Results 1 to 13 of 13

Thread: Server Stats: what, why and how to improve

  1. #1

    Default Server Stats: what, why and how to improve

    Across the different servers there have been a number of players who have taken the time to collect statistics to help give an overview of the round. These have ranged from hammers, to WW and artefact updates, to end-game statistics on attack and defence. being a bit of a stat-and-spreadsheet freak myself, I was wondering whether there is anything that can be improved both in terms of the type of data collected and of tools to simplify said collection.

    The reason for this thread is trying to gauge what we all think about the following two points:
    1. What type of statistics do we consider useful? Why?
    2. What are the main obstacles in collecting them and what could be done to help?

    Any idea is more than welcome, and I volunteer to try to help with anything within my abilities
    Luna (1/2 Ram Punch, UK1 2012/13)
    Who needs a meta when 27 players can win a round? Gogogogogogogogogo WOPPPaaaaa!

  2. #2

    Default

    Quote Originally Posted by kathy View Post
    Across the different servers there have been a number of players who have taken the time to collect statistics to help give an overview of the round. These have ranged from hammers, to WW and artefact updates, to end-game statistics on attack and defence. being a bit of a stat-and-spreadsheet freak myself, I was wondering whether there is anything that can be improved both in terms of the type of data collected and of tools to simplify said collection.

    The reason for this thread is trying to gauge what we all think about the following two points:
    1. What type of statistics do we consider useful? Why?
    2. What are the main obstacles in collecting them and what could be done to help?

    Any idea is more than welcome, and I volunteer to try to help with anything within my abilities
    A function that recorded end of server stats. So we werent reliant on players to record them would be ace. I think sams sql thing would address this if I understand correctly.
    A proud member of the CF leadership - Boots.

    "I don't have my head up my own arse, I actually am amazing" - Darkwing Duck.

  3. #3

    Default

    Yes; it would - but that file is under HQ control. Whereas there may be templates etc possible on a local level

  4. #4
    Cloud Strife

    Prolific MemberProlific MemberProlific MemberProlific MemberProlific Member septimus ii's Avatar
    Join Date
    Feb 2009
    Posts
    3,980

    Default

    It shouldn't take long to take screenshots of all the relevant screens, especially with something like Gyazo. It's just making sure someone is ready to do it, sorted before the server ends. I tried to do it with uk5, but I was on holiday, and with very sketchy internet access, and the PM crashed before I got it sent

  5. #5
    Paralysis by analysis Avi's Avatar
    Join Date
    Sep 2007
    Location
    Oxford, UK
    Posts
    1,807

    Default

    I've always found the end game statistics the most interesting, primarily because they were finite and more easy to judge and compare. Other statistical information is interesting but only "in the moment" of the game, eg who is holding which artefacts, which WW is climbing up or down. Most players aren't so interested in this information once a server is finished, they just want to see the winner.

    I'd say comparable statistics are useful; statistics which the game generates every round such as the population, attacking, defending, so you can rate your success. It gives you something to aim for which you can monitor throughout the round.

    Then a different type of statistic, such as the total number of players, % of tribes, length of server, they are less interesting to some but put each round in context so I think are important to record especially when every round has so many key differences, especially with the different Travian versions. eg. It makes an attack score of 163k on one server and 145k on another proportionate to know that one server had 3,500 players and the other had 1,700, or that one lasted 280 days and another lasted 350 days.

    I'd love to see more army statistics, but they rely on player contribution a lot, pasting of reports when the army is used, transparency in the numbers. I've done it a few times and it wasn't easy chasing everyone up for their numbers, and I've seen other people attempt it and also struggled to get everyone to come forward with all the details. Undoubtedly people are also disappointed that to make it usable data, there have to be boundaries and some army data complicates or exceeds those boundaries, eg. an unused army, or an army used but split in half.

    Main obstacle in collecting info/data on time, not a day too early, when the server ends and before it gets reset. Before this was less of an issue as we had better analysers which could collect the information for us, eg. Cry. But since the introduction of T4, analysers ceased to operate as intricately and less data is now available through them, eg. Getter. I know some players treat Getter as the bible now, but compared to what we were used to, Getter is very sparse. There have been individuals perfectly capable of creating tools to harvest the information but those tools were not deemed okay by HQ.

