But I just don't see how any of this could've gone any differently at the time.
With all due respect, the reason larger publishers have project managers and risk documents is to help deal with this stuff. I'm not in the gaming industry, but I was in the software industry. Every single time we did a project, someone wrote a risk document. The risk document would list all the risks we think the project could have, how likely they were, how severe they were, and how we could mitigate them. We didn't close a risk until it was no longer going to happen.
So when you say:
We absolutely COULD NOT have seen the problems that delayed the game at those points
I'm a little surprised.
Here's an example. In one project I worked on, there was a very minor known bug from an upstream provider which would not have blocked our ability to ship. The risk document had an entry like:
Risk: Upstream provider does not patch bug.
Severity: Low.
Probability: Medium.
Mitigation: Stay in frequent contact with upstream provider; Thomas will investigate workarounds; Wayne will check use cases to see which are impacted and write troubleshooting text as needed.
And then when we heard back from the upstream provider that they were going to fix it, we lowered the probability of that risk and added a new risk.
Risk: Upstream provider releases update that breaks some of our code.
Severity: Unknown.
Probability: Low.
Mitigation: Contact upstream provider to get preview release, run on testing server to ensure there are no regressions. ETA to make judgment call about which version we ship with: 6 weeks before launch.
We'd have these meetings at least once a month, and we were always closing and opening new risk tasks. And I have to tell you, there were a lot of risks we couldn't close and were ongoing bears... but we never had unexpected risks because we always took the time to expect them. Our relatively small projects generally had 20 or 30 risks identified.
So, here are some examples for SC:
Risk: Contracted localizer could become unavailable or fail to do job.
Severity: High
Probability: Medium (Notes: Contracted individual person with no backup strategy)
Mitigation: Contract multiple personnel to work on project; ensure backup personnel available to take over; structure regular milestones and progress meetings with minimum reporting guidelines--if a milestone is missed, initiate in-depth review.
Risk: Unexpected software issue impedes certification or release.
Severity: Medium-high.
Probability: Low-Medium.
Mitigation: Ensure SLA contract between original team and us guarantees support; hire additional personnel locally to debug, test, or develop changes as needed.
Risk: Project scope exceeds expectations.
Severity: Unknown
Probability: Unknown
Mitigation: Use regular meetings to review progress, design model to ballpark percentage completion and review against intended timelines, add personnel as needed.
Risk: Structural decline in market sector.
Severity: Medium-High.
Probability: Moderate, varies depending on release scheduling.
Mitigation: Ensure timely release; investigate opportunities for hedging risk by exploring alternate platforms; use a pre-sale campaign to ensure peak interest is capitalized on.
I can't pretend to know the intricacies of what's going on inside the project. Unlike some others in this thread I don't really have an interest in Trails or have all that many contacts in the industry. I also don't have an MBA or a PMP or whatever they want project managers to have today. But just on an intuitive level, "unexpected failures" almost always happened because people were winging it rather than taking the time to plan and plot out risks. And certainly the two things that have been made public on this project: the stuff with SpaceDrake and the game's overall scope being vaster than thought... both of these can absolutely, 100% be anticipated, mitigated, and responded to.
No need to hire me, Torrance is too long of a drive anyway. And I know that you guys are a pretty lean operation so it's not always feasible to hire someone full time. But something that might be valuable though would be sending at least one of your people to get project management training. Whether that's a PMP, or ITIL, or really any kind of process improvement stuff, the resources are out there, and they're not expensive compared to what they save you. Also, I do know the Southern California area well and I know there's a fucking ton of producers and former producers from in-industry that consult on project management gigs. Foresight can be 20/20 too.