Thursday 12 July 2012

The Quest for Ten Pissed Characters

When I introduced the What's Your Story posts over a month ago, I figured it would be a good way for the readers of The Adventure Gamer to get to know each other. With that in mind I think it has already been a success, and I'll be going ahead with it for as long as people keep answering the questions i posted in the sidebar. What has come as a surprise is the variety of occupations and talents that have already been put on show. Today I'm hoping to bring some of that talent to the blog, and our resident artist Dave_the_Turnip has very kindly offered his time in return for a nice little bag of CAPs. So how's this going to work?

It was Dave's excellent Sadwick drawing that got me thinking...

I'm happy about many aspects of the blog, but one I'm really not satisfied with are the Monkey Island characters that accompany my PISSED rating scores. It may not be apparent, but there are actually five different characters from LucasArts' series that can appear next to the rating. They are LeChuck (0-19), the Voodoo Lady (20-39), Bobo (40-59), Elaine (60-79) and Guybrush (80-100). What I didn't realise at the time is how bunched my scores would be for the first era of adventure games, which has resulted in about 80% of the games having exactly the same character. I want more variety and I need your help to make it happen!

I was naive to think I'd be pushing any of these out during the first couple of years!

With the community's help, I'd like to come up with 10 famous adventure game characters, each of which will cover a 10 point range (i.e. 20-29, 30-39, 40-49 etc.). My initial thought is to use some villains as the lower ranges and work up to favourite heroes for the upper ranges, but if someone has a better idea let's hear it! After all, I highly doubt there will be many games that sit in the 0-10 or 90-100 ranges, meaning characters selected for those may rarely (if ever) be seen. Whatever we decide on in the end, Dave will illustrate them for the blog using his own style, and in return he'll get a bunch of CAPs for his trouble. We'll start with some suggestions, with some possible polls if necessary.

Who knows? Maybe I'll soon be handing out April Ryan Awards or Larry Laffer Medals?!

68 comments:

  1. There is nothing inherently wrong with the PISSED system in terms of obtaining a gauage of the quantitative/qualitative "quality" of the games. What you may want to look into is the normal distribution of the PISSED ratings and determine the number of sigmas (standard deviations) from the mean each game achieves. If the number of sigmas is negative that means the game is below average, if it is positive then it is above average.

    I know I saw on your site at one point a google docs tracking of all the games you have played and their respective PISSED ratings. I can't seem to find it at the moment, but it would be a simple matter to implement this.

    ReplyDelete
    Replies
    1. Nevermind on the games list, found it. I'll be putting together my idea to see if it works out and to see where the mean and median game scores lie. Might be interesting.

      Delete
    2. For those interested here are the current PISSED statistics.

      Mean: 46.9
      Median: 50
      Mode: 52
      Standard Deviation: 9.7
      Sigma = (PISSED rating - Mean)/Stdev

      It should surprise no one to find that stinkers like Mortville Manor and Captain Blood (coming in at -2.1sigmas) are worse than 95% of all other adventure games.

      The current winner, Manian Mansion, is rated such that is better than ~91% of all adventure games.

      And, yes, I am bored at work...

      Delete
    3. For the interested here is the full breakdown of sigmas.

      Below the Root 0.5
      King's Quest: Quest for the Crown 0.1
      Déjà Vu -0.2
      King's Quest II: Romancing the Throne 0.6
      King's Quest III: To Heir Is Human 0.5
      Space Quest: The Sarien Encounter 0.8
      Tass Times in Tonetown 0.0
      The Black Cauldron -0.5
      Uninvited -1.7
      Leisure Suit Larry in the Land of the Lounge Lizards 1.0
      Maniac Mansion 1.6
      Mortville Manor -2.1
      Police Quest: In Pursuit of the Death Angel 0.5
      Shadowgate -1.2
      Space Quest II: Vohaul's Revenge 0.5
      Captain Blood -2.1
      Gold Rush! 0.3
      King's Quest IV: The Perils of Rosella 0.8
      Leisure Suit Larry Goes Looking for Love (in Several Wrong Places) 0.3
      Manhunter: New York 0.0

      Delete
  2. Oh wow, good idea!

