And I'm pretty sure there are many more careless mistakes yet to be discovered. Please fix the bugs before you guys sell anymore DLCs!
Issue added to the internal database.
Well, not be a meanie or to go off-topic much, but I would like to suggest the idea of creating a checklist to follow and maybe an automated debugger. That'd be a piece of inhouse software, basically a rough self-made parser to run a file through and e.g. see if it discovers the existence of any variables/parameters that are not on the acceptable list, then highlight them for the person who is running the check to fix (if it's something obvious enough that a QA'er can fixi it, e.g. 'cavarly' for 'cavalry') or get a designer/coder to investigate).
Processes and tools really can help avoid human mistakes (I'm a business translator by trade, I would know, we're a pretty hi-tech/process-intense job line these days unlike people who translated novels and poetry decades ago). Running a checklist can feel like the most soul-destroying experience ever (I HATE having to do checklists myself and always try to weasel out of them), but an hour invested now can save you ten hours in the future. Or something like that.
It certainly helps keep the product polished and thus the professional image intact at a relatively little cost — because it removes the tiny little issues that clients get so worked up about. Like, imagine you translated a professional article or a pretty novella, some 20 pages of excellent writing pushing your brainpower to its limits with all the juggling of languages and all the different fields of knowledge or areas of business. But your client can't see that (at best he feels you're just doing your job, maybe a little better than the other guys) and goes ballistic about the three typos he found. That's surely unfair from the translator's point of view, but then, running the spell-checker (which takes only a minue or so for that size of text) would in most cases have prevented the problem (and half an hour to read the whole thing again 'manually' and discover those typos that a spellchecker won't find because they are words that do exist, only not those words you need).
Besides, it's always easier to first invent the rules and procedures for dealing with a problem and then just execute them, than to reinvent procedures for every job over and over again. Saves you some time and energy and ensures consistency. This is another advantage of processes and checklists.
One other good way of avoiding certain tiny issues that attract disproportionate attention is the 'two pairs of eyes' principle. Basically a doer and a reviewer, two different people, not one. Sometimes human brains get fixated on some weird idea — for example they lock on a wrong spelling — and due to that cognitive distortion (sometimes a very brief one, you know, like a slip of memory) they are unable to identify mistakes in their own work, especially if they are double-checking it just after finishing it. A different person may still have his own momentary slips or tiny gaps of memory or something else like that, but usually a different combination of them than the first person. Between two people, it's always easier to fish out typos and stuff like that.
Besides, such mutual proofreading also sometimes prompts a discussion of the content, not just the form. For example a different coder/designer checking your stuff for you as a mutual favour could help you realize that maybe the numbers are a little off or perhaps something doesn't add up, or maybe there is a better solution to the problem which the checked code is addressing.
So it's generally always good to have some defined processes and checklists and help each other out with mutual proofreading at work.
