Valve Engineers Weed Out 'Lying' TF2 Game Servers 97
billlava writes "Tired of Team Fortress 2 servers that lie in order to attract players, engineers at Valve (creators of the Half Life franchise) have come up with a way to weed out servers that give false information about the number of players online, or custom server options. 'After kicking around some proposals, we came up with a simple system built around the theory that player time on a server is a useful metric for how happy the player is with that server. It's game rules agnostic, and we can measure it on our steam backend entirely from steam client data, so servers can't interfere with it. We already had this data for all the TF2 servers in the world, allowing us to try several different scoring formulas out before settling on this simple one that successfully identified good & bad servers.' Of course, this only works with their games running on Steam."
A real annoyance: (Score:5, Interesting)
I don't actually have a problem with reserved slots themselves. They have no value to me, and I've never been kicked out because someone else who had a reserved slot joined. My problem is that the server browser doesn't understand reserved slots, and shows servers whose "public" slots are full as having free space to join. Steam has a rather good server browser that refreshes quickly and has other nice features, but joining servers only to go through the entire motions of connecting and _then_ being told that I can't play because there isn't actually a slot for me is annoying. It also breaks the functionality that auto-retires full servers, which is a very nice feature.
Maybe it's a pet peeve of mine. On one other site I go to once in a while, pretty much everybody complained about getting into this kind of server once in a while--one of the ones that reports being full but is actually empty. I haven't once ever had that particular problem. It's a strange situation.
Re:False Positives (Score:2, Interesting)
Popular servers tend to have about 30 people online at once. At a turnover rate of two people per minute, you break even. I'd expect it to be less than two people per minute though (probably more on the order of one person per minute or less).
As long as the server isn't below that negative threshold, it won't ever get de-listed. If it does drop at times, it should have enough points "built up" to not get de-listed for some time anyway. Also, presumably, it would be bast on averages from lengthy time periods (weeks/months), not just on a hourly basis or something.
Re:Abuse (Score:3, Interesting)
Outside of a couple instances of 4chan-ing it, there's absolutely nothing to gain from doing this. The TF2 community is already highly compartmentalized with only maybe 100 truly active servers and/or communities. The rest are either empty most of the time or empty most of the time and employing these awful tactics described above. This fix just makes it easier for people new to the game to find a non-shit server that's actually populated. Plus they already have a several week baseline that they've been monitoring.
Re:"simple solution" (Score:2, Interesting)
While it is simple, it also has problems. What if I connected to a server and a minute later my PC crashed? Or there was a power outage and I turned off the game so my UPS would last longer? Or I thought I had time to play the game but it turned out I really didn't? Or ...
Those kind of effects would spread quite evenly on all the servers.
Even if some servers would have more restless players on average, isn't that exactly what the system is for, to warn players about a bad playing experience.
Re:Great! (Score:3, Interesting)
L4D server searches don't seem country limited to me at all. I play with 3 other friends all in the US and we get connected to German servers all the time, or some other server with a 150 ms ping. A manual server selection feature for the lobby leader would be a big step in the right direction.
Another big desire for us is some way to merge lobbies for versus mode, so I can create a lobby of 4 people, and then connect us to another lobby of 4 people so we can play some decent players, not just the 4 random yahoos that clicked auto join.
Re:So why would one "cheat" in the first place? (Score:3, Interesting)
If this is the problem (I think it's more complicated than this one factor), it could be solved another way. A "Quick-play" button, steam takes an empty server and puts the next 24 players into it. Steam could regularly cycle through servers, or just pick empty ones at random.
Anyhow, I think the root problem is too many people have servers up. It's so easy, valve is basically a victim of their own success. Now they have a bajimmillion(10^13.5) empty servers for players to filter through.