Already done by springies, some 15 000 games of past games tracked so far. Used among other things for ELO rating calculation.
If you do it by gadget, you might run into security issues - or you will have to upload stats as many times as there are players and aggregate on server.
However, if you wish to track Evo in a more robust manner (More than just kills/deaths, etc.) and are capable of providing a server for the gadget to talk back to, this could be very useful for providing persistent stats and drawing certain types of players.
Joined: 17 Nov 2005, 02:43 Location: Raegquitting Spring on 04/24/12
Well that's the idea. I want to display a rankings page on the site Showing playername, wins, losses, and disconnections.
@licho I understand what you mean regarding security issues (as the script would ahve to know the name and pw of the db, but wouldn't a simple enough solution be to just compile the end lua script?
Also, I'd put it in it's own db so that even if the username/pass were compromised, it wouldn't be any big deal. Plus, I can set it up so that the db gets emailed to me nightly so restoration wouldn't be any big deal.
^^ not desirable though
Surely there can be a fairly non-complicated solution?
Joined: 17 Nov 2005, 02:43 Location: Raegquitting Spring on 04/24/12
The thing is, autohosts are limited (especially springie - no offense intended licho). Map and mod options are extremely important in Evo and without them some things jsut don't work they way that they are supposed to.
Additionally, I don't want it to be locked down to just autohosts.
If it's used for ranking, why would you want to include each and every game someone every plays into it? That can really discourage experimenting with new strats, playing when intoxicated, playing vs more skilled opponent, etc. if this ranking is relevant in a way that ppl want to maintain a high rank. (Or they'll just smurf)
Sure. If you can decide which game counts you might maintain a higher rank. But that doesn't really represent your skill. E.g. if you just count easy games and always win you will end up with low skilled players in team games.
In long run some experimental games etc. don't really matter.
What you mean wins losses and disconnect.. !stats show that ... and in the long term it will converge to 50% wins (if you play team games on ELO balanced hosts).
What you mean wins losses and disconnect.. !stats show that ... and in the long term it will converge to 50% wins (if you play team games on ELO balanced hosts).
The issue is that Forb wants it for Evo games only. Springie shows your wins/losses for every game in the world. If you could change it to record a number per player per mod instead of all games, he'd have what he wants and the separate Elo rankings would be more precise. Different games are different, why pretend they aren't? If someone is truly a good player at every game he plays, his ranking will show up that way anyway. If someone is good at one and bad at another, he should show two separate rankings.
Uh oh, just change it on website? Its open source, you can modify it to count elo for each game independently. (Even retroactively). It just has list of games in database..
Uh oh, just change it on website? Its open source, you can modify it to count elo for each game independently. (Even retroactively). It just has list of games in database..
Okay, I'll do my best. But was hoping for Springie also to change and use separate rankings depending on game. I am willing to do that too, but I'd need help from you.
Car, there is no *ranking*, its all just view on data... Data itself contain all mods, what you do with that is your business. The output !stats show is just result of php page which runs few sql queries..
Users browsing this forum: No registered users and 2 guests
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot post attachments in this forum