Cleaning up the rankings

[wussy]

[acerimmer]

Hi Wussy

I wasn't actually referring to you my friend. You are one of the top players who DOES actually play.
The players I were referring to are the guys who can't even manage to play 3 ranked games in a 2 year period........not you.

Indeed we have played a few games......and most entertaining and thought provoking games they were.

We should play again soon I hope.



(Wussy was another member of the winning European team who won the Stratego Ryder Cup for Europe. All the players in this team......I have the up most respect for.......they ALL agreed to drop everything and play for Europe in this tournament. This is indeed a quality I hold in high regard.)

[OGETARTS]

Wussy/Luc.......you're alive!!!

After you play Ace, I would like to play a game with you if you have time. We've played maybe 3 or 4 times in the past, but that was a long, long time ago. I'm glad you still have the motivation (and fire) to play stratego. Good to see you online again.

OGE

[unbiasbob]

Wussy, you're a class act, dude

[unbiasbob]

[acerimmer]

I don't know if this english phrase is known in america......but basically it means.........it is the best...the king. The dogs boll**ks of anything......is the best.

[unbiasbob]

I think that Ace's idea to put dormant accounts at the bottom of the page is a good one. I also have an idea of what to do with players who barely play enough games to keep their rank status. I'm talking about the guys who have played barely the minimum games at this point in the year. What you can do is have a separate division for these guys and have a name for it, such as, the Twinkle Toes Division. This way these guys still get the recognition they deserve. I got the idea today when I saw that dirac stopped by and played a game. This gives him 20 games played on the year. I'm thinking that he could be the poster boy of this new division

[sadistic]

Yes, there should be a category for "pretty boys" or "posers", for those who like to look at their names but are scared sh!tless to get their hands dirty from a few losses. Those ballerinas, who avoid real competition and quit playing when their rank is high, need to be dealt with. Does this mean I should fire up my electroshock device? I think so, beloved fans. In the end, we might have a few stumbling "has beens" with nervous head twitches and other involuntary disorders but such posers are expendable anyway. Why not use them for our own entertainment, through Sadistic therapy techniques? Those of you who would question my methods, have never experienced a good "zapping". Trust me, you will laugh your a$$es off. Who's first?

[sir_dagonet]

Back to the earlier points raised in this thread, I agree that making dead inactive accounts is a good idea. Personally speaking though, I can HONESTLY say that I ONLY check 2007 Kleier rankings (so inactive accounts don't bother me that much) I am not a fan of the WSC rankings as an overall rank (I see them merely as a form guide). As for dropping Kleier - I think that’s ludicrous, the guys near the top are there as they beat other guys near the top - end of! For those that consult WSC, fine (your choice) but why drop a ranking system that many are happy with and that is highly accurate?

[unbiasbob]

Is there a post somewhere that outlines the differences btw the 2 rating systems? I have a general idea but not 100%

[unbiasbob]

I never checked wsc before this year. I think it was edfx who pointed out the system to me. He was all excited cause he was in the top 15 of that one for a couple of hours. He has since dropped out of sight. He was saying that it's more of a "what have you done lately" system than the Kleiers, weighing more recent games heavier. I dont think this is a bad thing as players do improve as they play more games (generally)

Here are 2 examples where the all time Kleier system fails miseably to gauge a players ability. Take Cleo and Pieke. Pieke is rated 1183 on that but he is far better as the 2007 ratings show. Cleo is a mere 1342 but a top 10er for 2007. Herein lies the fault with at least the all time Kleier system and perhaps the system as a whole.

[darthremark]

Hi Bob,

I searched Gravon and couldn’t find anything documented about WSC, so here’s the explanation:

The World Stratego Club formula was refined by Lonello (a great Dutch player and founding member of the WSC) from a formula used on the old Fairhousing site that MSN Gaming Zone players used from 99-01. It's not as accurate as Kleier, but does tend to be a reasonable gauge of relative strength, and is funner for competition. The exciting advantage is that it has built-in mechanisms allowing (ensuring actually) point swings. It also guarantees you at least a point for a win even if it's against an 800 player. Its disadvantages are "point creep" that over time reward those who play more games, and it allows better players to park to an extent. If you choose your games properly you can still catch them though.

I went through my old files and I’ve apparently deleted the documentation detailing Lonello’s adjustments. Maybe Spion could publish them on this site somewhere? I did find the original verbiage describing the Fairhousing formula from which WSC is derived. I’ll explain the modifications as I remember them afterwards.

=======
stratego club skill how to
this is the standard equation : loser's skill - winner's skill = a
a / 400 = b
b +1 x 14 = c
c is the value of the game and will be added to the winner's skill and subtracted from the loser's.
However, to figure skill levels as a result of a game is a little more indepth depending on the difference between the respective players skill levels before their battle began.
if the difference in skill is <= 50 the +1 in #3 (b +1 x 14 = c) above is changed to a 1.5 this allows for high impact close combat between players with close skill levels.
if the difference is 51 to 125 the standard equation above applies.
if the difference is 126 to 250 the +1 in #3 above is change to .75 for the loser only. this gives higher skilled players more of an incentive to player lower skill players without losing their shirt if the lower skill player happens to wins. it also decreases the impact of the lose on the lower skilled player in the most likely event of them being defeated.
if the difference is 251+ the +1 in the first #3 above is change to .40 for the same reasons as stated above.
The above formulas don't apply to draws. Both players involved in a draw receive 1 skill pt.
=======

Fairhousings formula has problems over time:
1) It can be difficult to advance without getting the “proper” matchups. If enough players are competing though this usually isn’t a problem.
2) Particularly nasty, it's possible for a high rater to lose points with a win over a vastly lower rated player.
3) It doesn’t tend to hurt a high rater very much if he loses to a significantly lower rated opponent.
4) It does not factor averages (specifically).

Lonello corrected 2 and 3 for the WSC formula. You will get at least +1 for beating anyone. If I remember correctly (and I might not ) if a low rater beats a significantly higher rater he can gain upwards of 25 points, whereas the high raters potential loss is capped at 15. The big action involves opponents within 50 rating points of each other. If you are right on the edge (exactly 50 points apart) you can get +25 for the win while sending the other guy back 25.

There are two secrets for leveraging the WSC formula:
1) To advance: Target opponents who are within 50 rating points of you. This gives you the biggest point boost upward. (It also sends you correspondingly backward if you lose).
2) To protect your rating: Don’t play anyone within 50 points of you, or anyone several hundred points below you.

As I said, it’s a very exciting formula for competition since there is real movement. If you follow the logic of the equation you can see what I’m talking about. The key with WSC is to play opponents with a similar rating. Over time this bubbles you to the top. And if you can string together some key victories you can bubble pretty quickly. It’s not as statistically accurate as Kleiers, but it does a decent job charting relative strength, and it’s very fun to compete with.

[unbiasbob]

Darth, great explanation, thanks!

[stratego]

hello,

maybe some should read the infos about the kleier rating.
in the end its a modern form from elo (chess)

if you read carefully, you know why the rating is fair and shows
a good result.
there is only one problem. the online games are "to fast".
old games count to long for the ranking.
wsc doesnt work (and didnt work in past.)

http://www.hermann-kleier.de/txt/rating.html

the context is in english - no german text.

stratego

[darthremark]

Hi SM,

I would counter that WSC does work as a fun system for competition. It's basically a point accumulation contest. It's not going to produce a statistically precise result like Kleier, but it does tend to be a reasonable reflection of player strength to within 10-20 positions. The players have to understand how it works and compete with that knowledge in hand. It would be a good idea to publish info about it somewhere on the site.

WSC cares only about the current rating of the two players, whereas Kleier attempts to reconcile the entire body of loosely related game results. Kleier does a good job at that, but doesn't really do much better at producing the true list of how players stack out at a given time.

Look at Vauban as an example. He has a rating of 1849, but has played only one player over 1500, and his average oppenent has a rating of 1383!

Achilles is another. He has a rating of 1189 with 2683 games played. If he were struck by lightining today and emerged as the best player the Stat world had seen he would have to play until he was an old man before Kleier recognized it, whereas he would probably go to the top of WSC very quickly.

Both formulas have their uses. Neither is completely accurate. I don't think complete accuracy is attainable from a formula.