Page 2 of 2 FirstFirst 12
Results 21 to 36 of 36
  1. #21
    Scholar Of Adventure & Hero Missing_Minds's Avatar
    Join Date
    Jun 2006
    Location
    work....
    Posts
    30,172

    Default

    Quote Originally Posted by myliftkk_v2 View Post
    The tools are garbage. There's not useful towards helping you maximize your preferences at all. They are lowest common denominator tools.

    You could build a real recommendation engine that facilitated successful groupings way beyond the stone age LFM mechanism. This has been done over and over in other industry verticals.
    I await to see some real examples rather than speculation. Let alone these examples must also have built in heuristics to prevent gaming the engine to troll/exclude players.

    Lastly, can you do so on a shoestring budget? Lets face it, getting any improvements in Turbine Soon(tm) standard has to be both easy and cheep.

  2. #22
    The Hatchery Enoach's Avatar
    Join Date
    Nov 2006
    Posts
    4,595

    Default

    please forgive me as I read the opening post my mind went right to a "Match.COM" commercial.

    There are so many elements that make gaming fun for different people. In my opinion the MMO game industry did come up with a Social Solution to grouping - The Guild System. Generally a guild is a set of like minded people (I say generally as in all cases people have used it for other purposes)

    Now the Guild System does not solve the issue of Pick-up Groups but does put the power into the players hands to gather people around them with common interests, goals and styles.

    The problem with trying to create a system that "Matches" people up is that we are not matching 1 person to 1 person we are potentially matching 1 person to as many as 12. I have trouble getting my family of 5 to agree on simple things.

    However, it is the Randomness of the PuG group that also gives it flavor. The changes made to the LFM system that allow a person to put up an LFM and just take the first 5 to 11 that join has helped make this easier for that person that doesn't want to screen applicants and just wants people to run with. It does not guarantee in any way that these people will be compatible but it would be similar to having the boss go "And here is Bob to help you on your project".

    What needs to be done is encourage more people to stop being LFM watchers and become LFM listers.

  3. #23
    Community Member Teh_Troll's Avatar
    Join Date
    Apr 2013
    Location
    Under the bridge
    Posts
    5,874

    Default

    Quote Originally Posted by myliftkk_v2 View Post
    Who said anything automatic forcing anything? It's a preference/recommendation engine, there's a difference. A computer can do this 24/7/365 w/o taking a break, you have friends that can do that?
    Honestly I can form any group I want anytime I want for anything I want . . . but I know that's not normal. I'm also an east-coaster with a 9-5 job who plays during the most populated hours.

    Sometimes I'll just throw up and LFM and see what happens . . . and often nobody ever hits it. I put my "BYOH Elite-Streak Zerging in progress" criteria and still nobody will hit it. What I'm looking for, what my playing style is, etc . . . is plain as day to see. I'm not sure how any automated system would fill a group any better.

    One thing that could help is extend the levels for elite streaking to maybe 3 levels above the quest level (a 16 can run a 13 elite) and maybe merge servers so there's a bigger pool.

  4. #24

    Default

    Quote Originally Posted by TrinityTurtle View Post
    Examples? What more would you actually want to see on the tools...
    MyDDO, before it's extended beta period's untimely demise, had a list of a dozen or so preferences ranging from RP interest to playing hours. If something along those lines could be incorporated into the game UI and combined with character/item metrics, establishing a baseline for grouping suggestions would not be a huge obstacle. As for implementation ideas... how about a selectively visible scale indicator floating above characters while in a tavern indicating how closely they match your personal criteria? All one need do is walk into a tavern, take a look and start a chat with those who share your interests. It would act a social icebreaker for new players and the terminally shy.
    The newest computer can merely compound, at speed, the oldest problem in the relations between human beings, and in the end the communicator will be confronted with the old problem, of what to say and how to say it. - Edward R. Murrow (1964)

  5. #25
    Community Member
    Join Date
    May 2013
    Posts
    669

    Default

    Quote Originally Posted by TrinityTurtle View Post
    Examples? What more would you actually want to see on the tools, rather than just "They're stone age and can do more?" What has been done in the industry verticals in other places, and how would they address the problems here?
    Ok, from a generic problem statement, essentially a grouping can be said to be about: content, playstyle, capability, and compatibility. LFMs, in the rudimentary form they exist now, essentially deal only with content and players try to shoehorn the other three elements into the message to wildly differentiating degrees of failure.

    Let's say that each of the four items is weighted equally at 25% of a successful recommendation, just because it's simple math.

    What would I track:

    1. Content - Why not keep a master list of what content I'm interested in running (a preference for content) that is persistent across logins (and content I own)? This is just data.
    2. Playstyle - Why not set my persistent preferential playstyle both at my toon level, and at the instance group level? Zerg, Flower-Sniffer, All Opts, Fit to Group style.
    3. Capability - Why not allow me to self-declare my toons needs and capabilities (which change over the course of builds): Boy Scout (always prepared), needs Heals, throws Heals, does Traps, solid CC, Nuke 'Em All, etc.
    4. Compatibility - Why not allow me a tag my aquaintances after a quest (calling them friends is a misnomer, since often times you want to remember those you both like and dislike) with range of compatibility facts: Know-It-All, Friendly/Helpful, Needs a Leader, Solid in a Pinch, Not if they were the Last Toon Standing, etc. They don't necessarily have to be super-serious as they would be your private, not public, tags - though the engine could "share" them in a valuable fashion.

    Now, how might a recommendation system work? For the group leader, you select the content and the playstyle and capabilities and immediately the game could determine what other players share, to what degree, intersecting preferences with yours. Number one, this will tell you lighting quick the relative chance of success your grouping attempt has at actually being a group. If there are very few natural intersections (say only toons w/ 2 of 4, for 50% confidence), you're obviously going to be at risk for either few takers, or applicants who do not share a high intersection count and thus might degrade your experience. The other thing this does is it allows players who aren't part of your group, and possible not in quest ATM to be notified that a grouping matching their personal preferences is available and the confidence level of the grouping that they will have a good experience. It would also enlighten players as to how many potential teammates share their particular preferences at a given time (if none do, can you really complain about your group not filling).

    As a group leader, you could in fact see the applicants tags that you had applied to them before accepting them into your group (or a recommendation from the grouping tool), and one better, if the you don't have experience with the applicant, it could look at your highly rated acquaintences tags and see what their ratings for a particular toon might be and offer a positive, or neutral, or negative recommendation on acceptance.

    ---

    As for other industries, essentially Yelp works off the above, travel sites, retail recommendations, and any serious preference marketing engine.

    --

    I could sketch out the whole thing technically, but I think the above communicates it, and more and I'd have to send Turbine a bill.

  6. #26
    Community Member
    Join Date
    May 2013
    Posts
    669

    Default

    Quote Originally Posted by Missing_Minds View Post
    I await to see some real examples rather than speculation. Let alone these examples must also have built in heuristics to prevent gaming the engine to troll/exclude players.

    Lastly, can you do so on a shoestring budget? Lets face it, getting any improvements in Turbine Soon(tm) standard has to be both easy and cheep.
    I'll answer the last first and the first last.

    That's a political problem, not a technical one.

    With heuristics, your question really is one of can someone be prevented from misrepresenting their experience with another, because that's really the only intersection between players of substance. It's trickier in other industries, say restaurant reviews, because no one entity owns and operates the entire environment. Turbine, though, does own all the data being created and consumed, and remember there is independent data that could be at nomially odds with a player given rating upon examination. Thing like quest deaths, logged activites, success/fail status, contribution, are hard to fudge. Could individuals still attempt to troll someone, certainly. Is it easier to catch a troll in DDO than in Yelp, absolutely (I'm fairly familiar with Yelp trolls and approaches to combating them).

    Remember, a rating cuts both ways. It's not entirely about whether it's impossible to troll someone, it's often about what the risk is in doing so. Also, it's community ratings in aggregrate that make an average, and having a community of synchronized trolls is even easier to catch.

  7. #27
    Community Member
    Join Date
    May 2013
    Posts
    669

    Default

    Quote Originally Posted by Enoach View Post
    please forgive me as I read the opening post my mind went right to a "Match.COM" commercial.

    There are so many elements that make gaming fun for different people. In my opinion the MMO game industry did come up with a Social Solution to grouping - The Guild System. Generally a guild is a set of like minded people (I say generally as in all cases people have used it for other purposes)

    Now the Guild System does not solve the issue of Pick-up Groups but does put the power into the players hands to gather people around them with common interests, goals and styles.

    The problem with trying to create a system that "Matches" people up is that we are not matching 1 person to 1 person we are potentially matching 1 person to as many as 12. I have trouble getting my family of 5 to agree on simple things.

    However, it is the Randomness of the PuG group that also gives it flavor. The changes made to the LFM system that allow a person to put up an LFM and just take the first 5 to 11 that join has helped make this easier for that person that doesn't want to screen applicants and just wants people to run with. It does not guarantee in any way that these people will be compatible but it would be similar to having the boss go "And here is Bob to help you on your project".

    What needs to be done is encourage more people to stop being LFM watchers and become LFM listers.
    You be amazed how many people meet nowaways using exactly what you described. Of course, on the right sites, those matches more than 1-to-1.

    Again, a recommendation engine isn't about forcing anyone to do anything they don't want to do. It's about conveying opportunities and your relative chance at a good experience in choosing them. Now some people will throw caution to the wind and head on into the greasiest/dirtiest bar in the land and then rave about it afterwards, but most people don't work that way. They want some level of confidence the experience they're going to engage in is going to satisfy them before engagement.

    Sure, does it wrap the "magic" of random PUGs in some science and statistics, and does that make them slightly less romantic for a few? No doubt.

    But, most people engage their own personal recommendation engines which suffer both from li

    ETA: Just a small comment on Guilds. Guilds don't capture your immediate preferences any better than the current LFM tool. What Guilds are is a limited set of old preferences (which you cast off when you leave a Guild) and then the ability to achieve group rewards (in the form of buffs/etc). It's like taking the same team of people to a hardcourt everytime you go, versus going with less than a full team. It doesn't solve your grouping conundrum when you want/need to fill additional slots.
    Last edited by myliftkk_v2; 06-26-2013 at 12:43 PM.

  8. #28
    Community Member
    Join Date
    Dec 2009
    Posts
    2,523

    Default

    Here is my "blue sky" idea.

    One thing I like about NWN is the automated queue. Hit a button, go do your thing and the server will piece together a group. When it is ready, if you chose to join then you are teleported into the instance. I would love to see this be a single queue across all the servers. Maybe a check box to indicate if you desire a balanced group of a next 6 player group.

    There is no shortage of low level groups, there could be 5 different buckets for quest/raids:
    11-13
    14-16
    17-19
    20-22
    23-25
    (Someday... 26-28)

    Turbine could have different a rotating set of what is run like the special events. To address the question about revenue... it could be a VIP only feature with tokens for sale, but not a permanent pass.

    Of course... it isn't going to happen.

    While I'm listing things that won't happen... a cross server AH.
    Last edited by Ancient; 06-26-2013 at 01:03 PM.

  9. #29
    Community Member Phidius's Avatar
    Join Date
    Mar 2006
    Posts
    6,592

    Default

    Anyone remember when the level cap was 10? Back then, a difference of 4 levels was a HUGE gap, and therefore worthy of a significant reduction in experience gained.

    Now that the level is 25 (and soon to be 28), 4 levels isn't as big of a difference as

    1) TR-cached goodies
    2) High level guild benefits
    3) Memorization of a quest having played it a bajillion times already.

    And that's not even touching the elephant in the room (aka, Epic Destinies) where a level 21 could be grossly overpower a level 25. Not that I want them to factor in the total levels in Epic Destinies to the equation mind you...

    But I would like them to rework the "over-level penalties" and the "Bravery Bonus" systems so that we could have more than a 2-3 level gap in a "normal party".
    "I require a reminder as to why raining arcane destruction is not an appropriate response to all of life's indignities" - Vaarsuvius, OoTS #674

  10. #30
    Community Member AtomicMew's Avatar
    Join Date
    Oct 2010
    Posts
    2,705

    Default

    It's really not that complicated. Lots of other games have figured it out. It really all comes down to the fact that turine actively punishes people for grouping, and does little to nothing to encourage it.

    Low hanging fruit:

    -Eliminate player scaling - seriously, being punished by grouping with other players is one of the most stupid game designs for a MMO

    -Death penalty - should accrue a personal penalty only, and not to the whole party

    Medium hanging fruit:

    -Daily quests - one big problem is that there are just so many different quests, it's hard to get players to want to run the same content. Give a few quests in different level ranges a large one time XP bonus per day. It'd be nice if there'd be extra non-scaling mobs/challenges/surprises too.

    -Temporarily downlevel high level toons to the quest level. This might be a bit hard to balance, but it really just needs to be done. If you have a friend who is at a different level than you, you still should be able to play with him/her.

    High hanging fruit:

    -QUEST DESIGN! Give us quests that require teamwork. Most of the quests since U14 have been linear, go there and kill stuff type quests.

  11. #31
    Scholar Of Adventure & Hero Missing_Minds's Avatar
    Join Date
    Jun 2006
    Location
    work....
    Posts
    30,172

    Default

    Quote Originally Posted by myliftkk_v2 View Post
    That's a political problem, not a technical one.
    Actually it is very much based in technical issues and is very valid. The fact you just ignore it shows how much you really want to show understanding of reality.

    The second part, gets very "political" as you call it. I call it bureaucracy if not worried of corporate data mining and personal privacy issues. (the latter is easily handled with expansion of the current annon function.)
    Could they mine player data for habits? Certainly. (they'll probably need more coding work done so again let alone additional storage and sorting filters, technical again. While you can make up numbers of how much things cost out of thin air, I don't advise it.) The problem with using that sort of mining, opposing companies could use that to mine information about players without paying. Would this really be a big issue in the corporate world? Honestly, no clue. But I do know that player habits in so far as spending money is very much protected.

    Now could you do matching in such a manner that it really couldn't be mined? I'm sure there are methods. However, then you open yourself up to rabid complaining about "why didn't I match upwith such and such. Where was my notification, I"m suing you for date rape, etc." all sorts of issues both real and perceived. As a business how much is that really worth to you? In this game... heh.

  12. #32
    Scholar Of Adventure & Hero Missing_Minds's Avatar
    Join Date
    Jun 2006
    Location
    work....
    Posts
    30,172

    Default

    Quote Originally Posted by Phidius View Post
    But I would like them to rework the "over-level penalties" and the "Bravery Bonus" systems so that we could have more than a 2-3 level gap in a "normal party".
    If they would come up with a "mentoring" system such that highers could run with lowers without penalizing either... That would help out a LOT in my opinion.

  13. #33
    Community Member
    Join Date
    May 2013
    Posts
    669

    Default

    Quote Originally Posted by Missing_Minds View Post
    Actually it is very much based in technical issues and is very valid. The fact you just ignore it shows how much you really want to show understanding of reality.

    The second part, gets very "political" as you call it. I call it bureaucracy if not worried of corporate data mining and personal privacy issues. (the latter is easily handled with expansion of the current annon function.)
    Could they mine player data for habits? Certainly. (they'll probably need more coding work done so again let alone additional storage and sorting filters, technical again. While you can make up numbers of how much things cost out of thin air, I don't advise it.) The problem with using that sort of mining, opposing companies could use that to mine information about players without paying. Would this really be a big issue in the corporate world? Honestly, no clue. But I do know that player habits in so far as spending money is very much protected.

    Now could you do matching in such a manner that it really couldn't be mined? I'm sure there are methods. However, then you open yourself up to rabid complaining about "why didn't I match upwith such and such. Where was my notification, I"m suing you for date rape, etc." all sorts of issues both real and perceived. As a business how much is that really worth to you? In this game... heh.
    You're arguing that monetary reasons are the cause for keeping a poor IT design? That's laughable especially given the explosion in open access tools and solutions to anlayze and/or use. Budgets represent political choices, nothing more. As far as we know, Turbine still pays it's bills, minus web domains registrations. Turbine flushed a load down the commode foisting this amazing new forum on us, so trying to tell me that they are somehow judicious budgeters of cash and priorities would fly in the face of the litany of evidence to the contrary. You hire the right people, with the right skills, you'll get what you want and budget to spare, you don't, we'll... I live in the reality based IT world where people flush suitcases of cash down on a daily basis for awful stuff nominally called software. Acting like there's no cash to do what is right has been disproven failed IT company after failed IT company that was given money hand over fist to fail. Technical challenges aren't the issue here, as I've already pointed out, the problem model has been solved over and over. Further, the only people afraid to write/change/update code are people who shouldn't be writing/changing/updating code in the first place and if there's a legacy roadblock because the code is a mess of spaghetti, that's a technical debt discussion about skewed political priorities again. But, I'm habitually unsympathetic to whiny developers, and god knows there are lots of them, who eke out a paycheck existence by being just slightly more capable of developing software than a monkey on a keyboard.

    Companies already mine the living hell out of your data (and we can have a long and detailed discussion about whether it's really your data - which it won't be unless citizens get a law passed saying it is), but you should already know that. The best ones (Target) use it to personalize your ads in ways you don't even realize (see case of Target ad and pregnant teenager), and the worst blanket your web experience with ads for things you've already bought and login to (DDO). If DDO isn't mining their players habits, they're even more incompetent that the most negative posters sermonize. And how would a competitor mine DDOs data anyhow, by hacking their database? Nothing being suggested leaves DDOs universe, so there's no risk to Turbine except continuing to give players a grouping tool that doesn't meet their needs.

    The only technical problem that can't be solved is stubborness. But, the market solves that.
    Last edited by myliftkk_v2; 06-26-2013 at 02:37 PM.

  14. #34
    Scholar Of Adventure & Hero Missing_Minds's Avatar
    Join Date
    Jun 2006
    Location
    work....
    Posts
    30,172

    Default

    Quote Originally Posted by myliftkk_v2 View Post
    You're arguing that monetary reasons are the cause for keeping a poor IT design?
    Money makes the world go round (sadly). You are arguing for an idealistic world which isn't reality, let alone your statements showing you are letting opinions cloud facts.

    Quote Originally Posted by myliftkk_v2 View Post
    Companies already mine the living hell out of your data
    Yes they do. But companies, in the case of brand names per say, such as Target or Turbine, don't make that data easy for competitors to get. That was the point that went way over your head.

    Quote Originally Posted by myliftkk_v2 View Post
    The best ones (Target) use it to personalize your ads in ways you don't even realize (see case of Target ad and pregnant teenager)
    "I owe you an apology. There were things going on in this household I was not aware of. She's do in...." was the response given to the Target "help" employee by the father of said teenager. Or so it was written in the article I read about that.

  15. #35
    Community Member
    Join Date
    May 2013
    Posts
    292

    Default

    Wow, thanks everyone for the responses, I expected one or two and not a bunch.

    Adding new players to DDO is essential for continued operation. Many people try the game but often become frustrated at some point above L8 and leave before the end.

    It is my first time through DDO and my Ranger AA is now L10. Up until now I have soloed most content but have grouped twice. In both cases the other players went too fast for me to easily follow although the they thought they were going slow. In these cases the meaning of "slow" was very different from mine. Sadly, above L10 a rogue that can disable traps seems mandatory for many quests but I can't afford to keep buying rogue hirelings from the DDO store.

    Having guilds is not that helpful to newcomers. There are about 2400 guilds on Thelanis but no way I can see to find one where the members play at slow speeds. Even if every guild has posted such information there is still no practical way to read about 2400 of them. [ I do ask in-game for "/advice" about slow-play guilds and no one seems to know of any.]

    My fleshware is older so I can never go very fast and the truth is real-time combat is difficult. Nevertheless, by choosing passive feats and mapping the most important things to my keyboard, I have managed to get this far. Sadly, without grouping with a rogue it seems I can go no further.

    So how can group matching be improved so I may continue with my questing? {My wife does not understand the importance of this work but there are thousands of litte virtual people depending on us to save them. We must not fail.]

  16. #36
    Community Member
    Join Date
    May 2013
    Posts
    669

    Default

    Quote Originally Posted by Missing_Minds View Post
    Money makes the world go round (sadly). You are arguing for an idealistic world which isn't reality, let alone your statements showing you are letting opinions cloud facts.


    Yes they do. But companies, in the case of brand names per say, such as Target or Turbine, don't make that data easy for competitors to get. That was the point that went way over your head.


    "I owe you an apology. There were things going on in this household I was not aware of. She's do in...." was the response given to the Target "help" employee by the father of said teenager. Or so it was written in the article I read about that.
    I'm not arguing for any idealistic world. I'm arguing my position as an advocate for customer-serving software design, same as I do day in and day out to corporate budgeteers. I don't pre-capitulate because of budgetary reasons, and I've won way more than my share of software battles than I've lost, and many simply because I found something that will solve the issue for free. Anytime some claims someone problem is too difficult to solve based on technical challenges, I call BS. That's just cause someone doesn't really want to look at the problem in detail enough to break it down in ways that it can be solved. And we're not even talking about a hard technical problem to solve. It's a recommendation engine. They have them to suggest beds for patients in hospitals, medical advice to practitioners (aka what Watson is doing now), who will pay their bills, Pandora, and all kinds of mundane problem domains where a customer would appreciate recommendation with a confidence score. It is most certainly not rocket science, it's not even model rocket science. It's some marketedly trivial data collection, filtered querying, message dispatching, and some scheduled statistical aggregation. It's not even remotely interesting intellectually like a making DDO combat work with gesture control.

    Your counterargument appears to rest on 3 pillars:

    1. Turbine has no money.

    Bull. They've got the trivial amount of cash it would take to build what I'm suggesting, because they sure can spend non-trivial amounts doing things players seem to scratch their heads about and collectively sigh, "We asked for that?".

    2. Turbine has no coding skills.

    Open for debate, but again, just hire some. You can't tell me Turbine doesn't have IT ballast they can throw overboard in exchange, every organization does. Most IT departments are worse than union shops when it comes to throwing ballast overboard.

    3. Even if 1 & 2 are invalid, Turbine can't stop the trolls.

    Who says they have to? Recovery from a trolling is more important absolute prevention. Is your debit card troll proof? Do you never use it? Most of us, of the non-tin foil hat variety, don't care that our debit card is poorly secured, why is that? It's because the turnaround time for the bank to rollback illegal transactions on debit accounts is practically instantaneous, and that's if they don't call your cell phone first to ask you about a suspicious charge.


    No one's arguing Turbine should give the data to anyone, even that which you store with them in the form of preferences, and I'm still not even clear where you're getting that feeling from prior posts. A recommendation engine that belongs to Turbine, belongs to Turbine, data included. How is that confusing? Unless you think someone is going to hack Turbine and reveal your preference for zerging to the world?

    How do you suppose Target knew she was pregnant?

Page 2 of 2 FirstFirst 12

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

This form's session has expired. You need to reload the page.

Reload