The RPG Duelling League
RPG Debate => RPGDL Policy Discussion => Topic started by: Dark Holy Elf on January 12, 2010, 09:20:42 PM
-
Yet another discussion which spun off of what to do with the main tournament.
Anyway, rerankings. The current system is as follows:
-Take the champion and runner-up of a division, and pit them in a round robin with four first-round losers from the division above. Top four from that pool end up in the higher division, bottom two in the lower.
It's not a bad system; we used it for nearly 60 seasons after all. Its main advantage is that is generally very effective at ensuring only deserved upgrades/downgrades take place, since each dueller fights five matches.
The problem is that it creates 42 matches to vote on in a single week. This is, needless to say, daunting - by comparison, the "big weeks" in the main season have only 16 matches. Between that and the lack of writeups for the reranking matches (inevitably, given the workload that writeups would require), the week is a low interest one. A quick check of the numbers shows that Week 6 creates by far the smallest discussion threads and also has the lowest votedraw. This is bad; it's a dud week.
My proposal is that we switch those 42 matches to a mere 6. These 6 matches would have writeups, like normal matches, and would generate more discussion since there are fewer of them, and each match has high, easily understandable stakes.
Basically, at each division border, we take the finalist from the lower division and pit him or her against a first-round loser from the higher division. Win the match and you punch a ticket to the higher division. For the champ and runner-up it's "one last win and you upgrade", for the first-round loser it's "last chance to prove you belong". Easy to get psyched about.
The only question is how to pick the first-round losers to face a downgrade possibility. I've tossed around a few ideas, but my suggestion is going to be to pick the two first-round losers with the lowest numeric ranking. Typically these are the duellers who need to prove they belong, and it's a fairer method of choosing than my other two thoughts, which were either to pick the duellers who lost by the largest margins (unfair because of the "lol you drew Fou-Lu" factor) and the duellers who lost to the duellers who lost to the finalists (even though it's a variation on what we did in the old system, it feels a bit arbitrary and luck-based).
I imagine the champ would face the lowest-ranked and the runner-up would face the second lowest-ranked in the usual fashion of seeding. Again, only first-round losers are eligible (this last part would hold regardless of what system for choosing the downgraders we use).
So, in summary of the idea, here's what I believe the ups and downs are:
Advantages:
-Fewer matches with more tangible stakes will generate more interest, leading to...
-Actual discussion in the season topic for week 6 (or whatever it becomes)!
-Better voting numbers
Disadvantages:
-Less accurate method of division rebalance, since upgrades and downgrades are more likely to happen or not happen just based on spoiling luck
-It creates 6 matches' worth of additional writeup work for the season (mitigated some if we kill the team match, granted)
I feel the good outweighs the bad, here. Thoughts?
***
Postscript of related queries, which may be worth thinking about here, or in the season schedule topic, since they relate directly to reranking week workload:
1. Should we kill the team match?
2. Should we put a break week after reranking week? (might also be a good time for NR/bonuses... dunno)
-
First, for the sideshow questions. Will add that I -do- like the idea, and while it does generate more workload, it's pretty minor, especially if we do shoot the team match.
1. Should we kill the team match? Yes. Honestly, it's been pretty much pointless for years.
2. Should we put a break week after reranking week? Also yes. Dunno what would be put as filling there, though.
-
Break week, see my schedule realignment topic. Team match also needs to go.
Two lowest numerical rank is okay, but it does lead to weird situations where our champs are constantly in the pool unless they win. It does seem like the best idea, so I'm supporting this unless someone else has a brilliant plan.
Also it should have numerical impact. What else. There is a risk of strong fighters getting spoiled and downgraded with this sytem, but with the lowest numerical ranking part I don't think it is a serious risk in Godlike. Heavy/Middle maybe, but it will generate more interest at least.
-
What they said.
For the break week, there've been more than enough suggestions on Hal's topic, so it'd be easy enough to just pick one a few weeks in advance and throw it up during the bonus week.
