Post Rank Change Debate

Discussion in 'Seasonal Rugby' started by HeathDavisSpeed, May 29, 2013.

  1. HeathDavisSpeed HT Davis

    Right, so I'm happy with the sim as it is. However, there is one major problem. That is that players at low post ranks are totally crap meaning that reserves games are almost farcical in their competence.

    I have two ways to change this:

    Linear

    Shift the post rank base (set at 250 players at the moment) to a higher number (say, 350).

    Pros: Low post rank players won't be totally crap
    Cons: It will disproportionately improve crap players and massively narrow the gap between the good and the bad.

    Non-Linear

    Use a logarithmic function to make the post ranks to skills relationship non-linear and then shift the post rank base

    Pros: Low post rank players won't be totally crap, the best players will remain the best.
    Cons: The skill gap between players will be much more significant at low post ranks (i.e. I'll be even better, comparatively) and players at high post ranks will see very little difference in skills based on the post rank element.
    I also need to find a non-linear function that doesn't basically flat line really quickly meaning the gap at the top post counts is huge and then quickly becomes negligible.

    Thoughts?
     
    Last edited: May 29, 2013
  2. loganb JEM Logan

    Without thinking too much, I'd have thought Linear would prove for a better competition.
     
  3. morgieb MC Burridge

    I kinda like the non-linear idea.
     
  4. Arheiner SIA Yates

    Non-linear is probably better for encouraging activity and having a realistic balance, so I'd say that.
     
  5. Benny BS Read

    I think you need to ensure that the players who've followed this strongly for multiple seasons are still superior to new players. It would be unrealistic if your test/rep players were not substantially better than rookies.

    I reckon you could come up with a way to support active rookies by determining post count assist by a system which tracks posts per season in the game, you could make the posts from older seasons be less significant but still effect the skills so you're established players get assisted but you wouldn't hax out new signups who are shwoing interest?
     
  6. Himannv LV Himann

    Keep it Non-linear but introduce a different factor like activity to give new players who are interested in the site some sort of incentive.
     
  7. Jabba HJ Bots

    Non linear rewards players who give a shit more
     
  8. Escath LE Schaw

    Piecewise, linear up to a certain point, non-linear closer to the top.
     
  9. HeathDavisSpeed HT Davis

    Yes, this is my biggest concern. I still want the good players to be good, but I also want the seconds to be of a reasonable standard. At the moment, we've got Test quality players at the top and club 5th grade standard playing 2nds.

    I did ask Cribb if he could set up a 2nd post rank page for the start of the season so that I could track current season activity and make overall post rank a combination of the two. So, that might be a possibility, but it means Cribb needs the time to do it in the next week or so.
     
  10. azzal AJ O'Neill

    Non linear sounds better.
     
  11. Benny BS Read

    Yeah exactly Heef. I spent 5 mins or so trying to think of a viable way to split signups of different age/experience and the reason I suggested SWPR is because it would allow you to tailor a player ratings to his performances/posts since they'd regsistered that way and it makes it more player specific than group specific if that makes sense - probably not:p

    Could for example come up with a number to factor towards skill for anyone that way.

    i.e Player A signed up in season 1 - track posts per season since registration and give slightly higher weighting t early seasons. Average all 5 seasons worth of date for said player and use the result to affect exisiting skill

    Player B signs up now - result comes from just this seasons posts and the current season being weighted slightly lower than early one will ensure rookies are less skilled than established players and keeps it in an objective framework
     
  12. Frizzed WC Welker

    Non-Linear seems the better option, however neither option is great.
     
  13. HeathDavisSpeed HT Davis

    What's the alternative?
     
  14. azzal AJ O'Neill

    The way it is atm?
     
  15. HeathDavisSpeed HT Davis

    At the moment, it's linear but with a finite base of a max. of 250 and there's now more than 250 players in the database for CricSim alone. So, it's no longer workable as it is - hence the need for a slight change.

    Changing the finite base will have a much more significant impact than changing from a linear to non-linear method as the finite base can remain the same as now using a non-linear method.
     
  16. maximin DJ Edwards

    Non linear sounds better imo
     
  17. Furball G Furball

    Whatever makes me perform as I should :p
     
  18. HeathDavisSpeed HT Davis

    You already are...
     
  19. Hurricane JD Hurricane

    Put the guys in 2nds on their own ranking scale.

    Ie you are ranked 280th overall - but 87th of the guys playing seconds.

    It means someone will get a number 1 ranking each week for playing seconds. But it means the standard of play in 2nds will be identical to playing 1sts. Which is not a bad thing except you won't be able to judge how good a player is going properly in 2nds in order to bump him up to 1sts. That is the only draw back.
     
  20. Weeman27bob BE Force

    I'd definitely be in favour of non-linear, but I'm happy for whatever.
     

Share This Page