Every one talks about QoS in the corporate world nowadays, with the SOA backbones and ESB products in a rampant evolution and implantation now it is imperative to provide  specific parameters like MTTF or MTTR.

Enter League Of Legends, the client-server real time tactic videogame that allows you to team up with the less cultivated 12 years old the Internet can provide. LOL got famous around 2011 but is has been in 2012 and 2013 that their reach on the Internet sky rocketed and so the amount of users.

In 2012 the League Of legends phenomena was too much for the Twitch streaming video service, Riot had to rely on the awesome Youtube live streaming service to reach the masses during the 2012-2013 third season. They dodged a bullet there.

There have always been problems with the League of Legends service: snowball matches, occasional ping, cursing, queue dodge but it was not only until October 2013 that the service really got worse with almost continuous server service interruptions in the European Union West Servers (EUW). Everything happened several days in a row: service not available or rampant ping or no chat or amazing levels of trolling or AFK (probably derived of the bad playing experience) and the list goes on.

The fan base knew things where really going bad when the twitter @lolstatus provided no information at all about the reasons of the problem, like when you try to evade confrontation only because you know there is no solution. Maybe that is Riot strategy, losing clients and leaving the available servers for the loyals that remain. Probably the worst handled SaaS status twitter account I have seen in my life, with only one tweet from years ago (they delete the new ones when the issues are solved, talking about sweep under the carpet).

Whatever the future brings, Riot gave this month an example on how not to handle your QoS, even if you have not a signed QoS agreement you have an implied QoS that your customers will accept and behind that level, only discontent may arise.