CR/SP and the players, what does it look like?

Discussion in 'Gotham City (General Gameplay)' started by Polymerase, Feb 27, 2016.

  1. Polymerase Well-Known Player

    For those of you who are interested, here's a heatmap for CR/SP, showing how many characters there are with the given combination.
    This relates to a question posted in:


    Notes:
    • Data are collected over a 24 hours period (there may be character duplications) from the census.
    • There's an oddity at CR14 (due to a query timing out, there might be one or two others - I've resolved this for future queries).
    • This is only for the US server.
    • Heatmap is only from 0-100 counts - for display purposes.

    [IMG]

    For an interactive map, please see: http://jsfiddle.net/chrisbridson/ase47f78/

    Enjoy.
    • Like x 3
  2. Van_Gho Committed Player

    dumb this down, I don't have enough brain sp to figure that map out
    • Like x 2
  3. Polymerase Well-Known Player


    CR is along the bottom left-to-right (CR 0 to 156) .
    SP is along the side, bottom-to-top (SP 0 to 264 ).
    Where the two meet, the colour represent the number of characters with that SP/CR combination - to a max of 100.
    Red represents a count of characters with equal to/greater than 100.
    The colour scale at the bottom represents the counts, to make it easier to visualise.
    • Like x 2
  4. Derio 15000 Post Club

    I dont think Census data can be properly used until the active character data problem is resolved.

    As you know a person who quit over 2 years ago can still be considered active thus their skill points are not going to increase.
    • Like x 2
  5. BipolarDiva Loyal Player

    Hmmm. Still don't get it.
    • Like x 1
  6. Veritech Loyal Player

    Another chart that loosely shows the effect that gu36 had. Note the distinct loss of density just after cr 100.

    Also look at that spread at current endgame! About 50sp to 265 range.
  7. Aren Sul Committed Player

    You had me at heatmap ;)

    I really like visual representations of data. The main problem is as Derio points out regarding "active players" and could affect Veritech's interpretation.

    Considering that, I wonder if that's why there is a mysterious gap between CR 85-95.
    • Like x 1
  8. Kimone Luthor Genetech Clone

    Aside from whatever potential issues were identified by the above responses -

    This is neat. Good show.
    • Like x 1
  9. Polymerase Well-Known Player


    Totally Agree.
  10. Polymerase Well-Known Player

    Part of you does :p

    I think it's caused by a drop in the number of players at those CRs (and an artefact from filtering on 100), going from 0-50:

    [IMG]


    It also ties in nicely with some (older, from 24th Jan - sorry I can't pull out the latest set at the moment) data for character/CR counts:
    [IMG]
    • Like x 1
  11. Aren Sul Committed Player


    Yep, that is a nice overlay and more than likely accounts for that gap.

    Cool.
  12. Superskull85 Devoted Player

    Well considering how the the heatmap is setup it doesn't really matter since their inactiveness would be drowned out. That said there are ways to determine activeness. Polymerase should already have the means to do that since a proof of concept was in the Character Name Tracker I posted and he posted in that thread.

    The general principle is to track a subset of characters each X CR/Skill Point interval. If they've been saved in the last 24 hours keep them. If not throw them out and select X new characters. Careful tracking and wall thought out queries to avoid timeouts can address your concern. However like I said inactivity is fairly moot given how the heatmap is setup. You can also track the delta in the comparison and make a heatmap of the trend. This would tell you how the community moves through the tiers (and thus active).

    BTW the "problem" was player made. Due to that it is not as simply as "fix it" since, really, Census is working the way the community wanted when it was first released.
    • Like x 1
  13. ProfessorKanua Committed Player

    This tells me that the average SP at endgame is around 120 and I'm okay with that.
    • Like x 2
  14. Polymerase Well-Known Player


    What ranges are you basing endgame on?
    The average for a CR of 156 is ~202 SPs.


    Here's a better image (simple boxplot for a CR 30-156):
    [IMG]
  15. Jeazy Baby Dedicated Player

    rase, post: 3605940, member: 491"]What ranges are you basing endgame on?
    The average for a CR of 156 is ~202 SPs.


    Here's a better image (simple boxplot for a CR 30-156):
    [IMG][/quote]
    much more accurate,
    a quick search on the census (DCUOLIVE) shows that there are

    6,686 CR156 characters
    of those characters
    3,335 > 213sp (49.88%)
    3,351 < 214sp (50.12%)

    I have no idea where that guy got the average to be 120sp for cr 156.
    Only (4.14%) 277 out of 6686 have 120 sp or less.
  16. xoHLxDPSox Steadfast Player

    221sp, working on 222 baby :D
  17. ProfessorKanua Committed Player

    Just eyeballing the intitial chart, based on the information given. Didn't state anything as fact. Also was looking at 150-156, since it can be assumed that people that chase Max CR also chase Max SP. You have to look at the drop-off when you go a couple of CR's down and then take an average. A second look at the initial chart tells me that an estimate of 150 would be more accurate.
  18. Polymerase Well-Known Player

    For Completeness, here's the US/EU data side-by-side (or above/below....).

    Heatmaps (50 Character cut off - red):
    [IMG]
    [IMG]


    BoxPlots:
    [IMG]
    [IMG]