SpielByWeb

tripitaka · Joined: 07 Sep 11 Posts: 37

I am sure this would be far too much work for very little gain, but if possible, it would be interesting to have an opportunity to see what your stats look like, head to head.

For example, a person may have won 30 games, 10 against, player1, 5 against player2, 7 against player3, 8 against player4. Now i know you can just look at the games completed page and work this out but a summary of that would be interesting.

Then an even crazier thought is a ranking based on the quality of the player you play. If you beat a newbie you score less than if you beat a seasoned player. something like that.

Anyway, just a thought

Trip

JokeToy · Joined: 12 Dec 10 Posts: 151 Location: Sheepscott, Maine

yes, it would be nice to have stats like that but it seems better for mikael to spend his "spare" time implementing new games to keep the site fresh.
_________________
thanks for reading this far, JokeToy

tripitaka · Joined: 07 Sep 11 Posts: 37

yep, for sure JokeToy

i didn't realize one individual designed, managed, (and all the rest) this site. that is amazing! i am very grateful. consider my request retracted!

JokeToy · Joined: 12 Dec 10 Posts: 151 Location: Sheepscott, Maine

yes, it is amazing what mikael has done. on top of everything else, he manages all the licensing issues too.

i strongly urge anyone who plays on SBW regularly to make a donation (you'll find a link on the "Your Current Games" page). donations are strictly capped to meet expenses. for all the enjoyment we get out of this great site, it behooves us to peel off a bill or two to help mikael keep it running Smile

_________________
thanks for reading this far, JokeToy

faticus · Joined: 14 Nov 11 Posts: 42

Wow, my hat goes off to the (solo?!?) developer of the site.

Consider the following "quirk" with rank by # of wins.

In short, a player can be ranked higher if s/he only plays head to head. For example, say John only plays head to head and wins 50%. Jane only plays 5 player games and wins 49%. John is ranked higher by number of wins, but Jane is probably the better player. Jane beats 4 other players every time she wins, but John only beats 1 player every time he wins.

A "fix" is to consider number of wins and losses against all opponents. So if John wins a 2 player game, he gets 1 opponent win and 0 opponent loss. However, if he finishes 2nd place in a 5 player game, he gets 3 opponent wins and 1 loss. The ratio of opponent wins to total opponents is arguably a better measure of skill.

Anyone else agree? Again, props to mikael for making an AMAZING system.

Saul Pum · Posted: Sun Nov 20, 2011 8:50 am Post subject:

A simple change that could be added is extend the number of miinimum games for winning percentage. For instance, minimum 500 games, right now the cap is 100 games. This should take very little time. Can also be ridiculous and go to 1000 for the long term.

Head to head percentage would be cool too.

faticus · Joined: 14 Nov 11 Posts: 42

If there was infinite time, the following seems like a great solution. It's the trueskill(tm) ranking system used for multiplayer xbox games.

research.microsoft.com/en-us/projects/trueskill/details.aspx

Alaina15 · Joined: 31 Oct 17 Posts: 11

Like this post & found interesting

Flores · Joined: 23 Dec 17 Posts: 12

Sobhana123 · Joined: 26 Jan 18 Posts: 45

hii. Thanks for sharing this topic it is very good and nice information .

Reek · Joined: 18 Apr 18 Posts: 16

Thanks