    My first thought was a kind of recursive thing were you use the existing PISSED spread for the illustrations. That would mean the 0-10 rating would get a character from the worst game rated yet (Mortville Manor or Captain Blood!), and the 90-100 rating would get something from Maniac Mansion. This would of course mean that Dave would have to get a few new drawings done as time goes and we get new bests (or worsts), or in-betweens, but with the promise of CAPs maybe he'd accept an ongoing engagement?

    ReplyDelete
    Replies
    1. I was thinking reading the post, that it'd be great if i could draw something specific to each new game, but as usual my plans are greater than the time and energy i have to dedicate to them. :)

      I wouldn't mind changing some of the drawings up every now and then though. The worst/best game spectrum is an interesting idea, and over time it would give Trickster a large stockpile to use, but again, might be too much work ^^

      As a general reply to the thread though, i think there definitely has to be a Guybrush Threepwood, Purple Tentacle, Manny Calevara (or Glotis), and Larry Laffer... at least!

      Delete
    2. Building on this idea, you'll only need the picture if a game actually reaches that score. Maybe take the first game to reach it and immortalize a character from there would work.

      Every few years a new character can be cycled in when there's a new drawing, or an iconic game is reached.

      Delete
    3. Yeah, given that the problem is most games are hitting the same picture, I'm not expecting new high-low games very often, and it likely will get less often over time. That said, I can see not wanting to commit yourself to an unknown number of drawings. Perhaps we should do a kickstarter to hire you to do more of them, with CAP as rewards.

      Or am I getting ahead of things?

      Delete
    4. Yeah, I thought about the number of drawings as well, but since the last new "best" game was back in February, I don't see it happening too often. It all depends on what's "too much" work for you of course. :p

      But as Canahunk said, while the load would be higher in the beginning, as we get further into the list and more representation, the number of new drawings needed would be lesser.

      Delete
    5. This is true. Perhaps then an initial set of drawings with more added to them over time.

      Delete
  3. 10 characters, low to high:

    Low range: Sludge Vohaul... Ad Avis (http://bit.ly/NrHVou)... LeChuck.

    Middle range: Murray the Skull/Baba Yaga... Cedric the Owl... Purple Tentacle... Manny Calaveras.

    High range: Sam and Max... King Graham/Roger Wilco/unnamed hero of QFG... Guybrush Threepwood.

    ReplyDelete
    Replies
    1. I'd like to back these!

      Also, some of these have many forms (such as LeChuck, who has been a ghost, zombie and demonic pirate), which could provide a bit more variation.

      Not sure how popular Cedric the owl would be for mid-range though! http://www.youtube.com/watch?v=F64wDQJLWMQ

      Delete
  4. Wow ... it occurred to me that I never even really addressed your request in my earlier comments.

    The problem, as you lay it out, seems to be that you don't like using the same picture for every single game (that poor monkey has been overworked). The problem as I see it is that uniformly creating more categories won't really help you all that much. If you move to a 10% bucket histogram you will end up with generally speaking 1-2 pictures that always show up.

    If you look at the statistics I listed earlier, virtually every game you rate ends up in the 40-60 range. If you are fine with just using 2 pictures, then you don't have a problem, but if you are wanting more, then you need to weight the ranges such that the pictures near to the mean have a smaller range.


    A sampple range distribution (with games rated so far rated in ()s)
    0-30% (3)
    31-40% (1)
    41-46% (2)
    47-50% (5)
    51-54% (5)
    55-60% (3)
    61-70% (1)
    71-100% (0)

    This gives you quite a bit more variety (using 7 of 8 possible pictures) and may better visually represent the quality of each game.

    My previous posts would require a constant updating process of re-evaluating each game every time you play a new one. Although more accurate, this would obviously become very cumbersome in the long run. It might be interesting to see it included in the spreadsheet you have posted online, just to see how the games ratings hold up in the long run though.

    ReplyDelete
  5. If you look at the PISSED ratings as a normal distribution (which is not an unreasonable thing to do based on the numbers spread in my above comment) the three lowest games appear to be outliers.

    Generally when you get outliers, it means that they are not actually members of the set of data (for various reasons). Does this imply that those games (Uninvited, Mortville Manor, and Captain Blood) are not really adventure games?

    Or is it just that adventure games follow a twin peaked distribution profile (what I believe will turn out to be the case)? What this would imply is that there is a kind of quality valley for adventure games. There are a few games just suck and there is really nothing that can be said in their defense, but there are very few games that bad, but with some redeeming value.

    In short, the data seems to indicate so far that games are either:

    A. decent/good adventure game

    or

    B. piece of crap

    ReplyDelete
    Replies
    1. This comment has been removed by the author.

      Delete
    2. Hi Mooki, I'm also bored at work, but I don't think trying to find statistical regularity in a sample that has 20 elements makes much sense.
      Theoretically though, the twin peaked distribution could make sense if we take into account the fact that the data comes from a single subjective source: a big bunch of the games are actually rated against each other (it's the same phenomenon that happens when teachers give grades, it's always actually relative to the overall performance of the class), and some others are just disliked.

      Delete
    3. You are correct. A good rule of thumb is 30 samples for a minimal statistical analysis. That rule of thumb applies to loosely correlated data. The PISSED ratings are highly correlated though due to the fact that one person is responsible for generating them. Therefore I do not feel that I have committed a statistical heresy by applying them to the smaller dataset.

      Your "teacher giving grades" analogy is only apt in the abscence of a curve. Ideally a teacher assigns grades based on a quantitative analysis of the students performance independent of the other students. This of course gets thrown out the window when the teacher starts grading on a curve. Then the relative independence of each student's grades are dependent on the nature of the curving system the teacher applies.

      There is clearly nothing independent about the ratings system in PISSED as other games are frequently cited when giving the grades.

      I have nothing better to do today...

      Delete
    4. And that is why I oppose curving classes. Sure, if you have 1600 students in a class, like first year chemistry, or the even larger number in first year psych, then yeah, you probably have a pretty good approximatoin of a normal distribution. However, when you have 20 people in 3rd year quantum mechanics or 4th year inorganic spectroscopy then you quite likely don't. For example; I'm from the genius year, where we had several students who got straight A+s; Unlikely you'd get more then one of those in a year, but given how many years the class has run, it becomes pretty likely that you will get one that contains a highly non-Gaussian distribution of intelligence, and trying to force these into a Gaussian (normal) distribution doesn't help anyone.

      That said, 30 is a very ideal case; What do you do when you only have 10 samples? Say, 10 students in a class, or 10 water samples, and need to see if they are a normal distribution? It can be done, but it isn't pretty.

      Now, how does Trickster's ratings do as a normal distribution? Would it work better if we modelled it as two Gaussian, one at each centre? I know it can be done, but I don't know the math to implement it.

      Delete
    5. Building a distribution with two peaks at this point would be kind of crazy. There are only 3 data points that seem to fit within the lower peak, and building a gaussian off 3 data points would be statistical malpractice. lol

      You can make determine the mean and standard deviation from any number of data points (even 2!). The problem is that any predictive results are HIGHLY suspect with that few data points.

      20 data points probably indicates that the distribution is somewhere on the order of 80% accurate. By that I mean, if you took any adventure game from this time period you could use the distribution to determine the rating of the game with a certain level of confidence about 80% of the time. I'm probably mangling the explanation trying to describe a fairly complex statistical concept in a short amount of space.

      Delete
  6. There's nothing wrong with Trickster's idea of 10 pictures, but Mooki raises a good point. The problem with calculating a mean and standard deviation now, however, is that the games will tend to trend upward overtime, as adventure game design develops.

    Trickster, perhaps you can come up with your own "grading chart," that is calibrated to your own inclinations. Already, some stinkers were nowhere close to scoring a 0, and you've already said that scores of 90 and above will be extremely rare. So I think it boils down to your making a "grading chart," perhaps something akin to:

    Below 30
    31-36
    37-43
    44-50
    51-57
    58-64
    65-71
    72-78
    79-85
    86 and above

    ReplyDelete
    Replies
    1. Actually Rotgrub, if you examine the running average of the games (i.e. taking the average PISSED score after each game is rated and looking at the trend) shows that each time you get one of the really poor games you get a larger spike in downward PISSED rating average score than a series of slightly above average games brings it up. However, if you discount the outlier bad games, the PISSED scores have demonstrated the expected gradual increase in average score that one would expect with a grade on graphical quality and dialogue. (i.e. the memory limitations of the games expanding)


      When you break the PISSED scores out by the individual factors you get some interesting insights into the trends.

