RE Reporting crashes:
https://forums.elementalgame.com/415288
I believe one change in beta 2 is that crashes are uploaded to the server via the Windows error reporting thingy. So we should have a lot more to work with.
But when I hear someone having a lot of crashes, that definitely makes my antenna stick up since our testing lab is finding crashing to be exceedingly rare.
Why crash reports matter:
The reason why crashing is important is because it really is often system dependent. And I'd rather us nail that stuff BEFORE release so that we don't have people out there saying "Yea, those losers at Stardock tried to claim that their buggy crap game was stable and blamed all their problems on people's hardware."
As any game developer can tell you, the dirty secret of PC game development is yea, the OS + hardware + drivers are a big part of the story.
Hardware + OS + Drivers = Lots of variables to juggle
The most INFAMOUS and scarring example was a certain reviewer trashing WOM saying it was totally buggy because it kept crashing on him when it turned out that yea, on HIS particular nVidia driver running on Windows XP on that particular nVidia card it was very buggy (because nVidia hadn't been updating their Windows XP drivers in awhile for that card).
We can always work around any technical issue. That's why you can say "Well, I play <game X> all day and it works like a champ". They've already worked around said issue.
Heck, I remember The Political Machine and a very specific but common embedded Intel graphics adapter that crashed 100% of the time IF you had never updated the drivers because the driver returned that it supposed feature X but it didn't really, it just crashed when that card feature was used. We had to put in a string compare into the code to fix it (check for that video card and ignore what it claimed it did).
Multithreaded programming is what makes FE's magical performance + AI possible.
And with FE, it's even trickier than what we faced with WOM because FE is so much more multithreaded. How it behaves on a dual core system will be different than how it behaves on a system with 6 cores.
Let me show you what I mean:
That's how many threads are running in Fallen Enchantress at once. A missing semaphore here, a typo in a critical section there and you can have all kinds of stuff. In WOM, we had relatively few threads (and you could tell with the late game performance). Most games use very few threads. But we use them because we can do a lot of very cool strategic stuff.
This why there is no big "please wait" dialog in FE. You get an hour glass but you can still interact with everything (Even build stuff).
On Scoring the betas
If you think of it as a 5 star rating system (5 star = excellent....1 star = terrible) you might be surprised (actually, you guys are likely much more seasoned PC gamers so you probably wouldn't be surprised) how good a predictor for a metacritic score the surveys can be if you get enough voters. Figure +/- 5 points either way.
We did this system with our previous games with the exception of WOM (because we had hired reviewers to internally review the game and give us a scoring...we won't ever rely solely on that kind of thing again).
Metacritic, which a lot of people don't like but is a fact of life, determines the fate of a game. A game that gets less than an 80 is basically doomed. Literally the sales difference between a 79 and an 80 is gigantic. Seeing the sales numbers on Impulse for a few years really was an eye opener on why people care about metacritic.
Comparative Metatric
For those of you interested in what various games have done, here are some Metacritic scores:
- Heroes of Might and Magic VI (77)
- Crusader Kings II (84)
- Skyrim (94)
- Sword of the Stars II (44)
- Stronghold 3 (47)
- BF3 (89)
- War of Magic (53)
- Temple of Elemental Evil (71)
- Sins of a Solar Empire (87)
- Demigod (76)
- Galactic Civilizations I (83)
- Galactic Civilizations II: Twilight of the Arnor (92)
- Master of Orion 3 (64)
