@max334 Scrambles?
G40 League House Rule project
-
JWW I will show you the path of pure aggression. :)
-
@Cow:
JWW I will show you the path of pure aggression. :)
I look forward to it. When can we begin? Show me the light. I actually had a question for you concerning some input you gave in a thread about blocking using ss in naval situations……
-
@Gamerman-
I’ve been meaning to ask a question for a while… in one of your previous posts here, you say something along the lines of
your PPG might change depending on if your opponents change in their own rankings. Does that mean that generating the list after adding in a new game is an iterative process, so you keep running some algorithm until everyone’s rankings are stable? If so, is it known mathematically that stability is reachable, or do you have to have some kind of cutoff in the case of players who keep swapping rankings each time the algorithm runs? -
It’s not reciprocal, and it’s all manual.
Your PPG is purely a result of how many games you won and lost against tier 1, 2, or 3 players. For example, if you beat a tier 2 player and lost to a tier 1 player, you would have 2-1=1 point, or .5 PPG
When a player crosses into a different tier, I basically go to each of their past opponents, and add or take away 1 point per game played, which ever the case may be.
So in the above example, if the tier 2 player who was defeated by our sample player were to fall to tier 3, then I would take away 1 point (because it turns out he beat a tier 3 player, not a tier 2 player), so his PPG would be 1-1=0, or 0 PPG.
This way, if you beat a player who is new to the league and I award you 2 points, but it becomes obvious a couple months later that this player is a tier 1 player, you get credit for beating a tier 1 player because that’s what he actually was - not a tier 2 or 3 player.
Clear as mud?
-
This way, if you beat a player who is new to the league and I award you 2 points, but it becomes obvious a couple months later that this player is a tier 1 player, you get credit for beating a tier 1 player because that’s what he actually was - not a tier 2 or 3 player.
So basically it isn’t iterative- if your adjustments change the tiers of a different set of players, you don’t make any additional evaluations.
-
Not sure exactly what you mean by “different set of players”, so I will try to clarify by saying:
When a player wins or loses and it causes them to change tiers,
I go to each of their past opponents and adjust the opponents’ point totals up or down to reflect the increased or decreased status of the player who changed tiers.
-
Note that your point rankings will change not only when you complete a game, but they could change up or down if you or your past opponents move up or down tiers.
From my original explanatory post (reply #1 of this thread)
-
But I later realized after posting that original post, that I shouldn’t have said “if you or your opponent change tiers”. It should only be “your opponent”. When a player himself changes tiers, that event has no further effect on his own points.
-
Updated
-
Not sure exactly what you mean by “different set of players”, so I will try to clarify by saying:
When a player wins or loses and it causes them to change tiers,
I go to each of their past opponents and adjust the opponents’ point totals up or down to reflect the increased or decreased status of the player who changed tiers.
I think the Eggman’s confusion is coming from the fact that he is not aware how the tier of a player is determined. It actually took me also quite a while to find that information. Gamerman, maybe you should include it somehow visibly in your sheet.
So the tier of a player is not based on his PPG but rather on his % of game won. Eggman is right that if PPG were used to determine the tier of a player then an iterative process/algorithm would be required to update the rankings. So instead Gamerman has a system in which the tier of a player is not based on his PPG but only on his % of game won.
The current tier system might not be as accurate… an example: a strong player playing mostly against other strong players ~ his % of games won should be around 50% ~ thus he would be Tier 2. On the other hand, a week player playing only novices… as he does know he can’t challenge the top players ~ he can easily have % of games won over 66% and thus would be considered Tier1. But I think in practice Gamersman’s system works quite OK.
I just wish the official ladder ranking would change… right now all you need to qualify to the final of the ladder is: avoid playing strong players and win over 5 novices. This does not feel correct.
-
I am not quite so sure that is how’s it’s done as there are 2 players below your % number that are still tier 1, and 1 player above that is tier 2.
-
I am not quite so sure that is how’s it’s done as there are 2 players below your % number that are still tier 1, and 1 player above that is tier 2.
I think Gamerman uses victory percentage from all games played by a player also those from previous years. That would explain your observation.
-
Thanks a lot, Darth.
That’s right. This thread is just for fun, and has no bearing on League regular season or playoff play.
These rankings are unofficial, and subjective.1) Regular season standings with W/L records will be posted regularly here. These should always match JWW’s, but I plan to post much more frequently.
2) I will regularly upload my Excel spreadsheet of the matrix of results, so players can see who has played who, and who won the matchups, at a glance.
3) I will post my subjective point rankings, which are designed to reflect strength of opponents. (Because players can choose to play who-ever, and there is no schedule)Here is the methodology for the point rankings:
1) I will consider the 2012, 2011 (and possibly 2010) league standings in determining quality of wins/losses from 2012 league results.
2) I will divide all league participants into 3 tiers, based on past records (2012, 2011, and possibly 2010)
3) I will award points for wins. 3 points against top tier, 2 points against mid-tier, and 1 point against lower tier.
I will deduct points for losses. 1 point against top tier, 2 points against mid-tier, and 3 points against lower tier.Note that your point rankings will change not only when you complete a game, but they could change up or down if you or your past opponents move up or down tiers.
I welcome comments on the standings as they are posted throughout the year. I will be posting the point rankings and spreadsheet for the 2011 season when it is finalized here in the next couple of days, for your enjoyment and/or curiosity. For the sake of consistency, I do not plan to change the point ranking system during the year, but will consider suggestions or comments for improvement for subsequent years.
To reiterate, this ranking thread is for fun, and nothing else. Enjoy.
Hi nerquen, haven’t encountered you before. I appreciate your interest in my rankings process.
I totally ignore win % when doing tiers.
Tiers are GENERALLY determined by PPG. .50 PPG and above is tier 1, -.50 PPG and below is tier 3.There are exceptions. It is subjective and judgmental. I may consider previous year’s results, or non-league results (tournament progress, etc). This is generally only for players with few league results. If a player has a lot of league results, I will go by .5 and -.5 cutoffs.
If you think a player is in the wrong tier, kindly offer that suggestion and the justification for your different perspective and I will consider it. What tier you are in actually has NO EFFECT on your own ranking or PPG. It only affects the players who have played against you.
Hope this helps
-
I just wish the official ladder ranking would change… right now all you need to qualify to the final of the ladder is: avoid playing strong players and win over 5 novices. This does not feel correct.
I agree, and have expressed similar sentiments on the boards before (see G40 league discussion thread).
When this year’s league rules were established, it was in light of very limited league activity in 2012. Only a few players even qualified for the “playoff” game in 2012, so the bar wasn’t set very high for 2013. Then interest in G40 exploded and now the playoff system idea seems ridiculous.
The moderator primarily in charge of the G40 league has been away from the boards for some time. I am even playing a non-league game with the moderator and we’re not out of round 1 yet. Perhaps the playoff rules for 2013 will be tweaked to reflect the hundreds of games that are being played this year - and perhaps not. We don’t have the person in charge around to talk to at this time
-
Ranking shmanking. I think the league is useful for another reason. I am finding that I enjoy the games against someone who is about equal to me in skill and the league rankings let me find them. Playing a really raw new player is not much fun because its too easy, although sometimes they can surprise you with oddball stuff and get lucky with it. Playing the tier 1 guys is definitely no fun for me, just stressful. My own skill level isn’t very advanced but I think that just about anyone who has played online against a variety of styles is actually pretty good even among the lower ranks. With the league ranking you can go through the list and maybe find an opponent who is about an equal match and then have a really good game that you both learn from. I think the ranking system you have right now is fine for that purpose. Let the guys at the top of the list worry about rankings and those of us down in the pack can focus on picking out equalish opponents to have good games with that are challenging but not ridiculously impossible. I myself count anyone in the tier 1 as impossible to beat but that’s OK.
-
Very well said, Variance. Thanks for the post - hopefully others will read it and see one of the most useful purposes of my spreadsheet, and that is:
Finding fun games to play.
I also think it’s too stressful to play tier 1 players, most of the time. :-)
I’d rather kick butt than learn lessons the hard way :-D -
Updated
-
Updated
-
Updated
-
Updated
Suggested Topics