      The average rating for "P" (Puzzles and Solvability) has taken a comparative nose-dive. At first it was near an average of 6, now it is hovering around a 4.5 (including the outliers).

      The average for "I" has remained relatively consistent. I'm not sure what will make this rating ever break free of the 4-5 range.

      The average for the first "S" (Story and Setting) peaked early at 5.5 and is now hovering around 4.5. Is this because there are very few new settings and games are getting dinged for not presenting new things (4 games in KQs alone so far). Or is it because generally speaking Adventure Games typically have simple plot?

      The average for the second "S" (Sound and GRaphics) is, as expected, slowly increasing as time goes on but not by a whole lot. An early 2 rated game brought the initial average down to ~3.5 but now it is firmly set at ~4.3.

      The average for the "E" (Environment and Atmosphere) has hardly moved at all since the start of rating. It started around 6 and has stayed there. This makes me question whether this is a good metric for rating Adventure games.

      The average for the "D" (Dialogue and Acting") has steadily improved right along with the second "S". This is expected as games are allowed to grow in size.

      Delete
    2. Really? This doesn't match what I found before this game at all. I found when I tried to fit linear slopes to them I got terrible R^2; What did I do wrong?

      Delete
    3. I'm not sure what you did wrong. Maybe nothing even, the running average doesn't really follow a linear relationship. It is more of a log relationship or an inverse power relationship depending on if the trend is increasing or decreasing. Both of those would have fairly poor r^2 if you fit with a linear. It depends on HOW bad your r^2 was. If you were getting in the ~0.5-0.7 range you were probably fine.

      Don't always rely on r^2 either. For comparing non-linear trends the r^2 is not reliable. By which I mean you cannot compare the r^2 of a power relationship with the r^2 of a log relationship.

      Delete
    4. I really like this suggestion Rotgrub! In fact, I think those ranges work really well. We'll see what else comes out of this post, but for now I think that's a good way to move forward.

      Delete
    5. Less then .1 in a lot of cases. I did try some other functions, but none of them seem to fit either, can I see how you plotted things?

      Delete
    6. Rotgrub's ranges do not provide as much diversification in pictures for rated games at this time as the percentile buckets I recommended do. This is dispite the fact that I recommende even fewer brackets. Rotgrub's breakdown would give the following:

      Below 30 (3)
      31-36 (1)
      37-43 (1)
      44-50 (6)
      51-57 (8)
      58-64 (1)
      65-71 (0)
      72-78 (0)
      79-85 (0)
      86 and above (0)

      The reason Rotgrub's percentile buckets are not ideal is that each bucket is the same size. In order to obtain a greater spread of images you will need to make the buckets near the average score smaller.

      On the other hand, it is everyone's expectation that adventure games yet to come will pull the averages scores up due to improvements in graphics and sound. See a previous comment of mine on the current trends in each category. The games at this time are NOT improving. The average PISSED rating over time is actually decreasing. A linear fit to the games give a slope of -0.218. This is due to the outliers I talk about earlier. If those are removed the negative slope is insignificant (-0.027 .. well within the margin of error).

      At this time, the next 6 games would need to achieve a score of 60 (very nearly the max so far) to bring the average up to 50. It would take the next 29 games being rated 60 to raise the average to 55!

      There are some great games coming in the future ... but not that many. The quality of adventure games has not improved THAT much.

      Given the unliklieness of the above, I think the focus should be on smaller buckets near the current average to provide a greater variety of images displayed.

      Delete
    7. Canageek:

      I assume you are using Excel (it's the only one I've ever really used so I can't comment on other spreadsheet input methods).

      You have a column of data in column "A":

      1
      2
      3
      1
      5
      2

      You calculate the running average by putting the following equation in column "B"

      =average(A$1:A#)

      Where # equals the row number the equation is in.

      Doing this provides the following r^2 values for each PISSED category (including outliers):

      P = 0.87
      I = 0.15
      S = 0.75
      S = 0.54
      E = 0.33
      D = 0.16

      Delete
    8. Ah, I was plotting the absolute value, not the running average. What justification is there for averaging things as you go along? I would have thought you should use the absolute value, since otherwise you are going to weight your line towards early games, since they have less points averaged in.

