|
|
On August 14 2016 22:43 LML wrote:
less stress on the server and testing with hand picked individuals (who all entered information about their OS/GPU). The advantage is that you get less problems reported at once and can focus on these, also reduces the chance of weird edge cases.
I read that the server averages around 20-40 users. That must be far away from anything near "stress" (I was lead engineer on a performance improvement project spanning several years). I can only advice doing everything to find bugs as early as possible.
Regarding focus due to less problems reported at once:
All you need is solid priorities. One way to assess severity could be the number of reported duplicates for a specific bug. Reports should never go directly to the developer. Instead use a bug tracking tool and make the beta-users do some work for you. No bug is accepted without steps to reproduce.
|
Glad to see SB is making so much progress since I first signed up (still no invite aw well haha). Much much better than Birdie or whatever his name was that gave nothing after promising a server with similar traits.
I said it a few months back but I really recommend you guys look into some DDoS protection if you plan to support Koreans. Fish has a history of DDoS attacks towards rival servers big as Brain and small as a UMS clan server.
|
pretty sure 2pacs response did answer the question, in that you don't want the masses to have access to an unfinished product. high expectations without the time to deliver would just lead to nobody coming back to the product and killing it before it could get off the ground.
|
On August 15 2016 07:10 brian wrote:
pretty sure 2pacs response did answer the question, in that you don't want the masses to have access to an unfinished product. high expectations without the time to deliver would just lead to nobody coming back to the product and killing it before it could get off the ground.
yes brian, from a product management point of view you are absolutely right. It's a thin line between getting users involved early and pissing them off because it is not finished.
|
Will this fix this issue that I have? It wont let me play games with other people online because it said my connection is either very poor or something about the UDP packets.
I can play other online games just fine.
|
I think I will play Broodwar again with this.
|
Norway28866 Posts
On August 16 2016 02:22 Bareleon wrote:
Will this fix this issue that I have? It wont let me play games with other people online because it said my connection is either very poor or something about the UDP packets.
I can play other online games just fine.
While I can't make any guarantees specifically for you, fixing this issue (variants of this issue are very common) is one of the top priorities of shieldbattery, and it seems like virtually everyone on the server has been able to create games and play games with multiple people without any issues.
|
waiting for my invite, get ready boys
|
|
|
Just want to say a big THANKS for doing all this work and helping the BW scene! It absolutely blows my mind that BW still has such passionate fans after all these years. Keep it up!
|
How big is the list of sign-ups for this beta? I've been through every wave without an invitation so far.
|
On August 17 2016 15:35 AsianEcksDragon wrote:
How big is the list of sign-ups for this beta? I've been through every wave without an invitation so far.
Without having any actual clue:
Assuming every player plays for about 1 hour a day on average and there are 20-40 players on the server on average (as mentioned somewhere in this thread), a very rough estimate would be 500 - 1000 invites given so far through all waves.
Now you can do some trickery *) based on the fact that you haven't been picked yet :p
If the list of sign-ups consisted of 500-1000 players, you would have been picked (because everyone would have been picked).
If the list of sign-ups consisted of 1000-2000 players, you would have had a 50% chance of being picked **).
550 - 1100 sign-ups give 90% chance to be picked,
5000 - 10000 sign-ups give 10% chance to be picked
We can therefore say with a certain confidence that the number of sign-ups is somewhere between 550 and 10000 
My guess because of ***) is 500 +/- 30%
Variance is probably much higher because the considered population of not-picked players consists of only you. That's statistically speaking a very small sample size :p
* Similar mechanisms are used in biology to estimate population sizes of animals :D
** Note that I completely ignore factors relevant for selection, such as submitted hardware spec, or close relationship with the developers, etc.
*** Note that the assumption of 1 hour a day and the correctness of "20-40 players on average" significantly influence the result. I honestly challenge the correctness because the perceived average is very biased (the player only sees other players when he is online himself, which is more likely to be at peak hours)
And yes, I know this is probably not the answer you were expecting. :/
|
Hyrule19228 Posts
that math is horribly off. Plenty of people idle when they aren't actually playing. Plus I'm sure there are people who got invites who either never accepted or don't actually play.
|
nice!! hope to get an invite soon™
|
Norway28866 Posts
did some very cursory counting, list of current playerbase is about 350. I have no idea how many people have asked for invite without getting one.
|
On August 18 2016 00:09 tofucake wrote:
that math is horribly off. Plenty of people idle when they aren't actually playing. Plus I'm sure there are people who got invites who either never accepted or don't actually play.
I should've applied and then streamed the lobby for hours just to make people rage.
|
On August 18 2016 00:09 tofucake wrote:
that math is horribly off. Plenty of people idle when they aren't actually playing. Plus I'm sure there are people who got invites who either never accepted or don't actually play.
[Edited out stupid flawed bragging comment, thanks Kau!]
Of course the math is horribly off. In such cases you can only hope that errors average themselves out. Quite often they actually do. Or you can attempt to mitigate systematic bias with intuition, as I did in the above example.
I remember a case where a project manager and I independently made an initial effort estimation and we both arrived at exactly 167 days. We laughed about it, but obviously no sane person would give such a precise estimate without a confidence interval.
|
Kau
Canada3502 Posts
On August 18 2016 02:04 imp42 wrote:Show nested quote +On August 18 2016 00:09 tofucake wrote:
that math is horribly off. Plenty of people idle when they aren't actually playing. Plus I'm sure there are people who got invites who either never accepted or don't actually play. Yet I nailed it with my guess (500 - 30% = 350) :p Of course the math is horribly off. In such cases you can only hope that errors average themselves out. Quite often they actually do. Or you can attempt to mitigate systematic bias with intuition, as I did in the above example. I remember a case where a project manager and I made an initial effort estimation and we both arrived at exactly 167 days. We laughed about it, but obviously no sane person would give such a precise estimate without a confidence interval.
I thought your guess was for number of signups, not the number of people already in.
|
On August 18 2016 02:19 Kau wrote:
I thought your guess was for number of signups, not the number of people already in.
Oooops! I somehow completely misread "playerbase". Embarrassing... edited & thanks
At least now, knowing that there are 350 players rather than an estimated 500-1000, we can update our intervals:
388 people signed up: 90% chance of being picked
700 people signed up: 50% chance of being picked
3500 people signed up: 10% chance of being picked
new guessing attempt: 1500 +/- 40%
(still not accounting for picked players not playing, etc.)
|
+ Show Spoiler +Wow +/- 40% nice estimate. Surprises me you didn't get reminiscent of another one of your crazy adventures as a project manager where there was some obstacle that you overcame and you got an award for it LOL. Please don't clog up this thread with this pointless math. There's an entire section dedicated for this type of posts, the blog section.
|
|
|
|
|
|
EPT