-
I like the team match personally. And I think anything that keeps some variety of the types of matches we have on the site is good. It's only one match every 6 or so weeks, so I can't imagine the workload is that bad for it.
Also in favor of a break week, with the option for us to add bonuses or something to it.
I voted for the rerankings to change, but I'm honestly okay if they don't too. Both methods have some advantages, not the least of which is that the rankings style we have now is another way to keep some variety on the site.
-
I like the team match personally. And I think anything that keeps some variety of the types of matches we have on the site is good. It's only one match every 6 or so weeks, so I can't imagine the workload is that bad for it.
That is true, but it has been a match type that has been used for nearly sixty seasons as is, and, at this point, provides nothing of real interest by now. We're not really removing them because they are -bad-, but because they have, at this point, grown pointlessly stale, and there are a few ideas for replacing them as we speak, even. It's not something that is cumbersome or hard to do, but it's just gotten entirely too old.
-
1. Should we kill the team match?
I'd be in favor of it, I think. The team match is just too complex to really get good commentary/thought put into it, so it's been years since I've been interested in it myself.
2. Should we put a break week after reranking week? (might also be a good time for NR/bonuses... dunno)
as mentioned elsewhere, a pure break week is a poor idea. I'm kinda uncertain on an 8 week season as well, but whether we go that route or not, there should very much be something there.
Anyway, I had a thought that... in principle at least, strikes a balance between the strong suits of elfboy's idea and the old system. Thusly- take the two lowest ranked losers from the high division, and set three matches, one between themselves, and each against the lower division Champion.
This gives us 9 matches (a little more work than the current Semi-finals, with 6 matches having no lead time), which should be viable in terms of work and gets the added hype and consequences of Elfboys plan. At the same time, it gives a broader pool of the higher division, so it has a bigger sense of the lower division champion having to earn entry (especially since all three having even records is possible, in which case no upgrade happens). It does mean that the runner up gets nothing, but I don't think any system with fewer matches than the current one could really allow the runner up a chance to upgrade.
-
I like Cmdr's idea the best. I think that the one match to upgrade approach would probably favor certain types of duellers over others (status whores) and would result in a lot of undeserved upgrades. Cmdr's idea looks like a nice inbetween for the two. As for writeups, what if we had one big one for each division instead of individual ones for each dueller? I think it would still incorporate writeups into the rerank week, but would keep the amount to a managable length.
-
I think the current system works best for rerankings. For the one match idea, someone could spoil someone and end up in the wrong division.
and Kill the team match. I am never able to vote since usually there is one finalist in one of the divisions I cannot vote on. Even if you could vote on 7 out of 8 characters, you wouldnt be able to vote on the team match. Kill it.
-
My issue with Cmdr's idea is that it practically begs for a rock/paper/scissors situation in which case there is no particularly elegant tiebreak (total votes is workable as one, but I'm not really a fan of it). Also writeups for round robins get necessarily more awkward and I think it would generate less interest to see the same duellers repeated in multiple matches. All that and it's 50% more work, too.
-
The work to my mind is the bigger issue. Your system gives us 6 matches with a two week lead time, mind 3 matches with a 5 week lead time and 6 with a one day lead time. While, on the whole, writeups are the easiest brand of writing to work with, this does require a bit more effort. I think, ultimately, whether we want to do that depends on how much stuff we're putting in the rerank week; if they're basically by themselves, it's no worse than a current semifinals week, which isn't too stressful, but if we're adding it in with bonus material, NR, etc, it becomes a current Week 3, which is excessive and undesirable (afterall, we're talking about breaking it up right now >.>)
The tiebreak bit... well, it really depends on whether you think divisional borders should be hard or easy to cross. Like I said, I think that in an Rock/paper/scissors situation, you'd just have no upgrade happen, so I suspect that your system would most commonly lead to both lower division combatants upgrading, while mine would make it... if not uncommon, then certainly not a practical guarantee. Dunno which is really better overall.
I dunno what to make of the round robin issue. It is worth noting that it's still significantly scaled down from the current system, so from a match perspective it shouldn't be a problem (or, rather, should still be an improvement), but certainly writing for the same character multiple times might lead to problems. How much I think we'd really have to try it to see, though.