      Delete
    9. "There are lies, damn lies, and statistics" Mark Twain (iirc)

      Your point about weighting to early games is true. Using the running average, however, does emphasize that the more games get played the harder it is to move the average score. This reduces the noise level of average games; making any perturbation on the running average line the indication that a significant game occured.

      Perhaps what you may have wanted to use is a "moving average" where at each step you average only the adjacent games together. Each data point would be the average score of that game and the games immediately preceeding and following.

      Taking the moving average fit to a linear trendline gives an R^2 value of 0.068. This is a truely atrocious R^2 value indicating that the ratings scores are not following a linear relationship.

      All that that implies is that as time progresses the quality of games is not improving or decreasing. This is intriguing to me, because intuitively one would think that due to the learning nature of "art" you get better with more practice. Adventure games have so far not demonstrated that principle. The only category that is improving is graphics (R^2 of ~0.4, really bad but starting to ressemble something you can declare a trend). Every other category has a R^2 value of < 0.2 which means there has been no linear relationship between time and quality.

      Delete
  7. You could do a poll of favorite adventure game characters, and use the ones with the most votes for the high scores, and the one with least for the lower ones.

    ReplyDelete
  8. What about some kind of "moving" intervals, based on the distribution from the particular day when the game was rated?
    So if a game falls into the first decile (the 10% with the the lowest score), it gets characters number 1, if it's in the second decile, the 2nd character, etc.

    ReplyDelete
    Replies
    1. That would be the idea that people are going on about with a normal distribution.

      Delete
  9. How about some additional icons for something like achievements.

    For example, Purple Tentacle - awarded to games with fiendish, illogical puzzles
    Guybrush Threepwood - awarded to games where it's impossible to die/get stuck
    Indy from FotA - awarded to games where there are several ways to solve the game.

    Good/bad idea?

    ReplyDelete
    Replies
    1. Guybrush Threepwood wouldn't be a good one for no way to die or get stuck since he CAN die in his first game (drowning .... eventually). I can't actually think of a game that has no deaths or ways to get stuck that don't also have the player as some sort of AFGNCAAP (Ageless, Faceless, Gender-Nuetral, Culturally Ambiguous Adventure Person), and even then it is exceptionally rare (from my experience).

      Myst and Riven are about it, but those games treat getting stuck as a form of bad ending.

      I love the idea of game achievements though. Perhaps they could be oriented more towards the TRickster's playthrough.

      The game gets an achievement if the Trickster gets stuck for more than 30 minutes.

      The game gets an achievement if the game tempts T to die just so he can see the crazy way it kills him (I'm looking at you Space Quest lol).

      Delete
    2. Guybrush -- Exceptional witty dialogue?

      I say if something stands out in a game, then an award gets created, if Trickster adopts this idea.

      Delete
  10. Ugh how did I pass statistics in college?

    ReplyDelete
    Replies
    1. I never took it, to be honest. Or in high school. I sat through three days of it then dropped it for...something else. I forget.

      Delete
    2. I started out trying to triple-major in College (Math / Comp Sci / Poli Sci), then dropped the Political Science. Strangely enough, even though I spent more time doing computer stuff than anything else, I ended up with just a Math major. The reason? One Math class - Statistics - too few to qualify for a CompSci degree. :-)

      So I like to say I majored in Mathematics because I didn't take enough Math to be a Computer Science major.

      Delete
    3. Irony at its best!

      I failed Computer Studies at school because I didn't have a PC at home to learn on. Now I'm an IT Manager of a global company. Things often don't turn out the way you expect them to. :)

      Delete
  11. Most of the characters mentioned thus far appear to be from Sierra or Lucasarts, so just for a balance I'll suggest Malcolm from Kyrandia and Simon the Sorcerer. Personally, I'd also like to see what Dave could make out of AFGNCAAP ("Ageless, Faceless, Gender-Neutral, Culturally-Ambiguous Adventure Person"; a great symbol for all the average games) from Zork: the Grand Inquisitor, but I don't think I'd get much support.

    ReplyDelete
    Replies
    1. Other famous icons would be a good idea (and i can't believe in my suggestions i forgot Tex Murphy) ><

      I already have an idea on how i could do AFGNCAAP, but that is a tough one :D

      Delete
  12. Quite tempted to make you a full 100... wouldn't take that long, and I'm sure there are 100 characters out there!

