1. We have added a Gift Upgrades feature that allows you to gift an account upgrade to another member, just in time for the holiday season. You can see the gift option when going to the Account Upgrades screen, or on any user profile screen.
    Dismiss Notice

VVV scoring system suggestion - Updated

Discussion in 'Civ5 - Hall of Fame Discussion' started by Xger, Apr 16, 2013.

  1. Xger

    Xger Prince

    Joined:
    Mar 22, 2010
    Messages:
    510
    Location:
    Denver
    Did you mean one for score and one for time games? The concern I would have there is that further splitting the players would seem detrimental to the idea of getting more players. Not to mention needing another 34 games (soon 43) for the second VVV.
     
  2. vexing

    vexing knows

    Joined:
    Dec 24, 2010
    Messages:
    2,668
    no, vvv is currently just points, separated by either g+k or vanilla, with no distinction between points that were earned for score games vs fastest finish.
    since they're completely different goals (you can get fastest finish and a low score, or you can get an amazingly high score and lose the game to an AI), i think they should be separated into a subcategory.

    ie there should be 4 drop down choices atm: vanilla - time, vanilla - score, g+k time, g+k score
     
  3. Denniz

    Denniz Where's my breakfast? Moderator Hall of Fame Staff

    Joined:
    Nov 28, 2003
    Messages:
    11,092
    Gender:
    Male
    Location:
    Dallas
    I would prefer to weight the scores according to difficutly, map size and spead. Difficulty can be liner but map size and speed probably should reward the middle ranges. Maybe some kind of bell curve.

    I think points for 10 places is good. But only if certain minimum are there like with the medals. Maybe range of 1 to 10 counting up. Five entries get 1 to 5. Something like that.


    Don't worry about the SQL or efficiency issues. We can work out efficient ways to do something as long as it isn't too insane.

    An idea\plan with proper math worked out is what we need.
     
  4. kirbdog

    kirbdog King

    Joined:
    Jun 14, 2005
    Messages:
    898
    Weighting scores by speed just discourages people from playing the speeds with lower weighting. I don't see the point in that.

    Weighting by map size only makes sense as a way of giving less weight to things like bashing out duel domination games. There's no reason to give higher weight to standard than large and huge. What makes more sense is to plateau the weighting for map size either at small or standard, making games with 6-12 or 8-12 civs equal in weight.
     
  5. Denniz

    Denniz Where's my breakfast? Moderator Hall of Fame Staff

    Joined:
    Nov 28, 2003
    Messages:
    11,092
    Gender:
    Male
    Location:
    Dallas
    No one wants to play the higher end of things. Marathon on Huge anyone? Quick Duels probably make up the majority. So throwing them out now doesn't seem fair. But lowering weight of the easier combos while reward towards the middle reward more difficult without rewarding the upper too where no one what to play anyway.

    I know it sounds weird but that's what I was think. ;)
     
  6. sanabas

    sanabas Psycho Bunny Hall of Fame Staff

    Joined:
    Nov 24, 2004
    Messages:
    4,269
    Location:
    Canberra, Australia
    If nobody wants to play the bigger, longer games (I'd rather play them ;)), then there's no reason to penalise them with a lower weight. Especially as it seems like any game in an empty or near-empty table will continue to be penalised already, if not ignored completely.

    I think one big issue is to cap the number of games that can contribute to one person's score. Set it high enough that anyone who wants to maximise their score will need to play quite a few games of a wide variety, but don't have it so that volume of games submitted ends up being the major criteria.

    Capping the number of games is also a way to have different scores for each subsection. I don't see the point of having all the subsections listed when they all show the same set of scores. And the ability to focus on one of them may generate more interest.


    Biggest issue I've got with what's been proposed so far is that it still discourages people from filling empty tables, and discourages them from submitting games that don't beat an existing result. If you've only got 1st place because nobody else has played that particular game, so what? It's an incentive for someone else to beat you. Do a really good time on an empty table, and it's likely to just sit there giving you nothing because nobody can beat it and so won't bother playing. All #1 times are definitely not equal, but I don't think the inequality of giving points to a sole game on a table is much different to the current level of inequality with medals. All bronze, all silver, all gold medals in the tables now aren't equal either, not even those with the same number of games. The only games that are really equivalent are the gauntlet winners. If I didn't have a new Vicky 2 expansion to play with right now, it'd be very easy to cheese my way to at least half a dozen or more G&K golds for next update, reach at least 2nd if not 1st overall. Will take minimal skill, simply game the scoring system and spend a big chunk of time.

    Alrighty then...

