, as you may have guessed, not everything was perfect at first. Deciding where to jump was the issue here. The easiest thing to do was just get the mob’s position, either the center of the mob or the leader’s position, and jump there. But that didn’t feel good so we had to resort to some of our old tricks.

 

 

In this case we used randomness again. Adding some randomness to the jumping target position meant that sometimes Phobos would end up just ahead of the mob, seemingly anticipating the mob’s movements, or a little behind them, giving them a chance to escape or head back and engage. The result was something much more interesting than having him always jump into the middle of the mob and attack everyone there.

 

Is there are morale to this story? Of course there is, thousands of hours of watching pre-Seinfeld sitcoms taught me there is always a morale at the end. It may be that little changes can make a big difference. Or it can be that one of the keys to making an AI more real is to make it fallible. Or that random numbers are awesome. But there is definitely a morale in there somewhere.

Read More

Last week, we talked about our “outside the box” solution to a particular problem with the Y displacement.  It wasn’t perfect, but we are almost proud of it (we don’t actually know what being proud is, but we read about it).

Before that, we had started talking about Sparta,  the new world we are working on. We showed a few bits of Sparta, and this week we will continue presenting new stuff!

 

Another one of the new enemies introduced in Sparta is the Rock Thrower Cyclops. He will throw rocks at you (I bet you didn’t see that coming). He may not be as tough as the armored cyclops or giants, but his rocks can do a lot of damage to the units.

 

Besides adding more enemies, there were a few things we tweaked in Sparta.

In the previous world, Delphi, you could find a lot of shops in the city, like a meat shop, a fruit shop, and so on. In Sparta, being a much more “war friendly” zone, you will be more likely to find war-related stuff . Which meant that, in addition to a change of materials and textures, there were lots of structural changes in the buildings.

For instance, instead of a meat shop, you will have an Archery Range.

And the buildings were just a small part in the process. You can push a little further the idea of a heavy armed city with props.  For instance, having some training dummies around. It really suggests that the people there are constantly training.

 

Also, to have a heavily armed city, you need lots of weapons.

 

And finally, we used the directional light to change a little the mood and the time of day.

Well, that’s pretty much it for this week, and probably for Sparta. Next week we’ll be introducing a new world! (maybe)

Read More