-
I like Cmdr's idea the best. I think that the one match to upgrade approach would probably favor certain types of duellers over others (status whores) and would result in a lot of undeserved upgrades. Cmdr's idea looks like a nice inbetween for the two. As for writeups, what if we had one big one for each division instead of individual ones for each dueller? I think it would still incorporate writeups into the rerank week, but would keep the amount to a managable length.
I'm liking this idea the best.
It gives us less (or the same) overall writing than NEB's idea, even. This seems pretty desirable.
It also makes Middle spoilers less likely to force low Heavies into the wrong division. And in the case of the rock-paper-scissors results, no upgrades or downgrades happen. It's intuitive and better than accidentally throwing a powerful Heavy or Middle into the lower division and then crushing it when they next get in. It's possible with this format that it can still happen, but it's far less likely than with the single-match system.
-
Duelists crushing a division happens. Sometimes it's a lucky draw, sometimes an undeserved downgrade, sometimes it's just that we found out something about someone that we didn't know before. Besides, someone has to win. I really don't see why that would be such a big deal.
I dunno, something about a round robin system just seems inelegant to me, even cut down to three people. Probably because it's just basically using the existing system with less people. I'd rather see those single matches, using numerical rankings should do a reasonable job of keeping undeserved downgrades to a minimum. I mean, it's not like we avoid them completely now, as shown by Worker 8 in Middle.
Also, after actually voting, it seems like keeping the status quo seems to be pretty close to the top but there's little to no justification for those votes in the topic. If you have a reason to keep it the way it is, speak up about it!
-
Tiebreaking to "no upgrade" works, but... basically, the only difference between Cmdr's and mine then pretty much becomes that upgraders need to beat two duellers to upgrade instead of one (and the runner-up isn't even given a chance). I don't view this as an advantage unless you're incredibly enamoured of making division borders hard to cross, which if anything seems the undesirable side of the coin to me.
And Djinn, it doesn't lead to fewer writeups, it leads to more, unless you're willing to make a big writeup for the round robin. I don't think that'll work at all.
-
Well, I think that if you have the champ winning once, but one of the upper division fighters losing both, that'd also be a divisional shift. But yeah, it still makes it harder to cross the border. I dunno, I do feel that divisional shifts are things that should happen, but I can't help thinking that only having a single match makes it too easy for spoilers to undeservingly upgrade. For example, I don't think things like Naoto in Godlike are really desirable, and that they would be routine in a one-match system.
On the other hand, I think the 7 week system is currently winning, so 9 matches combined with other week 7 stuff make it one of the busiest weeks, rather than an off week. So my idea may not be viable for that reason alone.
-
And Djinn, it doesn't lead to fewer writeups, it leads to more, unless you're willing to make a big writeup for the round robin.
That -is- why I quoted dude's post... >.>;;
-
I'm for keeping the status quo for reranking if we're committing to the 7 week schedule and adding the new stuff to the rerank week. Aim is to reduce the workload overall. I'm not against the proposed changes, just being pragmatic here.
-
I dunno, something about a round robin system just seems inelegant to me, even cut down to three people. Probably because it's just basically using the existing system with less people. I'd rather see those single matches, using numerical rankings should do a reasonable job of keeping undeserved downgrades to a minimum. I mean, it's not like we avoid them completely now, as shown by Worker 8 in Middle.
In general I'd say the round robin system is the most elegant (I'm interpreting elegant to mean balanced) system that we really fairly divise for this. Single matches have for more potential flaws into someone running into really specific spoilers.
That said, my current thoughts.
--Institute using the bottom rankings either way. I guess I never thought of anything like this because I never really thought about rankings before, but his is much better than basing who goes to pools based on how the person who they lost to does.
--I don't mind single matches (I do prefer round robins for the sake of balance, but if single matches can do a lot better on keep interest up, it wouldn't be a bad tradeoff). However, the one thing I liked about week 6 was that I always knew that I could just basically completely avoid writeups. So I'm definetely wary of wanting to create more of them. This could vary depending on whether we think we are going to looking at a 7 week turnaround or an 8 week turnover, and what potential other extra things we might be wanting to stuff in (Like NR matches).
--My gut is that my preference is to cut the 6 person round robins to 4. That cuts it from like 42 to...15? matches, which is still a major decrease and makes voting far easier. While it doesn't have the simplicity that single matches does, it still keeps writeups out of the game, balances upgrade pools better, and the 3 person round robin feels notably worse for reasons already pointed out (Upgraders fighting for 1 upgrade spot, rock-paper-scissors situations). Are there any weird ties that could come up with this method?
-
--Institute using the bottom rankings either way. I guess I never thought of anything like this because I never really thought about rankings before, but his is much better than basing who goes to pools based on how the person who they lost to does.
One thing I like about the DL's current system is it does put in that chance for double downgrades and the like. We have plenty who have deserved two downgrades, and the system makes that easy to happen if you suck enough.
Cmdr's system has a lot more work, as he just said. Elfboy's has the advantage of still being fairly light on writeup work. We're killing the team match which helps some, and we also have a week to sort things out with pools matches.
-
Kill the Team Match Up cause yeah, its pointless.
I think the idea of just 2 matches for the rerank per divison, against the 2 lowest ranked losers of round 1 works.
And yeah, a week between reranks wouldn't hurt just cause it helps slow things down.
-
Could we implement a 'Godlike Champ gets accolades, but not a guaranteed spot in the next season' rule?
We could even give them the reward of being able to be nommed immediately following their victory. That way, if people -really- don't mind seeing this Godlike again, s/he'll have the chance to participate again.
But it won't be a forced 'Dammit, stop nomming Myria, she champs every season and NEVER GOES AWAY!'
-
Voted for Option 2, I like the two duellers fighting the people with the lowest ranks best. :)
Alanna: They'd have to get spoiled twice here (both in the season itself and in the downgrade pool and most people who are good enough to be considered a spoiling to downgrade won't be in the bottom two for rankings typically anyway.)
-
Man, we have five people voting to keep things the way they stand but only one has bothered to say anything about WHY. And no offense to Alanna but even her justification was rather short and doesn't do a lot to sway people to that side. Would someone please explain to me why you want to keep what is, to me at least and I doubt I'm alone, the most boring slog of voting in the DL? Even when I'm voting on every other week, I generally don't in week six. I'd occasionally toss a vote at the team match or something but 95% of the time I don't touch the rerank pools. They simply have too many matches and nothing really interesting about them.
Not to insult you guys, maybe you've got good reasons and enough people care that it's worth keeping it. I just don't see it and when I don't see anyone defending it at all I find it hard to believe there are good reasons to keep it.
-
--Institute using the bottom rankings either way. I guess I never thought of anything like this because I never really thought about rankings before, but his is much better than basing who goes to pools based on how the person who they lost to does.
One thing I like about the DL's current system is it does put in that chance for double downgrades and the like. We have plenty who have deserved two downgrades, and the system makes that easy to happen if you suck enough.
How many people actually deserve double downgrades now though? And given how random the pools tended to be (how many pools did Fujin avoid in the end?), even a double downgrade was far from guaranteed even if deserved.
-
Man, we have five people voting to keep things the way they stand but only one has bothered to say anything about WHY. And no offense to Alanna but even her justification was rather short and doesn't do a lot to sway people to that side. Would someone please explain to me why you want to keep what is, to me at least and I doubt I'm alone, the most boring slog of voting in the DL? Even when I'm voting on every other week, I generally don't in week six. I'd occasionally toss a vote at the team match or something but 95% of the time I don't touch the rerank pools. They simply have too many matches and nothing really interesting about them.
Not to insult you guys, maybe you've got good reasons and enough people care that it's worth keeping it. I just don't see it and when I don't see anyone defending it at all I find it hard to believe there are good reasons to keep it.
