Hi Snarkie, thanks for your work on Neopets! I’ve been a part of the site since 2000 and it’s crazy to think how much it’s impacted my life as I’ve basically grown up with it.
The Altador Cup is by far my favorite event, it keeps me and many others coming back to the site. I love the competition and as a very non-athletic person in real life it makes me truly feel like I can personally make a difference (the AC counts as e-sports, right? :P)
I had a few questions regarding some comments you made in your Reddit AMA:
1. You said that “at one point, it was really hard to keep track of winners because we had to do a lot of stuff manually to tally, so we moved to a different algorithm that did a lot of that automatically”. When you say some things were tallied manually, does that mean that someone actually had to stay up until midnight NST to calculate results? I’m curious anything you’re allowed to reveal about the previous scoring system.
2. You said that “early on in the first year, we were finding that freeloaders were having a small impact even though we didn’t want them to, so we changed the algorithm on the fly”. How are “freeloaders” defined in this context – users who weren’t playing _at all_, or users who were _playing some, but not very much_?
3. Related to this – there was a good bit of controversy generated by an Editorial (http://www.neopets.com/ntimes/index.phtml?section=editorial&issue=735) which claimed that sending minimum scores could possibly hurt your team. We were always told by TNT in the past that “you can’t hurt your team, only help it”. Can you speak to whether someone could actually _hurt_ their team’s performance, or is it simply a matter of “helping less” vs “helping more”? It’s obvious that minimum scores are “less helpful” than better scores, but this issue has created a lot of arguments and conflict.
4. I saw in the AMA you are still friends with Breik, who was always Meridell’s representative. Did he ever find out about my Meridell fight song and national anthem? (http://www.neopets.com/~Mega2000)
Yo, Michaelruss. Thanks for the numbered questions! My brain finds that very soothing. ?
1) Not for daily game winners, no. At least not in the way you’re thinking. Some stuff was done the next morning. But for overall tourney standings we were doing some very weird calculations manually, yeah. (It was “easy” to tell if Team A had beaten Team B that day but who was currently in 1st, 2nd, and 3rd place overall was pretty complex.) We were also running GIANT queries on all the LIVE tourney databases periodically and at the very end to do certain calculations on all the data. That was the biggest no-no. For a lot of reasons. The main one being our database servers crying for a swift death. Well, and also the inevitable times when a 3-hour query dies in the middle and you have to start all over again boo hiss. So we modified the scoring system to allow us to track extra data and do calculations live as games were played. We also began running lighter calculation queries nightly on crons (automated scripts that run at certain times and do tasks for you) to store other data and make the bigger queries more manageable. We still verified all that info manually (spot checks on the daily stuff, and definitely lots of verification for the final winner calculations) but this way made it both much easier to calculate and more reliable.
2) Freeloaders were people who joined a team but weren’t playing at all. Anyone who made any effort wasn’t considered a freeloader for purposes of the algorithm and was helping the team. Also keep in mind that we changed things up after the first year because people whose favourite teams were knocked out early didn’t bother to pick another team and just stopped participating. So the algorithm got a LOT more complex once we had every team in it the whole time. There were a lot more things to consider in that case.
3) I’d been gone for a couple years by issue 735 so I can’t really provide insight into what they mean here. The algorithm we used in my day meant that any scores submitted helped your team. Higher/more numerous scores helped more, sure, but the system was specifically designed to make sure that every score submitted only contributed to the team. This was to ensure even low-skill/low-time players could participate. Parts of the algorithm did have to change every year or things would get weird so maybe it changed in a different way that year or they changed it completely by then? Or maybe they just phrased their answer poorly. We had YEARS of experience phrasing answers around this question very carefully to avoid just this kind of confusion… because we’d made that mistake before. Haha. I’d probably give whoever answered that question back then the benefit of the doubt.
(Quick semi-tangent: I think people were always assuming we did something really simple like divide scores submitted by overall team size or something, meaning that someone who submitted once and never played again was being counted as a full participant and ballooning team size, thus dragging the team down. Definitely not. ? There were MULTIPLE levels of participation that were compared and calculated in different ways. We ran scripts that simulated many whole seasons with all different kinds of player types and team sizes to test against. Saying it was a complex system is putting it mildly. Heh.)
4) Haha, I don’t know! I would imagine he’s seen it if it was made while he was still there? He loved Meridell and looking at all the fan-created stuff so I can’t imagine he’d miss it. I’ll pass it along anyway!
I hope that answered your questions enough (without giving too much away about the system, sorry). Thanks for asking!
Thanks a lot for the detailed response! It is very informative and reassuring to hear that everyone who plays always makes a positive impact for their team. We’ve had a few vigorous discussions about this issue over the years, so I think it’s really important for people to know they’re still helping even at small levels of activity. I totally understand you can’t give away the whole scoring system but there’s definitely lots of food for thought!
I’d love to know if Breik has a reaction to my Meridell page too!