Log in

View Full Version : Raid Score Redesign


Turiel
24-10-2009, 04:27 AM
Hey,

So as I'm sure most of you know, I'm working on SSE V3.

Part of this is a redesign of the raid system and raid scoring to take into account the different types (25n, 25h, 10n, 10h), achievements, etc.

I need some help working out a raid scoring system.

Here's what I want:

1) Award points based on a boss kill or achievement gain
2) Modify points based on difficulty (normal, heroic)
3) Modify points based on time-since-release (i.e. a guild that kills it on day 1 gets more than a guild that kills it on day 2).
4) Seperate 25 man and 10 man progression.
5) I've yet to figure out how we rank which achievements as more important than others.

So mathematically, how am I gonna work this out?

Do people have any other suggestions on what raid score should take into account?

Tarqy
24-10-2009, 02:59 PM
Are you intending to pull this data from the armory or are people going to be manually updating?

Ring0
24-10-2009, 03:35 PM
The way I see it raid progression should be scored on:

* Difficulty setting: Back in TBC some bosses were harder than others and it would've made sense to score them accordingly. In these times we live in all normal modes are accesible, to the point of being able to PuG them, and there's no reason to score the end boss of an instance as worth more than the rest unless Blizzard's design philosophy changes.

Hard mode kills would quite obviously be worth more points than a normal kill. At the moment in Colliseum there's no way of 'skipping' hard modes, as could be done in Ulduar. Icecrown however will go back to something similar to Ulduar's system with a mix of Colliseum.

Dungeon & Raid Difficult Display: A group’s dungeon or raid difficulty setting will now be displayed via an icon off of the mini-map.ICC's hard modes will be triggered, but via in-game UI (much like how you toggle between normal and heroic modes today) rather than triggered by game mechanics (like not killing the three drakes before you attack Sartharion).
No doubt some hard modes will also be harder than others, and while it makes no sense to give special scoring to normal bosses you could do so with hard modes if you wish.

It's very simple to rate progress according to this. Let's assume guild A kills Yogg-Saron, is awarded 1 point in raid progress.

Guild B then comes and drops Yogg with 2 keepers out. Each keeper that's left out adds one to the raid progress (regardless of which keeper is left out).

Guild A comes again next reset and completes Alone in the Darkness. They'll be at 5 points reflecting the harder difficulty. Now because killing Yogg with no keepers doesn't require that you beat any of the other hard modes it could mean that a guild that's defeated some of the easier hard modes would be ahead on rankings; this is all hypothetically speaking of course, no doubt any raid that can defeat Yogg+0 can do the other hard modes. In this case you'd make specific hard modes worth more than others to reflect the difficulty of the encounter, ensuring that guild A which has defeated Yogg+0 is at the top of the realm's progression.

So let's assume Guild B still has Yogg with 2 keepers, but also has defeated easy hard modes like XT, Leviathan and 2. With the above in mind let's say a 0 keepers kill is worth twice as much as a 1 keeper kill, which is about right in terms of encounter difficulty. Guild A will be ahead of guild B's progression no matter how many hard modes they defeat or even if they kill Algalon until they kill Yogg+0 because it is worth more points than all other hard modes combined, ensuring a correct progress assesment.

So in Icecrown there's 12 bosses, let's assume hard mode Arthas will be insanely hard and make it worth 11 points, +1 point for killing it normally. Any guild that can defeat Arthas hard mode will then be top of Shadowsong progression no matter if they've skipped the other hard modes. If hard mode Arthas turns out to be loleasy then you can lower the points awarded for killing it.

* Time taken to kill: This one is straightforward as well. You don't even need to award points for killing it sooner, simply ordering the progression table so that if Guild A and Guild B are tied on points the guild that had the kill registered first will be placed above the second one.

If you want to get really fancy can always copy WoWProgress' rating method. (http://www.wowprogress.com/post/31_Frequently_Asked_Questions) (http://www.wowprogress.com/post/31_Frequently_Asked_Questions)


* 10-man vs. 25-man: You're going to have to have two separate ranking tables here. Technically you could merge them together but then it creates a host of problems: what if they're a 10-man guild but PuGed a few spots and cleared the 25-man, does that count as a guild kill? Any 25-man guild will have no problems handling a 10-man version and so 10-man guilds get screwed over unless they expand to 25-man raiding or PuG the remaining spots, etc.