    The only problem would be putting them in an order that would satisfy most people.

    ReplyDelete
  13. Alright, sale alert! Am I given points per game? Because Steam has on sale:
    http://store.steampowered.com/sub/15168/?snr=1_614_615_summersale2012collections_616_14
    Cyan Complete Pack:
    Cosmic Osmo and the Worlds Beyond the Mackerel
    Myst V
    Myst: Masterpice Edition
    realMYST
    Riven: The Sequel To Myst
    Spelunx and the Caves of Mr. Seudo
    The Manhole: Masterpiece Edition
    URU: Complete Chronicles

    http://store.steampowered.com/sub/15409/?snr=1_614_615_summersale2012collections_616_20
    TellTale Collection:
    Back to the Future: The Game
    Jurassic Park: The Game
    Puzzle Agent
    Puzzle Agent 2
    ****Tales of Monkey Island Complete Pack***
    The Walking Dead.

    I think you can buy the games one by one as well, still 40% or more off, as I don't think most of those are adventure games.


    *Goes off in search of more CAP*

    ReplyDelete
    Replies
    1. The general rule of thumb with steam sales is to wait till individual games are discounted on a certain day (or seemingly with this sale, every 5-8 hours). If the game hasn't shown up by the final day, buy the pack.

      It saddens me that i've spent enough money on these sales over the years to come to that conclusion. ^^

      Delete
    2. Most of the single games do seem to be on sale; the packs are just on sale on TOP of that, to encourage you to buy the pack with a couple more games in it.

      Delete
  14. http://store.steampowered.com/sub/15239/
    Midnight Mysteries Collection:
    Midnight Mysteries
    Midnight Mysteries 2: Salem Witch Trials
    Midnight Mysteries 3: Devil on the Mississippi
    Midnight Mysteries 4: Haunted Houdini

    http://store.steampowered.com/sub/15352/
    The Clockwork Man Bundle:
    The Clockwork Man
    The Clockwork Man: The Hidden World

    Pendulo Adventure Pack
    http://store.steampowered.com/sub/15253/
    Runaway, A Road Adventure
    Runaway, The Dream of The Turtle
    Runaway, A Twist of Fate
    The Next BIG Thing
    Yesterday
    (These pretty much all look like adventure games from the screenshots)

    Most of these games are on sale individually. The packages add the *sale* prices together, then take a bit off of that, so if you were going to buy every game anyway, you save a little bit more.

    I hope there is at least a little CAP in this, as I stayed 20 min after work sorting through all the packages trying to find adventure games on sale.

    ReplyDelete
    Replies
    1. I'm going to give you 10 points Canageek for all your work in this post, but I think there needs to be some clarity around what sales and game announcements people get CAPs for.

      I'll think about it.

      Delete
  15. Newer adventure game:
    Adam's Venture Complete Pack
    Episode 1: The Search For The Lost Garden, Episode 2: Solomons Secret, Episode 3: Revelations
    http://store.steampowered.com/app/214310/?snr=1_200_200_251_201_12
    Might be action-adventure? Hard to tell from the screenshots. Lot of people talking, not much shooting or jumping, which normally means adventure.

    The Dark Eye: Chains of Satinav
    http://store.steampowered.com/app/203830/?snr=1_200_200_251_201_11
    Looks very adventure-gamey


    Ok, I'm done; there are probably more though.

    ReplyDelete
  16. Alright, I'll freely admit that this isn't an Adventure game; it is a roleplaying-board-party game. However, some people I know are making it, and the predicted finish amount just dropped to 99%, so I thought I'd use some of my 'celebrity' status to get away with posting it, just in case one of you wanted in. I know how much it sucks to see a Kickstater you really want die, and it must suck even more from the other side of the project.

    Project Ninja Panda Taco:
    http://www.kickstarter.com/projects/962662760/project-ninja-panda-taco

    Project Ninja Panda Taco is a game where you get to play a Mastermind trying to take over the world. Along the way, you’ll compete as a Nemesis against other Masterminds and as a Minion who loves to help, while receiving rewards for their hard work.

    Project Ninja Panda Taco is about Masterminds taking over the World and the Minions who try to help. The game is a low-prep, lighthearted, collaborative roleplaying game that is perfect for everyone around the table.

