WWAUT - Digging Too Deep

[Harb]

I would have thought the real life Paris catacombs would have been ripe picking for a great location. Something that already exists, has a great theme of walls of skulls and long tunnels and ties in to the above ground Paris map

Anyway looking forward to using night vision with my Commando perk

 

[Lieutenant Skittles]

I figured there would eventually be a catacombs map, but I thought it might be in the Paris catacombs. Whatever, this is much cooler!

 

[Lieutenant Skittles]

I would have thought the real life Paris catacombs would have been ripe picking for a great location. Something that already exists, has a great theme of walls of skulls and long tunnels and ties in to the above ground Paris map

Anyway looking forward to using night vision with my Commando perk

Looks like you and I have the same thoughts.

 

[Aleflippy]

Wow! Now that's not the kind of map we see often! It looks amazing and it looks both scary and stylish. I can't wait to blast some zeds!

 

[Knowsley]

Lets get ready for 20 more days of its only going to take 20 more days.

 

[NonCom]

Can you explain this?

Yes. People were added to the project.
Bugs were found that increased work on some people.

 

[NonCom]

Someone on reddit pointed out that it looks like Tripwire used the data from two weeks ago as the data for this graph's Last Week, so I made an updated little version from eyeballing it that is probably closer to what this one should look like.


Keep in mind I didn't have the real data so that might be wrong in places, and I could've mixed a few bars up.

The person on Reddit was not correct.
It looks like you know your way around Excel graphing, so have a look at this:

[/IMG]

Notice 3 columns of numbers - 2 weeks ago, 1 week ago and this week.
See how the "select data" dialog in Excel shows that the graph references columns, D and E - last week and this week.
The bars for anim1 and anim2 demonstrate this as well. Anim1 cleared 4.5 days worth of tasks in 5 work days. Anim2 has newly added tuning tasks.

Again, remember that these values change every day, as we load balance and add new bugs.

 

[Ryno5660]

New map looks great! Claustrophobic natural tunnels, potentially pitch black...Perfect hunting grounds for a fire breathing monster!

As for development time, you do what you do! New things, bugs, things that cannot be predicted, all a part of it no? I have no doubt you're doing your best, so thank you!

I am excited for this update!

EDIT: I must say, the way the paragraph above the 2nd picture is worded, makes me think there's a boss down there that isn't hans.

"It". Giant crawler Broodmother or something? Not gonna try and start a hype train about it, just my thoughts!

 

[sathenzar]

I can't wait for this update. It looks freaking sick!

 

[jorgamun]

The person on Reddit was not correct.

I don't mean to sound combative, but they are correct.

Here is an image I threw together illustrating what I'm talking about:

What happened is that the "Last Week" data on 7/16/2015 is being used as the same data for this week's graph for the "Last Week" column. So effectively the "Last Week" column for 7/24/2015 is two weeks ago.

The image that I uploaded uses the "This Week" column from 7/16/2015 and puts that as the "Last Week" column for today, 7/24/2015.

 

[s1nk0]

Looks AMAZING! Would love to see more dark, creepy maps in KF2.

 

[NonCom]

I don't mean to sound combative, but ...

What happened is that the "Last Week" data on 7/16/2015 is being used as the same data for this week's graph for the "Last Week" column. So effectively the "Last Week" column for 7/24/2015 is two weeks ago.

Not really, but I see why you would think so.

For instance, the anim1 numbers do appear static for 2 weeks, because 1. (I repeat myself) these numbers change every day and 2. most graphs are made on Thursday night and do not contain any changes that happen on Friday. This week, I made the graph on Friday.

So you see things like:

Anim4 shows 17 days worth on the (Thursday) graph of 2 weeks ago, but by Friday the number had grown to 32!
By the following Thursday, that was knocked back down to where he started at 17 and by today, it was down to 8.

Art1 went 27d Thr / 30d Fri to 17d Thr / 26.5 Fri to 15d on the next Fri (today).

This effect is known as the "beta bounce" (ask Stunt about it). You'll see the same devs oscillate between two fairly consistent hi/lo values of bugs as you approach a release. Gradually, the oscillation shrinks as the 'found versus fixed' rate starts to improve. It looks like a bouncing ball gradually losing energy.

As each new feature, gun or map is added to the project, bugs spike. Devs hold back on a set of lesser bugs to attack more serious issues, making their personal bug counts oscillate between "all I can manage" and "the few I'll get around to".