While I'm sure you could lawyer talk a way of rating it it complicates matters when having 10-man and 25-man progression separated does not.

As for achievements... I really don't know, I mean, other than hard modes, and perhaps meta achievements should silly achievements count for progression? Hard mode kills should be enough IMO, but there's a case for arguing that guild that can get Meta-achievement drake mounts for their raiders is technically superior to a guild that can't.

I don't know about other random ones, skill-check encounters are usually included in the meta, while stupid stuff or really hard ones are left out.

Should also give a point for getting immortal-type achievements I guess, but not Champion of Ulduar types.

Tarqy
24-10-2009, 06:02 PM
I personally would just keep it simple, since the actual number or points in itself is irrelevant and most people will just want a general overview of raid progress on the realm.

I think the most accurate way of ranking during progression would be to use a system where the points depreciate over time, so points are based on the date of a kill.

If you make the base points for a kill on the day of release (or date of first access if we have another coliseum system for heroic modes) 900 for example and then reduce this number by 3 points per day.

This would mean that the guild who have killed the most bosses in the shortest period of time would be ranked highest which is what we want to see.

I would also suggest that at the point when the instance is cleared by any guild (on heroic since normal is far too easy) that the date of the achievement is used as the basis for ranking and supersedes the points.

This would mean separate rankings for each new tier of content but it makes sense to do this anyway as new guilds are created and others disband.

oh and 10 and 25 should be separate for the 10 man only guilds but i would merge the normal/heroic points for them into one since a number of guilds will always clear normal on day 1 since its too easy.

Elexin
25-10-2009, 02:14 AM
I could be wrong, but I think "simple" and "depreciates over time" could conflict :P

Turiel
25-10-2009, 05:23 AM
I definitely want to do depreciation over time, because at the moment guilds end up hitting the same score when some are clearly ahead. Code wise its not that simple to check the date and sort by that in a tiebreak situation (it sounds simple, but it actually involves a huge amount of db queries and processing).

Tarqy - if 5 guilds clear ICC normal on the first day, then they'll be equal in ranking. I think this is acceptable enough. They will start differentiating very soon once they start the hardmodes.

Mirinette
25-10-2009, 08:03 AM
Why not just let sites out there that already do ranking do the ranking, and either just present their stuff via an iframe (ugly solution), or see if you can get them to implement and API to poll ranking data.

Whats the benefit of making your own ranking system? To further some peoples epeen even more, while diminshing the work of others? Any sensible person knows who the high ranked guilds are just by looking at the options already out there, so if it may be damaging to some guilds egos to look like they are as good as another guild out there, on a ranking, them knowing they are better, and the info, if read properly, telling people they are "better", should be enough.

Vegelus
25-10-2009, 08:19 AM
Becase, since vanilla wow, we always had one. Before all those sites even started working. The natural way of things would be still having one, as people use it.

Turiel
25-10-2009, 06:33 PM
Well there's a couple of reasons.

I asked in the original thread a few months ago whether we should use WoWProgress rankings or keep our own one. The general feedback was to have our own one.

The other scoring sites only rank the top X guilds, so a lot of the guilds on Shadowsong don't get counted.

We will also track 10mans (which some other sites don't do), and have a seperate progression ladder.

Mirinette
25-10-2009, 08:54 PM
Thank you Turiel.

I can only see one problem: Whatever weightings you choose will most likely not represent most of the realm, or a greater variety of opinions/input. My beef is that I am afraid it will reflect what the top guilds (I assume most of the regulars are from top guilds) view as hard/easy, and how it should be valued according to them. I feel it might genuinly devalue the hard work of some guilds, just because they evaluate fights differently. It might also make some guilds and/or groups skip or focus on fights in a disadvantagous way, on this realm, ultimately hurting our progress compared to other realms.

Just a few thoughts that have popped into my head, when it comes to having our own set of standards.

Tarqy
26-10-2009, 02:22 AM
If sorting by date is too much of an issue then certain encounters would possibly require multiple scores based on difficulty, freya +1 2 and 3 for example, but the resulting rankings should be accurate.

Turiel
26-10-2009, 05:08 AM
Thank you Turiel.

