"Is csn alive" app

[mogexpres]

Edit: http://apps.decision-cream.com/isCSnAlive/

Hey guys I'm mogo. I throw piss and sometimes medic. I'm here to gauge interest in a web app I'm thinking of making. The community here is pretty strong but sometimes popping is hard. When people see that there are only 12 people in the server they assume that the server is dead and go to a jump map. However, if those 12 people are all regulars who just joined then the server is pretty alive despite the sheer player number. You may say that one can look at the server list from the website or on the "server is full" menu for this info. This makes you look up and down the list and try to manually aggregate the notoriety of the players in the server and the time that they have played. What constitutes "just joined the server" is different for each player, each day of the week, and the time of day. These factors makes it so loading up TF2 and joining the server is the best way to figure out if those 12 people are at the tail end of a good run and just throwing piss at each other or are going super tryhard for a nice long session. This is kind of a long process with loading steam, loading TF2 and getting into the server all for a chance of disappointment.

What I propose is an app that grabs some hlstatx data on the current state of the server along with some player tendencies. I would turn these into hopefully 2 stats, Tryhard index and predicted lifetime of the server. Tryhard index would be a number between 0 and 1 aggregating the +/- of the hlstatx points. If everyone is gaining points its okay to assume people are trying to get objectives, kill medics, etc. If they are in the negatives they are probably throwing piss and trying to poop on people. The lifetime of the server will predicted using the time people have been on the server and a prediction of how long they will play. This is done on a per player basis and then added together with an option for hlstatx points being weights. The thinking is that the play time of regulars means more for the server health than a new user. Both of these are just preliminary ideas open to discussion and debate since they will probably change drastically as I develop it. Also I haven't looked at any of the data yet so these assumptions are just expert opinion.

Csn offers a variety of different experiences depending on the people playing and how they are playing. Sometimes I just want to throw piss and sometimes I want to win stuff which can be checked from the tryhard index. However, I also want the server to not be dead which can be checked by looking at the expected lifetime. This app will try to give players an indication of what csn is currently offering in these two dimensions. Hopefully this increase the health of the server while making popping attempts and revival after a server crash more successful.

Edit: I was thinking of writing it in R as a shiny app, free and open source.

[L0wsound]

That sounds pretty cool. Although I don't really see it gaining too much popularity..

Is this a school project? or a for fun thing?

We could most likely have it advertised on the server as well, and it should definitely have a giant banner promoting the popping or SB2

[Rumors]

I think the idea is pretty neat, but its also super-specific in what it does. So there's a chance it may hardly get used. If this app had other features that would complement this one I could see it being viable. With advertising on CSn and spreading the word, anything is possible. ^_^

[Someone Special]

Seems like a great idea, albeit hard to pull off.

[flufffage]

So tl:dr correct me if I'm wrong, but what im seeing is it would show the try hard index for each server, and projected pop lifespan, correct?

Would it also show which players are on? Which map is currently on?
Could it show pop announcements as push notifications?

[megamoose]

If the app could somehow get to the forums, this could be a very personal CSn APP. It'd be great, nifty, and we could possibly get more players on board with us then...

[mogexpres]

I was just going to do it for fun and cause I would find it useful for myself since I can plan out when I'm gonna play from my phone and whatnot.

I will have a full server list on there with their individual tryhard indexes and expected times cause obviously the weighting will most likely be player specific (some people may skew the stats more than others due to personal preference i.e. Like playing with some people more than others). Also showing what team everyone is on is also pretty easy. Could then add other hlstat data as new columns, maybe adding those server awards so people know about it. I can even get server actions if need be so I'm all ears if someone has something else they want to add.

Push notifications is a possibility but depending on the implementation it may require a third party service like ifttt or pushbullet. I could simply do a email notification but that probably isn't too useful for people. Steam notifications could be a thing but would only hit the people already logged on to steam where a personal touch is probably more effective.

Maps are another breed entirely, yes they can and should appear but I'm not sure what to do with that data. Some maps kill the server coughegyptcough and should probably be reflected in the estimates.

I could also do a time slider so you can ask the question "How alive is Csn in the last week" etc so people can see the activity of the server without having to delve into the hlstat pages. I can also see it as a plugin with much more careful verbiage. We could brand it with csn stuff and advertise the program to other hlstatx servers.

[flufffage]

I actually read it this time

I think it's a great idea. I don't know jack about making apps so it seems like a daunting undertaking to me, but if you are willing and able to do it, I'd definetly be willing to test and use it.

Besides, there's no rush, plus there could be new versions and updates as different functionalities when they're available.

I say go for it; I'll be excited to check it out!

Ps pls make sure android version and not just iOS pls and thank

[Someone Special]

I actually read it this time

I think it's a great idea. I don't know jack about making apps so it seems like a daunting undertaking to me, but if you are willing and able to do it, I'd definetly be willing to test and use it.

Besides, there's no rush, plus there could be new versions and updates as different functionalities when they're available.

I say go for it; I'll be excited to check it out!

Ps pls make sure android version and not just iOS pls and thank

[megamoose]

I actually read it this time

I think it's a great idea. I don't know jack about making apps so it seems like a daunting undertaking to me, but if you are willing and able to do it, I'd definetly be willing to test and use it.

Besides, there's no rush, plus there could be new versions and updates as different functionalities when they're available.

I say go for it; I'll be excited to check it out!

Ps pls make sure android version and not just iOS pls and thank

iOS needs like 100$-200$ for it to be on the app store, Android is free, Windows is nonexistent

[Balubish]

Who uses Apple products?

[Rumors]

Who uses Apple products?

I know right?




















Sent from my iPhone

[mogexpres]

Whoa guys, at most it will be a web app cause I don't care for app coding for so many hardware considerations.

So the only requirement is having a web browser on your device which I *think* is universal.

It has a Javascript frontend so if any problems happen on a specific device it probably can be solved without too much trouble.

I got a dumb phone and no tablet so I got no horse in this race

[flufffage]

Whoa guys, at most it will be a web app cause I don't care for app coding for so many hardware considerations.

So the only requirement is having a web browser on your device which I *think* is universal.

It has a Javascript frontend so if any problems happen on a specific device it probably can be solved without too much trouble.

I got a dumb phone and no tablet so I got no horse in this race

oh wow ok mb sorry i read "app" and i think mobile lol
rest in pretzel never forgetzel

[L0wsound]

Shoulda called it a Web app.
Where ya hosting it?

I could help test on a bunch of devices and browsers.

[mogexpres]

Yo this is a thing now
http://apps.decision-cream.com/isCSnAlive/

This is in conjunction with a hlstatsX interaction package I got that can be seen here https://github.com/sidjai/rhlstats. I got busy for a while and got kinda bored so I did this over the last week or so.

The final definition I landed on for tryhard is the amount of skill change in a session that is above your personal average. This kind of normalizes the weird parts of HlstatsX on an individual basis. Lifetime is the hours that a player is predicted to play further. From the glances I had when developing, the lifetime seems pretty good based on my own experiences. Like I saw Lia have a connection time of like 2 hours and it predicted like 3 more which seems about right

The stack is a r shiny application running this package which is running on a rstudio server on an amazon ec2 machine which is redirecting to this subdomain of my blog which is in utter disrepair.

TODO:
-Layout / style stuff is all default -- could make it look better (probably looks best on a phone now)

-Statistical model is a simple moving average which can easily be snazzed up with first priority going to adding seasonal variation for the day of the week.
-Prediction is only for the next session not taking into account the number of days between each session

-The archive is only the top 200 people which gives a lot of people the default values of 1 tryhard and 45 mins of lifetime (horizontal line on the graph) which makes the lifetime extremely conservative
Solutions:
--Expand the archive to like 500 which would most likely slow down the connection
--Get the top 200 in terms of connection time instead of Skill which may better represent the community
--make a new model just for new people based on what they already did on the server

-Fun server is not on the server list cause they don't use HlstatsX ;( Probably nothing I could do about that

tldr; App is out and is mostly working in terms of aesthetics, performance and the stats. Could expand in all three areas depending on need / requests.

[Lia_Rein]

' Like I saw Lia have a connection time of like 2 hours and it predicted like 3 more which seems about right'

Oh geez, this says a lot about my life :/ lol

Cool idea though, I'd check it out when I'm in a country that allows me to have wireless internet in my house lol

Also this "--Get the top 200 in terms of connection time instead of Skill which may better represent the community" seems like a good solution to me

[HausOfJoel]