    Collecting data as a player shouldn't be difficult, I think it just needs more people keeping an eye on every round, as the reason for some stats falling through the net was due to assumption that "other people would do it", or relying on one person to do it, and that person losing the data or - like septimus' experience - being unable to collect it. Perhaps a data collection team?

  6. #6

    Default

    Quote Originally Posted by septimus ii View Post
    It shouldn't take long to take screenshots of all the relevant screens, especially with something like Gyazo. It's just making sure someone is ready to do it, sorted before the server ends.
    Theres actually a fair bit of time involved and I think that could be a big factor in data not always being collected.

    Firstly you need somebody willing to grab the data and ready to do so at the right time. Unless someone has announced on the forums that they will collect it there is no guarantee that anybody will, the vast majority of players expect someone else to do it and then complain if nobody does.
    After that the information needs putting into some form of spreadsheet or template (unless you're going to be cruel and make Avi work it out directly from your screenshots).
    Then you actually need to hand the information over and wait on Avi having time to process and upload it so you can finally check that it all matches up to the initial information you collected.

    I collected the Uk2 data on 02/09/13. After checking it in the Hall of Fame tonight I've found quite a few mistakes (some of the players listed weren't even on the server). That means tomorrow I need to go through, find the mistakes and pass them to someone who can edit the post, this is yet more time.

    I'm more than happy to collect the data on any servers I play but I don't have the time or energy to keep an account active on each game world purely to collect endgame data.
    When you have lost hope, you have lost everything. And when you think all is lost, when all is dire and bleak, there is always hope.


  7. #7

    Default

    I think the stats that particularly add value are those that help to tell a story. There is a lot of numbers about, but they can be tricky to interpret (with factors such as server length playing a big role), or where raw data needs some processing to be really interesting (something more creative that averages with hide 100 sins).

    End game
    The screenshots for basic end game data are relatively straightforward. Presenting/typing/sorting etc. is rather more time consuming.

    The top 100 armies are incredibly time consuming - inputting data into a well designed spreadsheet. For the 'perfect' data you need the village types as well and adding the alliances etc.

    Keeping an attacks list up to do throughout the server involves not-insigificant effort either (or more than around a dozen people in total would have done it ever).

    Ongoing
    And end-game is one of the easier, it's focussed on a small number of accounts, army information is often shared (because there is a long history of doing so in the UK) and is a snapshot at the end.

    Capturing the top tens on a weekly basis has proved beyond the ability of everyone to do manually. And cry's is much mourned - which is why I'd love to see more data available to download (which makes doing across the server much easier).

  8. #8

    Default

    Quote Originally Posted by Samantha78 View Post
    The top 100 armies are incredibly time consuming - inputting data into a well designed spreadsheet. For the 'perfect' data you need the village types as well and adding the alliances etc.
    Yeah, I hadn't seen Avi' before I did my troop capture. Didn't think to take down village type.

    Having never done the armies before, I didn't realise how time consuming it was. And because of the way I did it, it got more difficult as time went on, due to the amount of amendments to positions, etc.

    I found most leaders were complicit to giving me the info i needed, and without their assistance it would have been a lot harder.

    Now that I'm going through the previous stats and collating them, I hate pictures with a passion, as it means I have to hand type the data, which gets very tedious...
    A proud member of the CF leadership - Boots.

    "I don't have my head up my own arse, I actually am amazing" - Darkwing Duck.

  9. #9
    Paralysis by analysis Avi's Avatar
    Join Date
    Sep 2007
    Location
    Oxford, UK
    Posts
    1,807

    Default

    Quote Originally Posted by Bloated Fish View Post
    Now that I'm going through the previous stats and collating them, I hate pictures with a passion, as it means I have to hand type the data, which gets very tedious...
    Tell me about it; some of the earlier rounds for hall of fame were hand typed from blurry youtube videos! Not my favourite past time.

    Which previous stats are in image form? Perhaps I can help.

    I also got lucky just through having very helpful contacts in my friends and in the leaderships when recording army numbers from a round. But to embark on that on a server I wasn't involved in and didn't know anyone on would be near-impossible, I think.

  10. #10

    Default

    Quote Originally Posted by Avi View Post
    Tell me about it; some of the earlier rounds for hall of fame were hand typed from blurry youtube videos! Not my favourite past time.

    Which previous stats are in image form? Perhaps I can help.

    I also got lucky just through having very helpful contacts in my friends and in the leaderships when recording army numbers from a round. But to embark on that on a server I wasn't involved in and didn't know anyone on would be near-impossible, I think.
    s1 and s2 were very well trained in sharing army hits - certainly for three or so rounds of each I'm fairly confident that 95% of proper wwk attacks were posted. (Tho to a certain extent, once there is a 'top armies' it's self reinforcing as players are enthusiastic about sharing the next year to see where they rank lol)

  11. #11

    Default

    not only would it be nice to be able to compare UK records, etc a worldwide one would be even better

  12. #12

    Default

    Quote Originally Posted by Jurassic Park View Post
    not only would it be nice to be able to compare UK records, etc a worldwide one would be even better
    I'm not doing world wide stats...
    A proud member of the CF leadership - Boots.

    "I don't have my head up my own arse, I actually am amazing" - Darkwing Duck.

  13. #13

    Default

    It is a great idea, but we tried all this before, and you always have people supporting the idea, promising to to it, but nothing gets done.

    I previously tried to do something where we got, not just stats, but also allowed for a victory speech, or a bit of a background of the history of the server, but with dozens of people promising reports, nothing came of it. Some servers have had some AMAZING twists and turns, and it would make for great reading if they were listed somewhere.

    For me, its not just about stats, but also respect. Some of the worlds greatest alliances, and players quickly fade into oblivion, be great to have better stats to show the world, just how good they were -> for example, doesnt matter how good any player from here is, as I have never heard of you. But having a place that showed world stats, wow, that would give the top players the respect that they deserved.

    You can create a site yourself at like wordpress.com (and its free) and then you can add in members, and they can upload the images themselves. So if someone guarantees that they will do the images / stats, you just give them access and they do it.

    If others from other domains wanted to do it, then this is what I would do. Find 1 person for domain, and they are in charge of finding, and managing the people that promise to do this, and you then give them access to the site, and then these people post the reports in the relevant sections.
    If you use wordpress, there are article directory plugins, so easy to have a layout such as:

    US servers
    - us server 1
    - round 1
    - round 2
    - round 3
    - us server 2
    - round 1
    - round 2
    - round 3
    - us server 3
    - round 1
    - round 2
    - round 3
    UK servers

    - us server 1
    - round 1
    - round 2
    - round 3
    - us server 2
    - round 1
    - round 2
    - round 3
    - us server 3
    - round 1
    - round 2
    - round 3

    well, you get the point.

    But doing this all comes down to people doing what they promise to do... which is not going to happen.
    One thing I quickly discovered is the instant the project started, peoples attitudes changed from "yeah happy to help" to "why should they work for free".... like I am serious, you will have people demanding you pay them as they think you are earning from it.

    which brings to another point. Who is going to see the site? This is alot of work if only you and a couple of others will see it. If it is only listed on a forum, then you are looking at perhaps 1000 people will see the thread in a year, 10% might follow through and go to your site, so expect only 100 genuine visitors..... (per year)
    I know your thinking everyone can find it via google, but it doesnt quite work like that. to elaborate, If you compare my site (travtool.com) to lets say travianhint.com, you will see my alexa ranking is insanely good, (alexa is not accurate, but gives an idea of how much traffic a site has, go to alexa.com and compare both sites and you will see how much traffic i get compared to his, you can also compare this forum as well), even though travian hint has more content, looks nicer, better tools and guides, as well as has been online for longer, the fact is that his site doesnt rank in the Search engines as well as mine, is that I threw $xxxx on SEO in order so that others could find my site.

    There something like 10 different travian sites that have gone under in the last 1-2 years, and I know for a fact that some of them tried to ask travian games to host it, so that the site could continue to exist, and well, we all know how that went!

    Most likely your project would quickly become another casualty, and unless travian changes it attitudes towards these projects, I guarantee we (as a community) will continue to lose more sites and tools. And the more sites and tools we lose, the more the travian community shrinks.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •