Articles

After spending some weeks on stage and gameplay testing, and briefly considering switching to squares, I had decided to stick with hexagons for movement. While we had enjoyed a good run of productivity on the art production side of things, now that we had a responsible and professional team that was good at communicating, spikes in Ovidiu’s and Matei’s day jobs were cutting into the time they had available for freelancing.

When looking back at the time we started with the second batch of models, I realized that about six weeks had passed and we weren’t done with them yet – while we were close, some details were still missing, and it would take a couple more weeks for them to be finished. This would end up bringing the total for them to about two months.

The quality, however, was up to the standard I expected from them. This was one of the two models:

I then took this character and sent it off to Jenna Bankhead for skinning and animation. It turns out that at the time she was planning a big move, so she contacted me and asked if there was any rush for the animations. I told her not to worry, and waited for her to get back to me.

By the time Jenna was done, however, we haven’t started on the new batch of models. Ovidiu and Matei had been required to go into overtime by their employer, which left them drained and not exactly in a very creative mood by the time they were done. Matei introduced me to another freelancer at the time, with the intention that even if any individual model was delayed, we could still continue moving forward in parallel.

This was a reasonable enough assumption: even if it takes us a month to complete a character, if everything else remained have more characters done in the same timespan.

It was time to take a step back and make a cold evaluation of the situation. The following were true statements:

• Models of the quality we were aiming for were taking between 4 and 8 weeks to produce with part-time freelancers.
• Every time we had to evaluate someone’s portfolio – be it modelers, concept artists or animators – my own productivity grinded to a halt, as I had to put on my producer’s hat and leave gameplay programming aside to focus on evaluation and hiring.
• Hiring another modeler would likely allow us to generate more assets in the same timeframe, at an increased coordination cost
• Coordination costs were already high enough, as more and more of my time was spent managing the concept -> model -> animation and rigging pipeline.

At this point, it seemed like I had three major options to continue with the project if I wanted to hit the targets in any kind of reasonable timeframe:

1. Forget about doing the development myself, hire a programmer and focus on production. This would increase the project monthly outlay, increase the coordination time, and require me to spend some time finding a good Unity developer, but once we found a good one it would result in accelerated development.
2. Hire at least one of the freelancers full time. This would reduce the coordination costs, allow them to focus, and make the production times more predictable. I was partly uncomfortable, however, with asking someone with a full time job to leave it to join an unproven start up.
3. Look for a full time freelancer who had a window of time available, and get him to fit in as much work as reasonable on that window. While this would have a similar effect to #2, it would also increase the coordination time if I kept on the current modeling team. I was not comfortable with replacing them, as they had provided good quality work so far.

Ovidiu and Matei were looking for other opportunities beyond their employer, which also consumed some of their time, so I aimed for #2. Matei would end up moving to Spain, and while I made Ovidiu an offer, I got outbid by a local media company. Considering the work he mentioned they were going for, I thought this was a better opportunity for him, so I didn’t try to up the ante.

That left me with the option of looking for a new modeler, which would require me to go into production mode for a while longer, evaluating new people, aiming to weed out the inconsistent ones before we even began working, delaying any work I could send on Jenna’s way for rigging and animation and, in short, restarting the whole art dev cycle again.

I really was not looking forward to that.

It became apparent I had failed to heed Douglas Stanhope’s advice about practicing excess in moderation, and had not considering how much aiming for a quality increase in a particular area would have cascading effects through the rest of the pipeline. Instead of picking an area of the game and polishing that, I was trying to polish too many things at a time – stage design, character modeling, animation, scenario generation and the several other areas we have discussed.

This was definitely doable, but at too high a cost. It would increase the total development time, multiply the amount of money spent on contractors, require me too focus solely on production and, in the end, raise the risk level exponentially just to have an increase in art quality. While I loved the work I was getting from the art team, aiming for that art style was something I desired, not something I needed.

Lesson learned: pick one part of your game and go nuts – polish it into oblivion – but make sure it’s a fundamental part of your game.

And so it was that I decided to somewhat reluctantly take my own advice that the money in the table is no longer yours, and decided to pull the plug on the project.

Next week: all together now.