    ReplyDelete
  17. While the idea of a distributed scoring system is not a bad one in theory, I really don't like the idea of having to go through and change the scores on all previous games each time I finish one. As I mentioned above, I think rotgrub's suggestion is a good one. With a grading chart like that, we'll get a bigger variety of images appearing, and we should be able to use every image at some point.

    I think we need to start focussing on which characters we could use, so I've put together a list of protagonists and antagonists to get things moving.

    April Ryan
    Bobbin Threadbare
    Dr. Fred
    Gabriel Knight
    Guybrush Threepwood
    Kate Walker
    King Graham
    Larry Laffer
    Laura Bow
    LeChuck
    Malcolm
    Manny Calavera
    Purple or Green Tentacle
    Roger Wilco
    Rosella
    Sam & Max
    Simon the Sorcerer
    Sonny Bonds
    Tex Murphy
    The Hero
    Vohaul

    There's a mix of protagonists and antagonists there. I would think, given Dave's style (correct me if I'm wrong Dave), that the more cartoonish characters would be more easily represented, so perhaps the likes of April Ryan and Kate Walker should be excluded.

    Thoughts?

    ReplyDelete
    Replies
    1. While i have a cartoony style, i could do the more realistic characters justice, at least in likeness :)

      Delete
    2. The problem with Kate Walker and April Ryan is, IMO, that their look is more "realistic", i.e. not iconic enough to be recognizable in the same way as, say, Lara Croft would be. I've played the Syberia and TLJ games, but I know I'd have trouble distinguishing either protagonist among a bunch of generic 3D characters. Do they even have a signature outfit?

      I guess if Kate was surrounded by mammoths and automatons I *could* make the connection...

      Delete
    3. Eh, the real problem is that there's so few recognizable female characters in adventure games at all. I could go for April Ryan, but swap out Kate with Elaine Marley maybe?

      Delete
    4. If you've played Dreamfall (The Longest Journey's sequel), I'm pretty certain you'd be able to recognise Zoe Castillo. ;)

      Delete
    5. Trickster:

      See my above comment. Rotgrub's method does not provide the spread of pictures you are looking for. The percentile range buckets near the average need to be smaller to get a greater variety of images.

      It is tough to get away from Sierra and LucasArts characters because they established the genre. They were the premier publishers during the adventure game golden-age. Therefore I make no attempt to avoid using their characters when making my list and justifications.

      1 Vohaul (the biggest bad guy from any of Sierra's series)
      2 LeChuck (the biggest bad guy from any of LucasArt's series)
      3 Purple or Green Tentacle (being a tentacle is pretty bad)
      4 Larry Laffer (being a depraved loser just trying to score may be funny, but it isn't all that great)
      5 Roger Wilco (he's a janitor ... a step up from depraved loser)
      6 Guybrush Threepwood (a wannabe pirate .. at least he's not cleaning toilets between adventures)
      7 Tex Murphy (hard(ly)-working private eye who has saved the world a few times)
      8 Laura Bow (a strong resourceful, intelligent, and respected adventure woman)
      9 The Hero (the great hero, stopped demon invasions, defeated cthullu-like monsters from the beyond)
      10 King Graham (the guy who started it all)
      11 AFGNCAAP (that's you, the player, the ultimate adventure gamer, who has beaten countless obstacles and adversaries to save the day, time after time, game after game)

      Delete
    6. Personally for me, I'm not too concerned that the scores cluster closely around a certain average. And I guess it also depends on whether the Trickster thinks his ratings will trend upward as we go into the 1990s, because if they do then there will be a somewhat more of a spread.

      If Trickster thinks that some games will go up as high as the 80s, then yes let's leave those categories/pictures rare. Let's allow the best to be an elite few. If, however, you don't see many games going above 70, then we can set the "bands" to be closer to each other.

      Delete
  18. I really like that mix. A lot. Do we have to pick and choose ten? If so, from that list, low to high:

    Vohaul (who/what better to represent the sludge of gaming?)
    Purple/Green Tentacle
    Malcolm
    LeChuck
    Roger Wilco (the perfect symbol of mediocrity)
    Tex Murphy
    King Graham
    Sam and Max
    Manny Calaveras
    Guybrush Threepwood (always first in my book)

    ReplyDelete
    Replies
    1. I wouldn't be opposed to that list. It's still very Sierra and LucasArts heavy, but that's not really a surprise.

