Hi all

In the weeks leading to the release of Heroes 5, I`ll put up weekly dev letters. The `weekly` part is not totally guaranteed though

Each dev letter will treat a specific subject, the first one to be published in a bit being about the town construction interface. If you have requests about a subject, please say so.

Heroes 5 dev letter
#1: the town construction interface

The town interface in Heroes 5 has been a typical case of usability versus visual quality. We are reaching the final version only now, and I`ll describe a bit the process it has gone through.

The very first interface test we made was actually quite close to the final solution being implemented now. It was half a screen used to display all available buildings (quite similar to previous Heroes games), with the whole town itself shown on the other half.




Another test was made later last year, showing only the town itself, with a camera centered on the building to be done. You would cycle through all of the buildings one by one, getting the view centered on them, switching from one position to the next. With this solution you could see the town details up close and the graphical quality was used to its full extent.



It was much too slow though, and it was too hard to understand how you could build up to a specific building, so the cityplan was introduced. There all buildings were shown with logical links, and with a status (a color code) to show what was already built, what could not be built for some reason, and what possible choices you had at this stage. The full screen was used for this feature, and when you clicked on one building you would switch back to the town view with the corresponding building selected.

This seemed like a good solution but was still quite slow, and needed the player to switch between different screens for the simple operation of building a new structure. The betatesters have witnessed this solution being integrated over the end of last year.

Finally we decided to go back to the original solution, mixed with the latest one. The cityplan would be displayed over one half of the screen, and the building would be shown zoomed in on the other half. You would use the cityplan to select and choose which building to make. A last twist is the added option to either display the whole construction tree, or only available buildings on the cityplan.



Hopefully this one will work ;p

Heroes 5 dev letter
#2 : handling a beta test

It seemed quite obvious that we had to include the fans into the development process in some manner. I don’t think anyone can be considered as being the ultimate expert on Heroes questions, but having a few dozens or hundred people sending you ideas or feedback helps getting everything into shape.

The risk however, is to get overrun by the feedback, or to give in to all demands. Most of them are contradictory, and quite a part would be conflicting with other decisions or would alter the direction we are giving to Heroes 5. So it is a huge work to integrate these multiple opinions together and mash something out of that. Simply keeping track of the feedback during the closed beta test was a headache – but most communities have helped getting summaries out.

A disadvantage of making beta tests is the amount of work required to prepare versions and update them in a timely manner. It can disrupt the production process for a long time, and either you get late delivering beta updates, or you get late actually making the rest of the game, and most often you get late in both respects !

Finally, a disadvantage I see now is that, at some point, this has to end. Beta testers, who were frustrated at the start because of the amount of bugs, and also because the changes were not necessarily going to their liking, become frustrated again as they feel they need to be a part of it even longer, until the game is perfect or close to it. But as we enter a polish phase, treating a flow of feedback becomes less interesting and more dangerous : you need to nail down decisions on design points, and constant rediscussions only lead to delay. A game can only be as perfect as production allows.

On the other hand, running a beta test is an exhilirating experience ! You get in touch with hundreds of fans from all over the world, you get lengthy debates, historical summaries, polls, etc, all over your head. It is also a challenge to keep the exchanges running, and to keep the development in check with the requests.

Preparing and selecting people is interesting in itself. With the registration form we have put up, a very precise image of the typical Heroes fan was laid out, thanks to the 15.000 registrants. The average age of the registrants, for instance, was logically quite high, above 24 years old. A large (and untended) community was `discovered` in Israel. The hardcore aspect of the russian fans was proven. The vast fields of casual strategy players in the US, whom we need to get to try Heroes, was underlined. We could also sort out which were the influential communities

Then you have to organize the work in advance with communities and internal testers. We had to make out with the constraints and tools we had, in terms of forums, bug reports, etc. The number of spoken languages was a challenge in itself. Each community created language-specific topics to harvest the feedbacks. The update of activation keys had to be taken care of : we added one month of closed beta testing, and thus needed to distribute new keys around. We just ran into a similar problem with the open beta : too many players downloaded the version when compared to our plans

In the end, running a beta test is about getting feedback, and we had a truckload in this respect. Even by weeding out the duplicates and the unrealistic suggestions, we had a lot more than a thousand items in the list to browse, in just three months. I think around a large third of this has been integrated, although not quickly enough for the beta testers to actually see their ideas in the game.

Besides suggestions, getting the general feeling on the game was needed. At the time (september) we had been working on the thing for almost two years, and we started to have a blurred vision of the priorities. The E3 and Leipzig shows garnered very positive feedbacks, and all journalists were already enthusiastic, but we needed to get players` opinion. This was also made possible by the discussions on the forums with the fans, and with the communities. Being able to get an opinion prior to release is invaluable.

Finally, running an open beta test after a closed one is about reaching out to more people. You present the game – in its current state, it was not a demo and wasn’t polished – to a vast public besides the first fans of the closed test. I think this challenge was very correctly met, even with the downfalls of the version itself

Since Fabrice didnt have time being so busy with the developing of this great game for all of us I have the privilege to present to you the third dev-letter which features the much debated issue of the battlefield.

Enjoy!


----------------------------------------------
Development Letter #3: The Battlefield
The battlefield is the real heart of the game, the place where all the development and recruitment strategies are put to the test, and also the place where you can spend half of the gaming time or more. So this is an area of the design where a lot of effort has been made.

We have started making tests on this subject around February 2004, which is already more than two years ago. At that time the question was about what system we would be aiming for, how it would be shown, how cameras would benefit from the 3D action, and how the interface would bring in additional gameplay or serve the action. The underlying question was about the size of the arena.

We studied the previous games and a worrying tendency was established : the ratio of creature size to the whole combat screen was declining very rapidly with each new version. If we kept such a tendency the average creature would not be more than a speck on the screen in Heroes 5. This is something that you can accept when you have a realtime game with hundreds or thousands of creatures displayed, but not when there are about a dozen. Plus, we wanted to have multiple sizes for creatures, and that would increase this problem again, with the smaller creatures (like the imps) being only a fraction of the others.

Besides this visual question, there was also the major objective of increasing the tactics during a fight. The designers at Nival were planning to give each creature a special attack pattern, like a breath area, or a shooting line. This idea would increase the need to search for specific situations where the position of enemy units would be used against them through a death blow. But for this idea to work, the size would need to be kept under control, to force tighter hand-to-hand brawls and decision-making. This need was from the start in line with the visual issue mentioned above.

Finally, a core objective on Heroes 5 was from the start to get a faster gameplay. As the battle screen is a place where you can spend most of the playing time, it is also a place to try to gain some time, by reaching conclusions faster. Most of the early fights are usually one-sided, hence the idea to create a smaller battlefield specifically for the first part of the game.

This is how the research on variable battlefield started. We would have a small battlefield, of 6 by 8, when the army was composed of four stacks of creatures or less, and the battle would switch to a larger one, 12 by 14 for larger armies. The early fights would be done much faster this way. The siege battles would in any case happen on a larger version of the battlefield.

It seemed ok during implementation and the first internal tests. However the larger battlefield fights seemed a bit too long though, with less interest, as the tactical element was much more difficult to trigger. There were much less occasions when the creatures placement would enable a special move to become interesting.

Another problem appeared thanks to the beta test : the balancing was turning into a headache with multiple sizes for the battlefield. It is not only about giving more movement points to each creature, but also balancing area effect spells, shooter and caster capacities, plus all the side effects on attack, defense values or hit points. It appeared to be too complex to try to balance close to 100 creatures efficiently.

So we decided around November last year to come back to a single size between the small and the large one, a medium 8 by 10 size. This would allow quick fights with good opportunities of tactical effects, and was also within range of the visual limits we wanted to keep to benefit from the animations and effects.

The objective was also to limit the strategy of creature splitting, where you basically end up with only one type of creature on the field, presented in several stacks. The gameplay in Heroes 5 is built around the need to use all creatures from one’s faction. Low-level creatures are still interesting in the late game, and high-level creatures are very costly. So a small or medium sized battlefield is more interesting for a mixed lineup.

We also integrated the ‘tactics’ phase for all heroes, so you can place your troops the way you want at the start of any battle. This allows you to adapt to the obstacles on the battlefield, as well as adapting to your foe. It also avoids a very initial phase of battles where you would have to move your troops in the configuration you want.

However during the final stages of the beta test where this medium battlefield was provided to the testers, the almost unanimous feedback was that this was too small, and that the fights could quickly become cramped. A bit more freedom was needed, increasing width and height by one or more cell.

After quite a bit of discussions around the christmas period and january, we decided to increase the size to 10 by 12. This size globally retains most of the advantages of the former one, while allowing more movement possibilities.

All in all this issue strikes right into the fine balancing between conflicting gameplay objectives, and nailing down bit by bit how the game will look like in the end. It also underlines very clearly the utmost importance of players feedback !

This message has been edited. Last edited by: Alderbranchh,