I can only see one problem: Whatever weightings you choose will most likely not represent most of the realm, or a greater variety of opinions/input. My beef is that I am afraid it will reflect what the top guilds (I assume most of the regulars are from top guilds) view as hard/easy, and how it should be valued according to them. I feel it might genuinly devalue the hard work of some guilds, just because they evaluate fights differently.


Our standard so far has been to assign the same amount of points to each boss, except for the end-boss which is double. So an easy boss is the same as a hard boss. What's going to change is two things:

1) Hard modes will give additional points.

It will basically count as a separate boss and give another set of X points.

2) Achievements will give additional points.

A lot of bosses have achievements that are for fun or indeed just for killing the boss. I imagine these will give half the amount of points as killing a boss. The achievements which are part of hard modes (usually part of a larger metaachievement) will give the same as killing a boss.


So yes, a guild that does no hardmodes or difficult achievements will of course be scored lower than a guild who does. I don't see how it can be any other way.


It might also make some guilds and/or groups skip or focus on fights in a disadvantagous way, on this realm, ultimately hurting our progress compared to other realms.


Not really sure how, could you give an example?

Turiel
26-10-2009, 05:26 AM
If sorting by date is too much of an issue then certain encounters would possibly require multiple scores based on difficulty, freya +1 2 and 3 for example, but the resulting rankings should be accurate.

There are 3 types of points awarded in this situation:

1) Boss kill (regardless of which Freya mode you do)

2) Hard mode boss kill. This would be the 'hardest' mode only: eg Freya+3. However then there are situations where there's things like Yogg+1 which is widely considered to be the hardmode and is part of the metaachievement, but Yogg+0 is the hardest mode. So I'm not sure, might have to take it on a case by case basis as to what 'hard mode' is.

3) Points for achievement. This is how the points will be differentiated between Freya 1,2,and 3 since there is a seperate achievement for each.

Mirinette
26-10-2009, 09:52 AM
Not really sure how, could you give an example?

Well, I had some thoughts along the lines of that. This is just a strictly hypothetical example. Take IC in ulduar. I find that the Steelbreaker hardmode is very simple (10 Player at least), lets say this "council of point deciding" finds the same thing, they give it a lesser value then other hardmodes. Guild X comes along, they read the values, or figure them out from previous rankings, and decides that because they are worth less they wont spend time learning it, and skip to something else. Without the proper research they are missing out on some valuable progress because a superficial system told them its worth less then doing something else.

What if some fights in ICC are the same way? Skippable, or doable in different ways, and whoever around here decides on the point values devalues something and alot of the followup guilds just skip it because its worth less in the eyes of some? It would result in a lesser degree of proper progress on the realm because we didnt use the same standards as other realms do. Believe me, because I have thought of this here, I am sure others have as well.

I just find having superficial scorings based on the perception of some to be bad. Thats why I dont really like wowprogress.com. Personally I prefer the guildox way of doing things, because its simple and it doesnt value things based on the input of some of the ranking systems users. I am also not a fan of Guild Bs efforts to be worth less because they werent as fast as Guild A, a timestamp will take care of that. Guild B will never be ahead of Guild A, but by comparing their ranking scores they can see that they have completed as much as Guild A, just not as fast.

Just trying to paint a picture here. It is ofcourse clouded by personal preference, but I do think more people then me have been thinking about it.

Faylin
26-10-2009, 09:55 AM
All of this sounds like much of a hassle to maintain. I personally hardly use the SSE list anymore, and just look on sites like Guildox or whatever, since tracking achievements is just so much easier.

Just skip it maybe?

Tarqy
27-10-2009, 02:17 PM
...

You're assuming that the points are going to vary between bosses and i don't think that was ever going to be the case?

I actually don't see the need to even display the number of points each guild has the only reason for them is as a means to calculate an overall ranking

Oh and Turiel can you approve some of the new members on the forum so they can sign for my TotC25 plz!

Senex
28-10-2009, 03:31 PM
Just use the Ensidia method:

- The guild that gets a server first Arthas (or Deathwing, or whatever the current patch's final boss is) normal mode kill become the Top Guild until said boss is killed in hard mode.
- The guild that gets a server first final boss hard mode kill becomes the new Top Guild until the next patch.
- Any kills or achievements made by guilds other than the Top Guild are completely irrelevant and don't count.