      Delete
    2. Yeah.

      I think these lists are always going to be Sierra/Lucasarts dominated. They're the only companies I can think of who had the same characters appear in more than one or two games.

      The main exception being characters taken from literature (Sherlock Holmes, Nancy Drew)

      Interesting that Manny Calavera appears in a lot of people's wishlists when he only appeared in one game. Just goes to show how much of an impact Grim Fandango had on people.

      Delete
    3. Uh, I tried to set up my own list, and it's almost impossible!

      LeChuck
      Purple Tentacle
      Malcolm
      Roger Wilco
      Simon the Sorcerer
      Laura Bow
      April Ryan
      Tex Murphy
      Manny Calavera
      Guybrush Threepwood

      I think that's a fair balance between LucasArts, Sierra and Others, and it has some female representation.

      What I can't decide about is what criteria to use. I put Simon in the middle as a fairly inept anti-hero, but does that invalidate putting Guybrush at the top? Why put him there at all? Hmm...should I do evil->good, antagonist->protagonist, sort by popularity, my personal feelings towards the games, or what? Hence, a fairly jumbled together compromise of all of the criteria. :p

      Still thinking.

      Delete
    4. Oh, and where's Larry?

      As Lars-Erik says, criteria is the main issue here. Are these ordered in terms of villainy, competence, popularity...? Imagine if there was a brief text illustrating the evil-good scale. LeChuck would be reserved for the lower tier: "Uber-Villain". Henchmen would follow, like Murray the talking skull -- it could be next as "Powerful Demonic Force". A mid-range option could be "Brave janitor" (Wilco). As much as I want Guybrush to be at the top, he'd be cool for an "Adventurer wanna-be". While the cream of the crop would be represented by Indy and "Adventure is my name".

      Delete
    5. I thought about both Murray and Elaine, but with the limit of ten slots, I couldn't justify using three or four of them just for Monkey Island.

      Considering the evil->good scale, I might have done this instead:
      Purple Tentacle (Yes, more evil than LeChuck, just face it)
      LeChuck
      Malcolm
      Sam&Max (thanks to Max they ended up just shy of neutral)
      Roger Wilco
      Simon the Sorcerer
      April Ryan
      Manny Calavera
      Tex Murphy
      Indiana Jones

      That drops out both Guybrush and Laura Bow though. Laura Bow I can do without, but it just seems empty without Guybrush. If I HAD to have Guybrush in, I'd put him at 5/6, either swapping out Roger Wilco or Simon, but that could drop Sierra totally off the list. :(

      Delete
    6. Well, we're close to figuring this out. How about this for a solution?

      Since Dave is the one that actually has to draw these characters, I vote we let him choose which ones he wants to draw. We can then figure out how to sort them from there (it will probably be obvious).

      April Ryan
      Guybrush Threepwood
      Indiana Jones
      King Graham
      Larry Laffer
      Laura Bow
      LeChuck
      Malcolm
      Manny Calavera
      Purple Tentacle
      Roger Wilco
      Sam and Max
      Simon the Sorcerer
      Tex Murphy
      Vohaul

      Delete
    7. My only suggestion for all this is have the list mirror itself. With the protagonist from a game in the high rating flipped to the antagonist from the same game or series for the low scale. If Guybrush is 91-100 and the best of the best then LeChuck should be the worst of worst at 1-10.

      Delete
    8. That's an impressive list. I kinda want to draw them all, so i think i'll start sketching this week and see what happens. :)

      Delete
  19. Gemini Rue is now on GOG for only $4.99! I haven't been buying new games for obvious reasons, but I've snapped it up on reputation alone.

    ReplyDelete
  20. I look forward to seeing what Dave comes up with. :)

    In the meantime, I wanted to hand out some CAPs to those I felt made a significant contribution to this discussion.

    rotgrub gets 10 because I really like his grading chart idea
    Mooki and Canageek also get 10 for the statistical effort they put in
    Lars-Erik, JosephCurwen, Andy_Panthro, Knurek, Ilmari and Charles all get 5 points for their contributions.

    Thanks all!

    ReplyDelete
    Replies
    1. Woot! Do I have enough to force you to play a couple of games yet?

      Delete