"
"
...
When something like this seems completely obvious, but hasn't changed, I find it very useful to ask myself why that could be - is there something I am missing in this picture that explains what is going on?
To help you out, in this case the answer to that is simple: "most players" never experience disconnections, so they can play through the lab in a single run without even the faintest hint of concern about being disconnected.
Given that, there is no compelling technical reason for GGG to add checkpoints because of disconnects: it doesn't help most players at all, because it solves a problem they simply don't have.
Also, it doesn't really help players who suffer them with anything other than the outright disconnection, since they can still die to traps, enemies, etc, if the root cause of the disconnects - network trouble, in about 7 in 10 cases - just causes them to lag out at the wrong moment, and stand in a trap, or a smash attack, or whatever.
So, there you go. You asked why this has not changed, and that is why. Enjoy.
Is the statement "Most players never experience disconnections" a hypothesis or a fact? What's the source if you're asserting that's a fact?
It sits somewhere in between the two: I have reasonable confidence in this assessment, based on experience with similar systems, and my experience working with people in the forums to identify the root cause of latency and disconnection issues. (That is, the 7/10 I noted above.)
Combine that with the player count numbers from Steam (which exclude, of course, stand-alone installer users) and it seems ... shall we say, highly unlikely that (a) the number of complaints is vastly smaller than the number of players, and (b) so many of these are diagnosed to non-GGG issues.
Not none, of course: sometimes there is no evidence to support a network issues, and sometimes it is in the NetworkLayer or SoftLayer data center / connections. I'm comfortable saying most though. (That also matches the general state of connectivity on the Internet, FWIW.)
"
An alternative hypothesis could be that disconnects and game crashes affect most or even virtually all players (or a sizable enough minority for GGG to think about), but that GGG figures players won't get overly discouraged or frustrated if the occasional run ends with the game crashing or a disconnect.
That is also a possibility. It doesn't seem to explain the preponderance of network issues identified in the tech support forums when people actually do run the tests, though?
It should be quite simple to eliminate my hypothesis, though: when you ran WinMTR or equivalent tests, did they show any network issues? (Pointers welcome to the forum thread, if you used one?)