We're not speaking up because our point was already made. ...Just not in this topic. Soppy put it best, but I'll summarize:
Changing it means more writeups. We're trying to make changes to -reduce- work, not increase it. The reason we're changing things at all is writer/editor burn-out.
While I personally feel that the one-on-one match style would be more interesting than the current style, it's also adding to our workload, and it'll also coincide with our plans for an NR exhibition match week, which is far more interesting to me and I'd rather be putting my work into that if we're going to increase our workload any. NR matches are also optional, so they're only extra work if you -want- to work on them.
Changing the reranking system would become a mandatory set of writeups.
While I wouldn't mind -some- kind of change, like the suggestion to reduce the number of people in the pools to decrease the number of matches on reranking week total... I don't think it's good to start mandating more writeups.
-
Well, is our primary goal to make things more interesting or to reduce writeups? Interest waning is a big issue in Week 6; I know I struggle to vote in Week 6 because it is a long process and anything to make that more interesting is good. I'm not sure if more writeups are necessary; I think rerank week matches would do fine without writeups sine uh they've never had writeups. Their existence is unneeded.
I'll phrase it this way; as a casual voter (or someone who doesn't do writeups), I will vote in rerank week more readily if it is changed to a more palatable format. I will never vote in NR. Something that affects how interesting the DL is what I think should be focused on because NR just appeals to such a small number of people, since it is really hard for even avid RPG players to vote in. I have played a ton of RPGs and I find myself barely being able to vote on anything.
-
(To Djinn, not Ciato) The problem with that answer is that the only real complaints of burnout I've heard as far as site work is concerned is from the missing comments, because there were often something like 10-12 or even more of them to do and they got shoved onto one, maybe two people. Also, missing comments have more of a content problem than writeups do since they're more original stuff and writeups are more spinning a couple abilities to hype a character up, with a line or two about the character themselves. Now, if I'm wrong on the burnout front feel free to have the writers or editors speak up, because a lot of this has been hashed out in IRC where not everyone is around to see it so it's very possible I'm missing something.
I feel that the problem with rerank week as it stands now is that very few people actually care, and leaving the reranking pools the same as they are is not going to fix that problem. Now, it's possible that there's a way to make it more interesting without adding even a little work to the mix but nothing pops to mind right away for me. I guess a possible compromise would be to use CK's system but leave them writeupless and just leave the current voting style in there, so at least you cut down the ridiculous slog of matches into a reasonable number, but I don't know if that would save interest or not. It would at least be better than the status quo, though.
-
Well, is our primary goal to make things more interesting or to reduce writeups? Interest waning is a big issue in Week 6; I know I struggle to vote in Week 6 because it is a long process and anything to make that more interesting is good. I'm not sure if more writeups are necessary; I think rerank week matches would do fine without writeups sine uh they've never had writeups. Their existence is unneeded.
I'll phrase it this way; as a casual voter (or someone who doesn't do writeups), I will vote in rerank week more readily if it is changed to a more palatable format. I will never vote in NR. Something that affects how interesting the DL is what I think should be focused on because NR just appeals to such a small number of people, since it is really hard for even avid RPG players to vote in. I have played a ton of RPGs and I find myself barely being able to vote on anything.
I'm mostly quoting this because Ciato's take on it is more or less how I see it. I also don't see why the writeups for the redone reranking system are necessary either: they can be just as easily applied writeup-less, and I'm not sure if giving these flavor text would really raise interest significantly.
-
Just polling: Would people who don't vote now vote on the 6 matches, but not if the round robin was scaled back 2/3rds in size?
-
I don't think it'll change the apathy much. People don't really care about the round robin matches outside of the really hardcore voters, and that's bad. It is a struggle to get people to care enough about the round to vote on.
-
Well, if there's only 9 matches to vote on instead of, what is it currently, 56 matches or something ludicrous like that? That'll probably help out a bit right there. I still think putting in writeups would help but even cutting the number down that much would do something.
-
15 by my method (9 by CK's...which uh...I honestly like the 6 match setup is better in terms of fairness and balance, which is a bad thing).