    I'll post one, in detail, later today. Things I'll be aiming for:

    -Sole games in a table still get rewarded.
    -Topping a full table gets more reward than topping an empty table.
    -Number of scoring games is capped, so that volume isn't king, and so that different sections have different scores.
    -Harder games get more points.
    -Standard+ games get more points than quick/duel/tiny.
    -Not insanely complicated to calculate.
    -Can be set up to produce either 6 sub-scores, summed for overall VVV score, or 6 time sub-scores, 6 point sub-scores, summed to create a fastest finish VVV and a best in-game score VVV, then put those together for a combined VVV.
     
  7. kirbdog

    kirbdog King

    Joined:
    Jun 14, 2005
    Messages:
    898
    It does sound strange. Besides, I'm sure there's one or two guys who like marathon or huge or both, and if they want to fight over medals, I don't think there's any reason to give them lower weight.
     
  8. vexing

    vexing knows

    Joined:
    Dec 24, 2010
    Messages:
    2,668
    i'd weight by size as follows:
    duel 1
    tiny 1.2
    small 1.4
    standard 1.6
    large 1.8
    huge 2.0

    difficulty i'd cut in half current scaling, ie
    settler 1
    chieftain 1.5
    warlord 2
    prince 2.5
    king 3
    emp 3.5
    immortal 4
    deity 4.5

    speed i wouldn't change, although marathon takes more turns and effort, it's also relatively easier mainly due to players being much better at combat than AI.

    for points prior to difficulty / size multipliers, i'd give
    for 6 or < participants, (participants - 3) * (2/3) ^ (place - 1)
    for > 6 participants, (3+ sqrt(participants - 5)) * (2/3) ^ (place - 1)
    both only given where points >= 1

    the (2/3) ^ (place -1) just means each place would get 2/3rds the number of points the place before them got, eg instead of our current 9 / 3 / 1 it'd go 9 / 6 / 4 / 1.333

    the sqrt above 6 participants is intended to continue giving rewards for a larger amount of competition but limit the scale of that reward.

    here's how the formula would work out into actual points:
    Column 1 Column 2 Column 3 Column 4 Column 5 Column 6
    participants 1st place 2nd place 3rd place 4th place 5th place
    3 1
    4 2 1.3
    5 3 2 1.3
    6 4 2.6 1.8 1.2
    7 4.4 2.9 2.0 1.3
    8 4.7 3.2 2.1 1.4
    9 5 3.3 2.2 1.5
    10 5.2 3.5 2.3 1.6 1.1
    15 6.2 4.1 2.7 1.8 1.2
    25 7.5 5.0 3.3 2.2 1.5



    this would be a pretty good reward based upon merit
     
  9. kirbdog

    kirbdog King

    Joined:
    Jun 14, 2005
    Messages:
    898
    I'd suggest these size weights:

    Duel 50%
    Tiny 75%
    Small 100%
    Standard 100%
    Large 100%
    Huge 100%

    The only problem with small maps are the really small ones that currently are the "low hanging fruit" for banging out fast victories.
     
  10. TractorBoy

    TractorBoy Prince

    Joined:
    Jul 14, 2012
    Messages:
    435
    Location:
    England
    That seems ideal to me.
     
  11. Mesix

    Mesix One of Porg

    Joined:
    Mar 17, 2006
    Messages:
    4,358
    Location:
    Ahch-To
    I really think that scores should be based on how close an entry is to the best finish rather than a fixed amount for each place (something like the EQM system in Civ 4). To me, if the first place finisher is turn 180, the 2nd place is 182, and third is 240, the 2nd place should have a score very close to the top finisher and there should be a much lower score for 3rd.
     
  12. Mesix

    Mesix One of Porg

    Joined:
    Mar 17, 2006
    Messages:
    4,358
    Location:
    Ahch-To
    Actually, the recent Gauntlet is a really good example of my point.

    1...325
    2...326
    3...335
    4...339
    5...352
    6...355
    7...357
    8...371
    9...380
    10..391
    11..395
    12..404
    13..429
    14..433
    15..493
    16..513
    17..546
    18..581
    19..606
    20..674
    21..706

    1st and 2nd are separated by only 1 turn, and then 3rd by 10 turns. 2nd place deserves a score very close to the 1st place finish IMHO.

    Also, in response to an earlier comment, I am not in favor of rewarding more entries simply because there are more entries. Once a table reaches a certain level (perhaps 6 entries or 10 entries), the score should be capped with 1st place getting full score and each place under first getting something less than full score.
     
  13. Moriarte

    Moriarte Immortal

    Joined:
    May 10, 2012
    Messages:
    1,871
    I like it! Thought:

    For each game, which within 5% (in turn times) of the leader the player receives a 20% boost towards his final score.

    So for current medal values in gauntlet L it would stand like this:

    1...325 6 points
    2...326 4.8 points (4 +20%)
    3...335 2.4 points (2 +20%)

    And so on until the 10th place. 5% modifier can be, of course, adjusted upwards on epic/marathon and vice versa on quicker speeds. I don't know, maybe progressive modifier would be better than a static one? (20%)
     
  14. vexing

    vexing knows

    Joined:
    Dec 24, 2010
    Messages:
    2,668
    i disagree. there is always some min number turns required to win whatever settings, and the difficulty of achievement rises exponentially the closer you get to that minimum. the reward should therefore rise equally so and 180 getting a good amount more points than 182 is appropriate. the problem with 3rd place in your example above solves itself via more entrants.

    note that i also don't agree with the current scaling (3x the points for the winner vs 2nd place). my proposal above isn't quite so harsh to lower placements.
     
  15. Miraculix

    Miraculix Warlord

    Joined:
    Jan 2, 2005
    Messages:
    154
    Personally, I would like to see the following changes to the scoring system:
    • Award people who make the effort to open up an empty table (i.e. some small amount of score for single entry tables).
    • Assign scores below third place. This will encourage mediocre players like myself to try to improve from e.g. 15th place to 10th place in a competitive table where they have no chance of earning a medal. Also, it will provide a better way to rank the large number of players (73%) in the overall Gauntlet table who never earned a gauntlet medal.
    • In each category, discourage people from submitting all their top games in only one sub-category. For example, in Tempi Trophy, the scoring system should discourage a player from submitting e.g. 20 gold medal quick games while submitting only a single 2050 AD game to each of the standard, epic and marathon sub-categories. One way to achieve this is to put a limit to the number of games that can count in each sub-category. This will also change the fact that each player currently has exactly the same number of points in each category.
    I believe this will also improve on the current lack of diversity. Examples:
    • There seems to be a large majority of quick/duel/domination games.
    • Of the 218 vanilla deity games submitted, 50% are in the 10 most popular tables.
     
  16. Xger

    Xger Prince

    Joined:
    Mar 22, 2010
    Messages:
    510
    Location:
    Denver
    Take two on the system:

    Here seem to be the general things desired:
    • The top performances need a higher weighting
    • Scores need to go further down the table
    • Map size and difficultly need to be accounted for
    • There should be two seperate VVV's; one for score and one for time, though there wasn't much discussion if they should combined for the overall ranking, if not, which would display on the page?
    • Playing an empty table needs something to encourage new tables to be filled
    • More entries should be more points, up to a limit

    Each entrant grants the table 10 points, up to a max of 100.
    At 6+ entries, 1st gets a 2x multiplier, 2nd a 1.5x multiplier. At 4-5 entires, 1st gets a 1.5x multiplier.

    Points are awarded as a fraction of the number of entries, up to 10th. First place gets the full table value, 2nd gets (n-1)/n*(table value) points where n is number of entries and capped at 10, 3rd (n-2)/n*(table value) and so on.

    Map size would be a end point multiplier of:
    • Duel:55%
    • Tiny:70%
    • Small:85%
    • Standard+:100%

    Difficulty would be:
    • Settler:1
    • Chieftan:1.5
    • Warlord:2
    • Prince:2.5
    • King:3
    • Emperor:3.5
    • Immortal:4
    • Deity:4.5

    Here is one and another example tables based on time finish:
    Column 1 Column 2 Column 3 Column 4
    Table 1 Table 2
    Player Points Player Points
    vexing 540 HideInLight 340
    moriarte 337.5 xger 229.5
    glory7 180 moriarte 136
    HideInLight 135.5 kirbdog 119
    Bram 90 browd 102
    Bignitas 45 Jinbe 85
    FoamFollower 68
    jamgdz55 51
    titantoma 34
    Dignitas 17
    11th+ Place 0


    Please keep in mind once again, the constants are just chosen at the moment so there are concrete examples. I'm attempting to accomadate as much as possible as there are people who want things exactly the opposite, so please keep in mind this is supposed to be a system for everyone, not just one person or one playstyle.
     
  17. sanabas

    sanabas Psycho Bunny Hall of Fame Staff

    Joined:
    Nov 24, 2004
    Messages:
    4,269
    Location:
    Canberra, Australia
    You've used 9, not 4.5, for the deity table.
     
  18. Xger

    Xger Prince

    Joined:
    Mar 22, 2010
    Messages:
    510
    Location:
    Denver
    That's what I get for rushing it out before leaving work, I had a stray *2 in my calculations :blush:
     
  19. sanabas

    sanabas Psycho Bunny Hall of Fame Staff

    Joined:
    Nov 24, 2004
    Messages:
    4,269
    Location:
    Canberra, Australia
    OK, here's my take. I think it's easier to automate, as everything would run off one formula, rather than have clauses like 'if there's more than x games, multiply first place by 2' or 'use this formula, but cap it at 10 entries'.

    A finished game = 100 points, before adding the various multipliers.

    Map size:

    Duel: 0.4
    Tiny: 0.6
    Small: 0.8
    Std, Lge, Huge: 1.0 - could make large & huge 1.1 & 1.2

    Difficulty:

    Settler: 1.0
    Chief: 1.2
    Warlord: 1.4
    Prince: 1.7
    King: 2.0
    Emp: 2.5
    Immortal: 3.0
    Deity: 4.0 - I don't think there's a lot of difference on the lower levels, which is why the multipliers are closer together.

    Speed:
    Quick: 0.8
    Normal, Epic, Marathon: 1.0

    Number of games in table (n):
    log (n+1)

    So a single entry in a table would have a multiplier of 0.3. 3 entries in a table, it would be 0.6. 9 entries in a table, it's 1.0. Really full table, 25 entries, would still only be 1.4. To get the multiplier to 2.0, you'd need 99 entries. So topping a competitive table, 7 games, would be 3 times as valuable as a single game. I don't think there's that much difference between topping a 7 entry table or a 15 entry table, and it'd be 0.9 vs 1.2 for those. Reward for filling an empty table, big increase to reward as an empty table gets competitive, not much extra difference as a competitive table gets a few extra entries. And no need for a manual cap.

    So to me, that gives an incentive to play standard+ size games, an incentive to play the harder difficulties, and will help fill the tables. An all standard deity game on an empty table would give you 120.4 points. A cheesy settler duel marathon game that takes 10 minutes would only get 12 points on an empty table, 40 points if you're the best of 9 entries. http://hof.civfanatics.net/civ5/tab...e=10&leader=27&mapSize=4&speed=2&dtSc=0&exp=1 is the most competitive Emperor table, first place there would currently get 250 points.


    For games that don't win, I think it should be based on how fast you are compared to 1st, rather than simply 10 for 1st, 9 for 2nd, 8 for 3rd, etc. But I agree with Vexing there should be more difference for being 20 turns quicker (say 180 v 200), rather than the 200 turn game getting 90% of the winner because the winner is only 10% quicker. So the formula I'd use would be:

    Quick:
    15/(15 + your turns - winning turns)

    Std:
    20/(20 + your turns - winning turns)

    Epic:
    25/(25 + your turns - winning turns)

    Marathon:
    30/(30 + your turns - winning turns)

    So, on a standard game, if you finish 20 turns behind the winner, you'll get 1/2 of their points. Finish 40 turns behind the winner, you'll get 1/3 of their points. Finish 100 turns off, and you get 1/6 of their points. Also an incentive for players to get the best result they can, even if they can't get to 1st on a particular table.

    Score:
    Really, the simplest way to do this would be to acknowledge score is irrelevant for most games, and only look at it for time victories. But as the current setup gives medals for scores, then I'd just use (your score/winner's score).

    So every game would generate a finish time points total, and a score points total. If you win using the actual time VC, then that would generate just the one score [(yours/winners) x the relevant multipliers] but use that for both finish time points and score points.


    Working out the overall total:
    Simply rewarding volume is a bad idea, I think. Treating score points and finsih time points separately is a good idea. So for each of these, I'd do the same thing for finish time points & score points, to generate a time total and a score total.

    League of nations: Best 2 games for each nation qualify, sum those to get your total points. Max of 68 qualified games.

    Map quest: Best 3 games for each map, sum those for total. Max of 60 games.

    Tempi: Best 15 for each speed, sum for total. Max of 60.

    Go the distance: Best 10 for each size, sum for total. Max of 60.

    Machiavelli: Best 12 for each VC, sum for total. Max of 60.

    Inferno: Best 8 for each difficulty, sum for total. Max of 64.

    Sum all 6 totals to get overall VVV fastest time points, likewise to get overall VVV score points. Sum the two to get overall VVV points.

    So volume will help, but only up to a certain point. And it will help much more if the volume consists of standard+ sized deity games with every civ, rather than just as many duel games as you can crank out.

    I think that addresses most of the stuff, and would be a simple formula that could get run on every table with a new entry when an update gets published. It'd look somewhat confusing for new players, but I think be understandable enough in practice. Play higher difficulties, standard+ maps, get more points. Get close to the top times on competitive tables = more points. Fill an empty table, get 30.1 x size/speed/diff multipliers.
     
  20. vexing

    vexing knows

    Joined:
    Dec 24, 2010
    Messages:
    2,668
    sql supports inline case statements: case when numgames > n then y else z end

    anyway, re your formula i fear things where the total is based on a specific number of games. in your system there will eventually be a set of 68 games that reward the most points and if you place well in those, there will be no vvv gain in submitting other games.
     

Share This Page