You can see it on individuals and even the whole project:

This is the found versus fixed rate for the whole shebang. Note the classic bounce effect - 53 bugs found a week ago, 55 found this week. 51 bugs resolved a week ago and 77 resolved this week plus a batch of audio tweaks (that is the shift we look for as things begin to stabilize).

You can also see the peak of changes at the end of a week, like the 16th through the 19th. A week later, we're passing the same bug counts again, just a little better off.

When green is winning, life is good. That means you are getting closer. Soon, despite best efforts, QA will just find fewer bugs and fixes will overtake the bug count.

Again, I see why you might think so, but that is just what it looks like. The bounce. I can't be any more transparent than showing you the actual settings in Excel.

 

[Durpy]

This map description makes it sound like as if there would be a new boss of some sort, or a new zed. Who knows?

 

[Ryno5660]

With the ability to go completely dark, I wonder how this map will fare come wave 10. And then when you fight the boss, potentially you're in total darkness. More ideas popping into my head!

 

[Durpy]

With the ability to go completely dark, I wonder how this map will fare come wave 10. And then when you fight the boss, potentially you're in total darkness. More ideas popping into my head!

Well if the boss remains to be Hanz, it will be like fighting the Glow In The Dark Nvidia Monster

 

[jorgamun]

Not really, but I see why you would think so.

I would think so because the data from the graph on 7/16/2015 for "Last Week" is exactly the same as the data from the graph on 7/24/2015 for "Last Week". It's not a matter of interpretation.

I see what you're saying about how you produce the graphs - however, that's tangential to the issue pointed out.

Is the "Last Week" data taking from the "This Week" of the last graph? If so, then your current graph is incorrect, because the data of this week's "Last Week" does not match the data of last week's "This Week".

If the data of the "Last Week" column isn't actually being taken from that section, then it's a misleading descriptor and should be modified.

However, regardless of intention - the data of this week's "Last Week" column is the exact same as last week's "Last Week" column. The data for last week's "This Week" column does not appear on this graph.

You are suggesting that the beta bounce is responsible for perfectly resetting every data point from last week's "This Week" column to that same exact value from last week's "Last Week" column. This would be true for every single category as far as I can tell. I highly doubt that every category would rest perfectly at the old data points like this.

Also I'm seriously semantically satiated with the word "week" now. What does it even mean anymore?

 

[Lorkenpeist]

Many, many words

I must say NonCom, I really appreciate your detailed explanations. You really know your ****, and thanks to you taking the time to explain things to us, we get to know a little of your **** too

 

[xbalday95]

I would think so because the data from the graph on 7/16/2015 for "Last Week" is exactly the same as the data from the graph on 7/24/2015 for "Last Week". It's not a matter of interpretation.

I see what you're saying about how you produce the graphs - however, that's tangential to the issue pointed out.

Is the "Last Week" data taking from the "This Week" of the last graph? If so, then your current graph is incorrect, because the data of this week's "Last Week" does not match the data of last week's "This Week".

If the data of the "Last Week" column isn't actually being taken from that section, then it's a misleading descriptor and should be modified.

However, regardless of intention - the data of this week's "Last Week" column is the exact same as last week's "Last Week" column. The data for last week's "This Week" column does not appear on this graph.

You are suggesting that the beta bounce is responsible for perfectly resetting every data point from last week's "This Week" column to that same exact value from last week's "Last Week" column. This would be true for every single category as far as I can tell. I highly doubt that every category would rest perfectly at the old data points like this.

Also I'm seriously semantically satiated with the word "week" now. What does it even mean anymore?

you see, I think you missed the part where he says he makes the graphs Thursdays, and doesn't include Fridays. The "last week" bars for this weeks graph could be showing days left as of Friday. Or Saturday even if I'm not mistaken.

 

[jorgamun]

you see, I think you missed the part where he says he makes the graphs Thursdays, and doesn't include Fridays. The "last week" bars for this weeks graph could be showing days left as of Friday. Or Saturday even if I'm not mistaken.

Nah, I didn't miss that. My question is: why not use last week's "this week" data for this week's "last week" data? The data is there, regardless of when the graph is generated. You can go to last week's WWAUT and see that it doesn't match what is shown.

The fact that this was generated potentially yesterday should not affect it.

Let's assume that somehow this would affect it (I can't wrap my head around it, since there's literally a missing column of data points between the last two weeks, but I'll grant that I could be missing something). If this is the case, then the methodology is faulty and should be modified so that it works correctly.

 

[Ryno5660]

It may be the numbers of last Friday, not last Thursday. It's been said that it changes day-to-day. Why the numbers of that specific date weren't used may come down to how the graphs are made, and it gets technical/semantic, as you say.