Game: Gauntlet – Seven Sorrows

Gauntlet: Seven Sorrows was the last iteration of the venerable Gauntlet series originally created in the golden years of the coin-op arcades. While this new Gauntlet aspired to be something special during development, it didn't get a chance to realize its full potential. In the end it turned out to be a somewhat mediocre hack'n slash affair with decent looking graphics but not much else going for it. The game's greatest accomplishment is that it overcame serious time and technical hurdles to (at the very least) ship.

My contributions

Level scripting. When I joined the project near the beginning of full production the level layouts already existed and it was my job to create enjoyable encounters or puzzles. There were a few layout changes that were made upon my request but those changes were entirely in the environment artist’s hands to provide since the levels had proceeded beyond the gray-box phase of development.

Ancient Forest

This forest level was laid out in a simple loop with a rather large temple at exactly on the opposite side from where the player spawn point was located.  There were also two camps, one on the left and one of the right path, leading up to the temple and surrounded by wooden palisade gates.  If I recall correctly, the original intent was to have the player choose either the right or left side of the loop, because neither side afforded the player any real advantage making it a superficial choice. More often than not when branching is introduced in level design the net result is that it really doesn’t matter which way the player goes as the level of difficulty and consequently the rewards involved will be about the same either way.  Also, eventually the paths have to rejoin otherwise it gets very expensive on the production side of life.

Nearing the end of production, when it became clear that the scope of the game was to be drastically reduced, it seemed obvious to me (as well as others on the team) that trying to eek out every minute of game-play was going to be pretty important. I can’t say for certain how short the game would have been if not for level changes designed explicitly to extend the experience, but short enough that I strongly felt the game would be a rip-off at the full price-point for last generation console titles (about $50). So, in an unusual about-face on my normal principals I chose to make changes that put value first instead of fun hoping that both could be accomplished.

I spent the next week or two fighting the tools (in a solid toolset most of these changes could have been made in a day or too) to modify the flow of the level entirely through scripting.  In the end the solution was to have players run the entire loop 1 1/2 times before getting to the boss who was spawned at the temple.  While that sounds like a simple solution, it becomes pretty obvious to the player that they are running on a treadmill (or maybe more like experiencing a chase scene from Scooby Doo where the background repeats) and once they feel that emotion, instead of staying immersed, well…oops. My bad — sorry guys.

I certainly made the level last longer but in the grand scheme of things is it better to have a level be very short but more enjoyable or longer at the sacrifice of a tighter or less repetitive experience? Since the common complaint of reviewers were that the game was too short and too repetitive even to this day I am unsure of whether or not I made the right call on this level.

Dungeon Deeps

This, as the name implies, is a dungeon. It is a very dungeony dungeon, full of jail cells and danger…with sometimes the danger being spawned from aforementioned cells. Did I mention that it was a dungeon?

Unlike the Ancient Forest where the layout wasn’t optimal, the level layout for Dungeon Deeps was just plain tiny and by the time things started to heat up the level was over. So again I went about looking for ways to add content without any drastic changes to the layout. I did manage to get a few tweaks to the level layout so that I could implement stories (or floors) in a very hackneyed way by teleporting the player to the beginning of the room and changing the encounters which added a few minutes of gameplay. This more flexible layout also gave me enough control to add a transition point from imperial guardsmen to trolls whom were at war. The most interesting choice that could be made in this level was to allow the two sides to fight one another, thereby making the battle easier for the player, or just jumping in.

Underworld Crossing

This is my favorite out of the four levels I scripted but it also had the most changes to the layout which helped immensely. Underworld Crossing can be described in the form of a question, “What if Moria [from Lord of the Rings] had been built by elves instead of dwarves?” The visuals of this level tried to answer that question.

What made the level special in my mind was that a focal point was added later on in development upon my request. The large female statue which the player revisits a couple times acted as a solid visual anchor but also helped to communicate how close the players were to finishing the level puzzle. The statue itself wasn’t a new asset, it was relocated from a level that had been cut, thus it was a pretty easy grab and something of a morale win for the artists since it was a pretty cool art piece that would have otherwise went to waste.  As a note for other designers — if you see something cool that is about to be cut, but could be saved through use somewhere else, don’t hesitate to ask for it if it makes sense in your level. Usually a level isn’t hurt, at least from the player’s perspective, by having too many cool things.

What I also liked about Underworld Crossing was that since the scale was significantly larger than previous levels such as Dungeon Deeps it didn’t force a lot of backtracking or tedious navigational challenges to make the level last longer.

Deep Fortress

This level is an extension of the same visual theme as Underworld Crossing but a little more urban. I didn’t have to do all that much to make the map work and the layout didn’t receive any changes, but it really doesn’t stick out as anything special from a design point of view. I added a couple timing-based puzzles centered around activating crystals to this level; however, those were there to slow the player down and provide ambush points more so than being intellectually stimulating.

Otherwise, the level was a typical hack’n slash fest lacking in truly memorable moments. That may not be the worst thing pacing-wise since the next level was also the final and most spectacular boss battle in the game.

Image Gallery

Please note: If anyone has any additional screenshots of the above levels, please drop me a line!

How Levels Were Made

In Renderware Studio! For those of you who never used these tools, be thankful. I won’t go into the gory details here but there were reasons many of  levels were cut  from the game in the last few months before ship and RWS was certainly one of them. I have no idea what improvements Criterion, who was acquired by EA during all of this, may have made to the toolset but in 2005 they were not well-suited for full production.

Aside from working with tools that didn’t, the process was pretty simple — greybox in maya, hook up the interactive elements in Renderware studio (this is where I came in), test and repeat. Well, it wasn’t quite that simple. The levels reached the outsourcers hands before they could be considered fun which in turn tied the designers’ hands when making layout changes that could of helped. Given the aggressive schedule put forth that was probably the only way to do it, at least from a scheduling point of view, but going waterfall on what needed to be iterative process resulted in some pretty lackluster designs.

The World’s Smallest Post-Mortem

What went right

  • In the end, the team pulled through and finished the game. This was no small feat given the circumstances.
  • San Diego sure was nice :)

What went wrong

  • The design tools were unreliable, slow, and often unpredictable
  • Many consistency problems crept into implementations between designers which resulted in a lot of rework. This was largely due to the tools and the complexity of implementing anything in them but could have been reduced through better communication
  • The scope of the game was drastically reduced, both in terms of features and content, at the tail end of production to hit the holiday window. As a consequence we lost some talented folks and moral, already low, tanked completely inside of design

Upon Further Review…

The bulk of reviewers out there felt that the game was Gauntlet. It wasn’t anything special beyond the tried-and-true formula, so if you liked previous Gauntlet games then you probably liked this one. Seven Sorrows certainly didn’t change the minds of any gamers whom were previously not Gauntlet fans nor did it endear itself to any new audiences. Financially speaking, the game resulted in a significant uptick in Midway’s financials for the quarter but at the expense of killing the franchise’s longterm appeal. I hope Warner Brothers, who I think now owns the IP, considers dusting it off at some point and gives the world another chance to experience a new Gauntlet title.

The lessons I will take away from this project have more to do with how to make a game than how to design one. If the tools are bad, design becomes non-iterative, and the scope is reduced drastically in the 11th hour, then making a good game becomes almost impossible. In hindsight it is easy to see that. There were other key lessons centered around the people that made the game but it would be unprofessional for me to bring those to light here.

Additional References

For more information on Gauntlet: Seven Sorrows, please check out the following links: