Wednesday, 12 June 2013

1800: A postmortem

It is pretty much a year to the day that I penned my last postmortem for Astavoid. At that time Astavoid was my first game and the amount learnt during that process was immeasurable. However, what it did do was open my eyes to just how difficult it is for an novice, indie developer to make a difference and just how difficult the goals set out in my project would be.

Roll on a year and I can safely say I have learnt a lot, lot more. Since that article I had six months of what felt like going round in circles but this led to a period of focus around Christmas / the turn of the year. The two factors that have contributed to the focus was the movement to Unity from Corona, and the other was the one game a month initiative.

To quote the one game a month (1GAM) website for those not aware of its goals: "Starting a game is easy. Finishing a game is hard. One Game A Month is a gamedev challenge where everyone's invited. Your quest is to create one game each month for the entirety of 2013. It might sound crazy, but the truth is that it isn't all that hard."

So this is exactly what I set upon during the early part of 2013 and actually at the time of writing is what I am adhering to. It is because of this approach that 1800 came about but it almost didn't happen.

During March 2013 I had just finished my game 1GAM for February and decided because Easter was falling early that year I would create an Easter game.  What I struggled with was a decent concept and played around with a few prototypes for a week or so.  What the 1GAM concept had taught me was rapid focus and discipline, and therefore if it isn't working move on. If it is unlikely to be completed within a month then its too complicated. If you can't explain your core game in a sentence, it is also too complicated.

The Design
With only 20 days of the month to go after my deliberations I set my thoughts on a design that was more about playability than graphics.  Something that I could programmatically generate levels to add content rather than bogged down with design.

I hit upon the idea for 1800 almost immediately.  Inspired by games like hundreds and gauge where style and accessibility win out over fireworks and sparkle, I felt that this would fit well with my new found approach and limited timeframe.

Additionally, I wanted to keep the file size of the app small and already at a disadvantage with Unity iOS Basic's unoptimised /compressed framework I didn't want to bloat with large spritesheets.

All these factors brought together led itself to a minimalist design that unwittingly hit upon an en vogue concept of flat design.

I won't talk at length about flat design within this blog as this article does it far better than I could, however, it essentially means applying no gradients, shadows, bevels or embossing to your graphics and having thereby having just flat coloured assets. This coupled with a hybrid pastel colour and 1800 immediately felt very appealing to the eye.

The main criticism, and rightly so, of Astavoid was its lack of depth.  One of my main goals I wanted to achieve from 1800 was to make it more "sticky" and by that I mean keeping players in my game longer.  While 1800 wasn't about making money but more tailoring my craft it was important to have the game engaging and encourage another quick go.

The mechanic of the game was simple.  I had created a rectangle and the player must stop a cursor as close to the middle of this rectangle to achieve as high a score as possible.  I would calculate the number of pixel offset from the middle (central axis) of the screen and subtract from a total score of 1800. Great I had one level now what.  Based on this calculation I started including lines and further arrows and carried out further mathematics on the differential between alignment to central axis.  This gave a little bit more depth particularly when varying the speed of arrows and lines.

Within about 8 days (25 hours development) I had finished 1800 and was ready for people to have a play and provide feedback.  I decided upon using Testflight to let some of my very helpful twitter followers test the app.  I cannot under estimate just what a useful step this was in the development of the game.

I'll be honest I was nervous when I released for testing.  Like a proud Dad I didn't want anyone to be disparaging of my efforts but even worse tell me it was a good game when it wasn't. What I was pleased with was just how consistent the feedback was from my 12 or so testers and that in the main very positive about how addictive it was despite its simple facade.

My twitter followers are probably an over critical bunch.  I mean that in a good way in that they fit a demographic of game developers and seasoned gamers.  I too fit this and having played games for nearly 30 years a game like 1800 wouldn't normally jump out at me as something I would be looking out for.  However, there was something about its mechanics that really appealed, not just to me, but to my wife and family as well.

A slightly discouraging feedback that came from the testers was that they had completed the 20 levels in under 10 minutes. This wasn't really what I was looking for when talking of stickiness in my earlier comments.

Some more feedback was having implemented a unlocking level star system, a la Angry Birds, it wasn't clear to the testers just what triggered an unlock nor was there really any incentive to go back and improve scores once all 20 levels were complete.

In addition, the most important feedback came back about my UI design and workflow.  I had overlooked a major factor in fuelling my stickiness and that was providing a mechanism to retry a game immediately.

Based on the feedback from the testers I went about implementing some new features that would provide the goals I was looking for.

First of all I changed the mechanic of the star system to only provide 3 stars if 1800 is achieved on a level.  This on some levels was very challenging so would hopefully provide the desire to come back and improve scores.  I also introduced a new game mode called time attack which was unlocked after 40 stars (out of a possible 60) was obtained thereby providing further depth to the game.  Time attack was using the same scoring mechanism but this time over 60 seconds rather than just one shot.

The next big implementation was to add a retry button into the in-game menu system that allowed you to replay the same game immediately or go back to the level select.  In hindsight once implemented I had no idea why I didn't do this in the first place but I guess thats the value of beta testing.  

A few minor iterations of testing followed but at this point I was pretty happy to stick it up on to app store and kick off the review process.  This was around the last week of March 2013.

Measurement and analytics
So that I could ascertain whether this design concept had been successful I decided to include the mobile analytics SDK from Flurry, to understand how my potential players interact with 1800.  This I felt would let me know remotely what is and isn't working and adapt appropriately with updates. I also included Flurry's interstitial ads for a little bit of monetisation.

Additionally, I added Adwhirl into the game a decision I didn't take lightly as it went against the minimalist design of the game but I didn't want to completely rule out making money.  However,  I strongly felt the design and concept of the game didn't lend itself to In App purchases.

More on the analytics later.

While I was waiting for the game to be reviewed by Apple I started on my marketing trail which would consist of twitter and forums. In my experience of Astavoid I encountered there are many reviews sites willing to take my hard earned cash.  Equally I didn't have an advertising budget given that 1800 would be predominantly free.

Instead I decided that each day for about 10 days I would post a teaser image or description of the game on to my twitter account (@gameproject10k) and the touch arcade forum.  At the very least I would get the name out there and a little bit of interest.  I wasn't expecting much from this approach (given I only have 250 twitter followers for example) but was more than I did for Astavoid.

1800 was approved in the first couple of days of April 2013 so I had decided to launch on the second Thursday of the month, the 11th.  This allowed me to step up my twitter campaign, begging for retweets etc as well as drumming up a little following on TA forum.

I had decided to defer the prMac press release (I did this for Astavoid too but on launch day) until a few days after 1800's launch in an effort to understand what effect this had on launch day.  It is very difficult to understand the worth of a release day press release campaign when twitter bots look out for new apps and tweet its arrival.  The only metric I had was I created a prMac press release as the only thing I did for Astavoid and to this day (June 2013) the launch day downloads still account for a third of its total.

Launch Day, downloads and stats
The day arrived and the usual nervousness kicked in as to whether it was going to work on all those devices, whether I had screwed something up when creating the release and generally how it was going to be perceived.

I had set myself the goal for 1800 of approximately 5,000 downloads in total.  The reason for this goal was to improve on the 3,500 Astavoid downloads had achieved in the 12 months prior to writing 1800.

Launch day was great, I managed to get over 2,100 downloads which beat my Astavoid launch day without any real PR. While taking a break from constantly refreshing chart positions and reviews I took a break to read the Toucharcade App to see what other launch day apps were out and I was competing against.  I was very pleasantly surprised to be greeted with the following:

Amazing.  The game had found some love on the Touch Arcade forums and like my assessment earlier forum members felt they shouldn't like it but they just did and kept coming back for just one more game, time and time again. Initial feedback was that perhaps my goal for stickiness seemed to be on track.

But this euphoria was short lived, no lack of exposure on the TouchArcade forums could stop the next day's download stats from being lower.  But realistically what should I expect given this was really the only marketing push it was getting.

That said the next big push was coming on the following Monday from prMac so I sat back over the weekend, refreshed the website, rankings and charts on a one minute cycle and see what would happen. By the Sunday just 4 days after launch 1800 was getting just 254 downloads, I say just as in context this was disappointing given the launch day.  However, on reflection some games don't even get that from their entire lifetime.

On the Monday, my prMac press release went out, a few TA forum quotes and a little background about what I was trying to do and I was ready to see its effect. I timed the release to go out mid morning in the USA which would still be picked up by european agencies in the afternoon. 

This I thought were decent tactics and better for me to measure how useful the prMac press was to future game launch campaigns.  However, in a good way, this was eclipsed by a feature Touch Arcade ran on 1800. Titled, '1800' - A Surprisingly Difficult Reaction Game, a fair appraisal of the game mechanics and a little bit of retro love put 1800 in a good light. I was naturally delighted and amazed.  The Toucharcade forums had clearly caught the staff writer's eye and despite the number of games that must fall across the desks / devices they took the time to comment on my efforts.  This has an enormous effect upon the next's week's activity.

The downloads went (relatively) mad, on the day of the PR launch I had 1,000 downloads, the next day to coincide with both the PR piece and the article I had 3,000. Over the next week the downloads halved day on day but by the following weekend I was still around 300 downloads a day.  

During this time I had been shortlisted and finished third in the Touch Arcade game of the week poll. Beating Mr. Crab, amongst others, which happened to be App Store's editorial choice for that week.

1800 was featured in a few countries' puzzle pages which certainly had an effect on sustaining chart positions in their respective countries.

The events just discussed make an interesting download trend diagram.  The usual whale's tail is in place but with a few spikes.

If you refer to the download graph below you will see that the first spike is launch day, the second the effect that prMac and the Touch Arcade coverage gave me. The third smaller one towards the middle of the trend was on the eve of the iTunes 50 billion app download.  While I can't be sure this was the case for the latter I can't find any coverage or article that could have accounted for this, nonetheless nearly 900 downloads were achieved on this day.

So without question from a downloads perpective this has eclipsed my goal of 5,000 downloads and a small one off feature having quite an effecrt  At time of writing (early June 2013) 1800 has been downloaded over 15,000.  Not bad for my second game that was written in a little over 25 hours and no marketing budget.

Chart Positions
So what countries did 1800 fair best in?

    Games      Games/Action       Games/Puzzle     Top Overall
Spain 33 14 7 106
Argentina 61 28 11 241
Paraguay 76 34 11 315
Poland 52 26 13 182
Grenada 110 45 14 264
Ukraine 57 23 18 211
Venezuela 95 44 19 291
Portugal 70 21 23 270
Chile 119 53 23     -
Domnican Rep 168 61 26  -
Guatemala 110 47 27 397
France 131 49 28   -
Peru 146 57 29   -
Italy 170 56 34   -
Mexico 174 65 34   -
Slovakia 154 55 35   -
Czech Rep 176 69 35   -
Colombia 199 78 39   -
Belgium 261 78 71   -
New Zealand  - 161 84   -

    Games      Games/Action       Games/Puzzle     Top Overall
Spain 90 27 19 214
Grenada 81 28 20 174
Slovakia 228 67 33 -
Peru 210 62 37 -
Czech Rep 352 108 41 -
Chile 259 76 44 -
Venezuela 333 75 49 -
Argentina 338 94 53 -
Portugal 328 79 59 -
Poland - 118 59  -
Ukraine - 109 74 -
France - 111 88 -
Italy - 165 103 -
Guatemala - 285 114 -
Colombia - 257 142 -
Mexico - 301 158 -
Domnican Rep - 357 170 -
New Zealand - 267 199 -
Belgium - 264 235 -

What about other analytical factors we spoke of earlier.  Has the design led to more sticky games.

Pictures paint a thousand words!

Session Length

Over recent weeks the median length this has tailed off a little from 2 minutes a session to just over a minute, but nonetheless a median of over 1 minute playing the game across over 35,000 sessions is pretty good given that a game on any of the levels could be played in a matter of seconds.

Games played
The statistic on session length and short game bursts is backed up by the new game event I added in which shows the number of games played per session.

At time of writing 670,000 games of 1800 have been played.  A statistic which I still struggle to comprehend that my little game has been played that many times and occupied that many people's time and attention.

The last statistic was to ascertain just what the demographic of 1800 had become.  When I talked earlier about accessibility just how far reaching had it proven to be?

As can be seen from statistics that could be obtained it does show a pretty wide audience for 1800 which aligns itself to its accessible and casual nature.

Probably the million dollar question (pun intended :) ) is just how much money has 1800 made.  Not a lot, but to be honest when I originally wrote this that wasn't the point.  I wasn't comfortable in my artistry or execution to ask for someone's hard earned cash but instead decided to subsidise with ads. A decision I wasn't entirely happy with as it goes against the minimalist design of the game.

What was interesting as 1800 was accessed by more and more people was just how many people wanted to pay for it.  A concept and request I thought I never thought I would see.  People even asked for my PayPal account to make a donation.

Obviously not comfortable with donations I decided the easiest thing to do was to add in an ad removal option that would provide some money and give the option for donations.  While the sentiment was great this hasn't been a money spinner but instead 6 generous souls have paid for this option.

So just how much has 1800 earned and by what means?

Flurry:        $12.93
iAd:            $28.77
AdMobs:    $14.47

Total:          $56.17

Not bad I guess but certainly not life changing and certainly not enough for me to consider giving up my full time job.

Beyond Financial Reward
1800 for me has been a success which is weird thing to say considering what I have just stated above concerning the financials. Making the game accessible has meant the audience has been wide and diverse and that has naturally led to larger downloads. Hitting upon the "just one more game" obsession has seen the large number of games played.

Purely by chance the design element was a fortunate fluke and had no idea at the time it was en vogue.  Maybe I am doing myself an injustice and that my artistic eye hit upon something that was right.

What I have learned from my 1,000+ hours developing games is the craft of writing games and not the monetary reward.  If you put your heart and soul into something then the rewards will come.

The exposure in TouchArcade certainly helped a lot.  It will be interesting to see from my next game just how much exposure and cross promotion from this base will influence the download figures but without such influential placement and articles I have to be realistic that achieving these levels of downloads could be challenging.

But despite all of this 1800 for me has gone beyond monetary reward.  I won't betray the confidence of the person who wrote to me but instead will summarise their story to make my point.

A player of 1800 contacted me about a week after launch.  They told me their story of how they were once policeman who was called out to an accident one night. While at the scene they themselves were hit by a car.  A significant period of time passed which was spent rebuilding their life as the injuries sustained meant they would never work as a policeman again. The injuries had meant that multi-tasking and concentration had become a struggle.   Most recently they had been taking exams to build steps for their future.

The email continued to say that having played 1800 over a period of a few days not only had it helped them improve their concentration and reactions but had helped them achieve higher marks in their exam.

To me that is priceless.

1800 is available for free on iOS for iPhone, iPad and iPod via the AppStore

Follow me at @gameproject10k for further updates and adventures.


Anonymous said...

1800 is a great game and it is just the start. I am sure you will make lots of great games. Keep going...

Breakdance McFunkypants said...

This is such a wonderful story. It was the first thing I read this morning and it will cast a positive light on the rest of my day. I'm so happy for you. You DESERVE this feeling of success, pride and satisfaction, and extrapolating from the amazing things you're learned so far on your 10k journey, imagine the amazing things to come. CONGRATS! I'm proud that you are part of #1GAM.
- Christer Kaitila

Post a comment