Difficulty and Endings

Introduction

Way back when, arcade games were fed ten pence pieces (big ones!) and the games only had one objective: to get the next big ten pence piece as soon as possible.

 

Single-screen Games

Anyone who remembers their first encounter with Space Invaders, or The Invaders, will probably recount something like this:

"What's this?"

"Space Invaders?"

"Let's have a go."

Ker-ching.

"Oh, I've been blown up."

"Oh, I've been blown up."

"Dodge the bullets then."

Oh, I've been blown up."

"That was a good 10 seconds."

Ker-ching.

And so it went on. Repeat until out of ten pence pieces. And we did, week after week.

 

Initially we had games like Space Invaders, Asteroids, Breakout and Galaxians. All of those games had a static screen display and if you became expert enough to beat a sheet of aliens or bricks or rocks then you got presented with another sheet, sometimes closer, sometimes more populated. There were no game endings, you couldn't win, it was you against the high score table. How long could you last? How far could you go?

 

We were quite happy with this. One ten pence piece had the potential to go for hours. We hoped.

 

We learned how to extend our Pac-Man games by a fair amount. We became interested in the fruit displayed at each level, and to find those out we needed a strategy. The first couple of levels were not too difficult after a bit of practice, and one of our number, Gary I think it was, managed to find out a route that you could take on levels 5 or so to 21 or so, that would get you all the power pills and the ghosts wouldn't catch you. You still had to get it right, and there were some tense moments. After level 21 or so, the ghost algorithm must get modified slightly because the route didn't work any more. We never went much further than that.

Multiple Screen Games.

A couple of newer games in the arcades started adding some different sheets. There was one with a 3-stage rocket that you had to dock the bits together for more shots: Moon Cresta, that had about 4 different attack waves. Then there was a game that we called "Cosmic Bunny" because the last stage had a big mothership to break through and in the middle was what looked like a floppy-eared bunny. Still though there was just a sequence of a few different sheets and then you went back to the first one, presumably with the difficulty racked up by increasing meanie speeds, accelerations, numbers or firing rates. Rarely did we manage to get through all the different sheets to find out. The games were being made harder as we learned various play techniques.

Horizontal Scrolling Games

Our pixel worlds then got expanded by Scramble. Now we had a horizontal scrolling tunnel system to navigate. With a bit of coaching I did eventually manage to get through all 5 stages of this game, and I got dumped back at the beginning of stage 1 again on completion. I couldn't repeat the feat on the second pass, so I'm guessing the algorithm started racking up the missile launch rates and fuel usage.

3D Worlds

Not soon after, we got Battlezone. Your tank gets dropped off in the middle of nowhere and other tanks appear and try to shoot at you, one at a time. Sometimes we tried to make it to the hills, but we never got there. You get only 2 types of obstacles: pyramids and cubes, and 3 types of enemy: tanks, super-tanks and missiles. There are bonus saucers too that don't fire at you. There's only one enemy at a time so how hard can that be? Well, the only tactic we found that worked was that you have to reverse away from the enemy, with a bit of turn so they can't target you, and hope that you don't reverse into anything. Of course as soon as you do reverse into an obstacle you won't have time to get out of that mess before the enemy gets to you. You only get one bullet at a time. If it hits something you get reloaded quicker than if you miss. The missiles became more agile and erratic with their turning so we had to wait until they got really close. If there's anything else after that, we never saw it. Our only improvement came when we could get the saucers quickly.

 

Vertical Scrolling Games

Shortly after that; the backgrounds got bigger and prettier as we got vertical scrollers like 1942, Xevious, 1943, SlapFight, Flying Shark, and Terra Cresta, amongst others. This was more of an influential era for me as I had started writing my own games. I'm guessing that the arcades went 16-bit at about this time so they got more memory and more processing power. Those games started to introduce power-ups and the concept of  restart positions, As they all scroll slowly they introduce challenges at a controlled rate and since the enemy are all related to where you are then the difficulty can be ramped up in a controlled manner. The concept of continuing (for an extra ten pence piece) was also introduced, which satisfied the machine's insatiable desire for money, and let players advance through the game to... the ending? Or did we still not get an ending?

 

Graftgold produced some 8-bit conversions of Flying Shark. We were drafted in at the last minute so we were working to a strict time budget. We borrowed the arcade machine and videoed  playing the game all the way through. It has 5 major sections, broken up into about a hundred restart points. There's still no ending, you get to go back to the beginning and start again. We didn't get any official documentation for this game, so we weren't sure if the difficulty level was then raised, but we likely did do that. Flying Shark is a pretty tough game in the arcades. With a lot of continues I did make it through, but I can't do it now. I can get to nearly 40% through on my first life, but the loss of all your power-ups every life means that you struggle to get all your power-ups back before you get shot to pieces again.

 

When it came to Rainbow Islands, we found a few big changes in the overall approach of the game. We got the arcade machine to study and again we videoed our arcade genius, David O'Connor, playing the whole game through. Firstly, some of the power-ups can be made permanent, which is definitely a target to aim for. You can make the single speed-up shoes permanent, the double rainbows and fast rainbows permanent, even the flying wings. This gives the player an opportunity not to get overwhelmed on a restart. Next, we found two different endings to island: the Dragon World. Depending on what you've done up to that point, the game ends, with a short sequence, and a hint that there's a way to do better. Our conversions did stop at island 7 as well, since we had quoted on the 7 islands that everyone knew about. Only later did we get a surprise.

 

Little did we know that there are 3 more secret islands, and 2 more endings. The last 3 islands are even bigger than the 7th, and the 9th brings in way more colours than we would have been able to do. We brought the ultimate success ending forwards to the successful end of the 7th. We also found for only the second time, that there are cheat codes in the game. Each secret room has a code at the top, and if you put in that code with button presses and joystick moves on the start page, you can start with the item that was in the secret room that showed the code.

 

The meanie speeds and fire rates have a base value per island. Algorithms might be repeated with new graphics and extra features, but they slowly ramp up as you progress through the islands. The meanie speeds and fire rates also include a factor based on how many secret rooms you have visited. Thus, the better you play, the more the game pushes back.

 

The only other arcade cheat mode I know about is SlapFight's. If you start the game and don't fire at all, and you get shot, you then restart with all the power-ups except any speed-up, which you can gather pretty quickly. You come back with about 3 wings on each side, which do tend to get shot off pretty quickly, but having homing missiles immediately means that you can make good progress at least to the first big tank. Shortly after that, the pesky blokes wandering through the protective tubes that fire in 8 directions when they reach a junction nearly always get me.

 

Ivan Ironman Stewart's Super Off-Road Racer

We also converted Super Off-Road, as we called it. We videoed the game over a couple of levels but it had a more simplistic approach. Firstly, there was no ending, it gets harder and harder, as you might expect, but I'm sad to report that it does 'cheat'. The game difficulty ramps up from level to level so the accelerations and top speeds of the computer-controlled trucks rise. It takes the time gap that the player wins by to decide how much to ramp up. The guys here realised that if you just about win each race by waiting around before the finish line until the computer truck almost gets there then you get a longer game. The game also gives the computer trucks extra turbo power after a couple of levels to make sure that they win! Evil incarnate.

 

We also got a peek at some of the original Z80 code for the game that figures out the wheel positioning across the landscape. It's possible with a detailed non-flat background that all 4 wheels may not touch the ground, or that the suspension may get compressed too much on one corner. The algorithm tried to resolve the positioning by adjusting the chassis position for each wheel in turn, but that can lead to unresolved issues. There was a loop counter that tried 20 times, and if it hadn't sorted it out by then it just fires the truck upwards at speed. You might see that happen on any truck on the arcade machine. I believe we tried initially to use that algorithm, but Steve did a rewrite to improve it. Probably less of a squirt upwards would have been a good start.

 

Game Ends

So should games have endings? Well, they tend to need a lot of dedicated graphics, so I'd say that on the 8-bit machines it's an unreasonable request, unless you're going to do a tape load, and I never liked that idea.

 

Most of my games followed the normal arcade idea of the time to just let the player keep playing on to get the highest score they could. I decided though that Intensity should have an end sequence, since it was more of a puzzle game where the player can head to any of the five exit platforms, where the alpha side was the hardest, and the epsilon side the easiest. The end sequence was designed to use the sprite multi-plexor to turn what appears to be background characters into a large moving object. That's all I'll say.

 

C64 Uridium doesn't end, per-say, The player gets insulted by the meanies to spur them on to the next iteration of taking on the dreadnoughts with elevated fire-rates. Uridium 2 does have an ending. Having taken the technological leap to load sets of 4 dreadnoughts, meanies and graphics, we figured we might as well load an end sequence. The graphics artists were keen to draw some bigger graphics for a home-coming sequence. I can't help thinking that we could still have let the player carry on after that, but it would take an hour's play to get to that point, which is enough adrenalin for one day.

 

Seeing the end sequence is supposed to be a reward, but should it be available to the few, or the many? That's down to the game's overall difficulty level. If Rainbow Islands can regulate the meanies so that they get a little fiercer the better you are, then everyone still gets a chance, but there is a required investment of time to learn what's going on, such as how do I get the correct coloured gem that I want?

 

Games seem to have tried to change into mini-movies. Movies have beginnings and ends, so it does seem more appropriate that these types of games have end sequences. They can re-use many of the 3D models for the game, so why not? Other games have become simulations, like F1, so presumably there's some continuity to the next season, though there are likely some celebrations for winning too.

 

To End or Not to End.

I don't see that games HAVE to have an ending sequence. If it's appropriate, then of course there should be one. I haven't seen many end-sequences, I'm not that good a player, and I don't feel inclined to cheat. There's more curiosity about how to complete the game properly, which leads me on to how hard is the game to complete.

 

I've played many arcade games where I know I have no chance of getting through them. I've tended to find the Irem games way too hard for me. I can't make a dent in R-Type II, nor Dragonbreed, though I continue to try as they are fabulously pretty and I want to see more, but as money-making arcade machines I see why they need to get the player off the machine as soon as possible.

 

Similarly I've played games like Space Harrier and Aero-Fighter where you can buy your way through the game with continues. I also have 2 different versions of Kyukyoku Tiger, one of which I can't get through the first big boss, and one where the game is so easy it's trivial, badged as Twin Cobra. The difference is that Twin Cobra is a two-player simultaneous game. The functional difference appears to be how and where the player comes back after losing a life. In the single-player K-Tiger, you get sent back to a previous restart-point. This buys you time to rebuild your weaponry a bit before you get back to where you were the previous life. With Twin Cobra, even in one player mode, the game keeps scrolling on, except when it stops for big-boss encounters. This is because as a two-player simultaneous play game it has to leave the other player going. As an arcade game collecting money, it works fine. It's tuned to be tough for 2 players, especially the big-bosses, and you have no way to re-accumulate weapon pickups because you're just thrown straight back in the ring. There's also no let-up in the onslaught either. No contrast, no time to regroup. The only way to beat the big bosses is to lob all your bombs at them, die, come back with 3 more bombs and repeat.

 

As a computer game where you're not paying 10 pence a play, you can force your way through and at least see all the graphics and big bosses. Being up for a challenge then, I have played a lot of K-Tiger because I want to beat the first big boss and still have some bombs for the next level. I only managed that once! Now I have played Twin Cobra I have seen what's coming after. It's not pleasant. My conclusion is that I'd play Twin Cobra in the arcades (with someone playing P2), but there's no challenge to it at home because it's easy to progress, but not without losing a lot of lives, which doesn't matter because I'm not paying. Does that make sense, or is it just me? I want a challenge that I can learn to beat. That's what generates adrenaline and makes a game fun to play. If you know it's nigh-on impossible then the instinct is to go play something else.

 

To Continue or Not to Continue

Reviewing every game I've done recently, I've never put in a continue option in any original game.  Rainbow Islands was an arcade conversion of a game that does have a continue option. Continuing is purely a mechanism for arcade machines to get another coin out of a customer. It's no problem for the arcade machine, you have a paying customer who's going to last even less time than starting again and all you have to do is show them a bit more of the game. There's no real reason to do it in a computer game. Now if I could get 1 pence in an online transaction per game played and the game was downloaded free... Perhaps continues would cost 2p... 

 

I've seen people complain that there's no continue option in Fire and Ice. Well, if there was, I'd clear the player's score each time, though some games don't, and probably you could force your way through the whole game quite quickly, and then you'd see the ending and likely never play the game again. For me, that's not what gaming is about. You don't get a continue option in real sport. Oh, my team lost 2-1, no wait, I'll invoke my continue clause and we'll play a third half, and a fourth, until such time as I'm winning. Of course there are "Fergie Minutes" when Manchester United are playing extra time and need 2 goals...

Gaming should be about getting better, knowing that the game is not impossible, which for computer games means that we have tested the game and played it through. Knowledge is gained every play, you start to know what's coming next, whether exactly or roughly. The achievement is getting to the next level, or the highest score, not just seeing what's in there.  

 

I did put in a training option in Fire and Ice so that you could go to each of the first 4 lands with 9 lives and have a play. That gives the player an opportunity to see the landscape and get some practice. Paradroid 90 lets the player start on any ship previously reached, so again you can practise and learn. Uridium 2 does something similar too. 

 

If we go to all the trouble to create an end-sequence for the few good enough to complete a game, then it does seem to be rather spoiled nowadays by someone posting a video of it online. Endings are going to have to become more customised so that it's different every time. That'll require more thinking about, maybe using effects rather than graphics.

 

In conclusion.

Ultimately, the objective is to make something that's fun and interesting to do or play. It doesn't matter to me whether there's an ending sequence. Did I have fun getting there? And that's down to the playability.