Super, will the apathy change much either way? I'm not sure the cut from 15 to 6 is a major one for internal voters (Can't say for external much). I guess my gut is that I'm not overly fond of the 1 match gateway. Just seems like to could easily lead to weird results based on a worthy upgrader drawing someone who deserved a downgrade, but was a semi specific spoiler.
-
I vote on rerankings as they are now with enough interest, but I can safely say that I would pay more attention to each individual match if there were only 15 total instead of 56.
-
CK's version isn't a one match gateway, it's two matches for the champ plus one between the two people he's facing and no chance for the runner up to upgrade at all, and he has to have a better record than one of the two to upgrade. What's yours, the same basic thing but with the runner up included?
-
Yes, mine essentially is the same thing with the runner up thrown into the mix (Which makes upgrades less elusive for the champ and gives the runner up a shot at least!).
-
--Institute using the bottom rankings either way. I guess I never thought of anything like this because I never really thought about rankings before, but his is much better than basing who goes to pools based on how the person who they lost to does.
--I don't mind single matches (I do prefer round robins for the sake of balance, but if single matches can do a lot better on keep interest up, it wouldn't be a bad tradeoff). However, the one thing I liked about week 6 was that I always knew that I could just basically completely avoid writeups. So I'm definetely wary of wanting to create more of them. This could vary depending on whether we think we are going to looking at a 7 week turnaround or an 8 week turnover, and what potential other extra things we might be wanting to stuff in (Like NR matches).
--My gut is that my preference is to cut the 6 person round robins to 4. That cuts it from like 42 to...15? matches, which is still a major decrease and makes voting far easier. While it doesn't have the simplicity that single matches does, it still keeps writeups out of the game, balances upgrade pools better, and the 3 person round robin feels notably worse for reasons already pointed out (Upgraders fighting for 1 upgrade spot, rock-paper-scissors situations). Are there any weird ties that could come up with this method?
Dhyer in his 1st and 3rd points speak for me as well. But I'll actually go ahead and put a vote behind this stance. It is, indeed, a nice hybrid of Elfboy's suggestion and our current practice. (The other possible merge would kill the writers/editors.) I'll go further and say that having a boundary line of 0.500 is quite elegant, IMO. More = Higher Division, Less = Lower Division, Exactly 0.500 = Status Quo
For those of you who would care, I also had an alternate strategy that would also mitigate the draw problem in Elfboy's idea and keep the direct impact, but would require a rerank period of 2 weeks. (Which would mean no break week, especially if the main Season (i.e. sans Rerank Week) would be 6 weeks.)
Basically, I would have the lesser champ against either Last Season's Promoted Champ if (s)he lost in Week 1 (Yes, I have that hardcoded) or the 2nd Lowest upper in Match 1 and the Runner-Up vs the Lowest Upper in Match 2. These would be in the 1st week. In the 2nd week, Loser 1 faces off against Winner 2 in Match 3. Upper Division: Winners 1 & 3. Lower Division: Losers 2 & 3.
-
I don't know if this has been suggested, but I'm jumping into the conversation!
What if you let the winner from the the respected lower division face off against the opponent who lost to the eventual champion their first week of fighting?
Like if we take last season for example, Gades lost to Fou-Lu in week one. since he lost to the person who smashed everyone, he never really got a chance to shine that season. Let him face off against the winner of heavy.
Or something...
-
The problem with that stance is that the match has a stake if the Godlike loses. It would be akin to if the Vikings win the Super Bowl that Alabama would face off against the Browns for a Pro/SEC spot. Or Virginia Tech (lost to Alabama in Week 1) would face off against Boise State with the former's membership in the ACC on the line.
In this case, you would be punishing the first loser to the eventual champion by giving him/her the "do-or-downgrade" match.
I will say, however: If the main problem with all of these methods is the need to write writeups within 24 hours, why don't we just move Rerank Week for Season X to the Semifinals or Finals of Season X+1 if we adopt a method requiring writeups? (Again, this problem Dhyer's method does not have... which means that we should've considered for more options before launching a poll.)