CTP PBEM Ratings (01-AUG-2001)

[Max Webster]

Wow. Lots to read since I visited last. I appreciate Quinn’s efforts and I think he is doing a good job. I realize now Quinn that my comment was kind of bogus, as you say ratings do not take into account diplomacy, which can have an effect on a player’s position in a game.
Excuse me but I sucked back one beer too many (lung would be proud) and was just spewing. I usually try and keep my mouth shut and not attract too much attention.

One thing though about the ratings, shouldn't the number of games a person is involved with effect the persons standings? Is this possible or necessary. Just think its kind of weird with one person doing well in one game being at the top of the ratings.

Btw. no you don't need to knock off points, I kind of like seeing my name up there. But it sounds like BM is determined to see that it doesn't stay up there for long.

[jpww]

Even though I hate everybody, I love Quinn's rating system

Keep up the good work

[Lung]

Don, i would be proud of you, but having a technicolour yawn is going too far!! Unless of couse you redeem yourself and have another beer (Technicolour yawn - spew/chuck/vomit/ralph/hurl/chunder [did you ever wonder what that line from Men At Work's Down Under meant? well, now you know!] )

Also, i totally agree with you about you-know-who. If she's so damned good, why won't she take on the rest of us?!? Between Mobius' propaganda and environmental vandalism, Tweety's "playing for both sides", and Lung's bad, bad ways, i'm sure we could put Klair in her place! After all, she's only a girl

World famous? I doubt it, but i'll settle for notorious

[quinns]

Thanks Max and JPWW! ... and Lung ... thanks for .... umm... erm ... keeping this thread lively!! . (It took me awhile to think of something to thank Lung for!)

Max brings up a good point about why the win/loss column doesn't reflect directly in a player's rating. Other players have asked the same question. I have tried to explain this several times... but here goes again, (I'm not a very good 'splainer!)

Using the number of Power Graph wins, to determine a rating, COULD BE misleading if someone wanted to abuse it. For example, lets say someone wanted to boost their win/loss record to get a better rating, (IF the ratings were calculated based upon win/loss, that is). Let's say this person was a fair player, but not great. This person could start 5 games against 20 VERY WEAK players. He or she would have the highest power graph rating in all of the games, and would be reporting about 8 to 12 POWER GRAPH WINS PER WEEK (with no losses). Now, I think most of you would agree that this person's rating should go up somewhat, but not that much, as all of his or her wins were against very weak players. The CTP Ratings system does not reward this type of "multi-wins-against-weak-players" tactic, but would only reflect a minor increase in rating for the wins against the weaker players.

Another reason for the win/loss record not always correlating with the rating, DEPENDS HEAVILY UPON THE PLAYER'S MOST RECENT RESULTS. That is, if someone has a lot of wins in the win/loss column, but loses 5 or 6 times in a row, that person's rating will drop fast. The reverse is true as well. If someone has a poor record but wins 5 or 6 times in a row, they may pass someone with a better win/loss record.

Does that help?

Regarding Klair. I do wish she would join another game to get some more depth, or at least prove that the "one game" lead isn't a fluke. But she is playing according to the CTP Ratings rules. And according to those rules, she is in first place. From what I hear though, she really IS, a very good player! Klair? Are you there? DO YOU WISH TO COMMENT!!!!!!?????? HELLO!!!!!!!!!!!!!!!!!!!!!

[Lung]

Quinns, your perception of apathy amongst CTP pbem players was wide of the mark. In fact, your ratings are far more popular than you ever realised.

I just realised that your ratings system had influenced my gameplay! In Lung's FAST! pbem, i held off finishing off St. Jon for one turn because of the turn number! This is because we were up to turn 208, and i had the opportunity to finish off Jon. However, in a rare bout of misplaced mercy, i held off, knowing that for St. Jon to survive another turn would give me a bonus win against him. Then, come turn 210, i kill him!! This then gives me 3 more kills against him, as he automatically loses 3 times to every player!

As it turned out, Stavros came out of the blue and hammered the final nail into St. Jon's coffin!!! DEPRIVED!!! At least someone killed him off

Omigod! They killed St. Jon!

[TheBirdMan]

Stavros did it ????

Lung - do you really mean, that Stavros did it???? Whowww man, I didn't expect the german to be that strong

[blackice]

Does that help?

No not me heh he commented on number of games played not win/loss big differance.

Posted by Max: One thing though about the ratings, shouldn't the number of games a person is involved with effect the persons standings? Is this possible or necessary. Just think its kind of weird with one person doing well in one game being at the top of the ratings.

The system is not obviously win/loss his point was it disturbs him to see a one game player being on the top of the rankings.

Should there not be a conssesion for participaction...after all there is a penalty for removing ones self from a game or two...

I think that more describes what he was asking...

Beging to ask the question if one can not continue for what ever reason then can, does the penalty affect there ranking? Is there a time limit? Or do they start from square one as a beginer?

[blackice]

Another reason for the win/loss record not always correlating with the rating, DEPENDS HEAVILY UPON THE PLAYER'S MOST RECENT RESULTS. That is, if someone has a lot of wins in the win/loss column, but loses 5 or 6 times in a row, that person's rating will drop fast. The reverse is true as well. If someone has a poor record but wins 5 or 6 times in a row, they may pass someone with a better win/loss record.

Humm I will point out here that you can have five wins and one loss and the one loss will be a greater influence then the five wins... several examples of this exsist... truely the same must be true in the exact opposit?

[quinns]

To convince Blackice of the validity of the CTP Rating system --- This may take an act of God!

Ice, the win/loss column DOES show how many games a player has played... just add the wins plus the losses together. I wish you would please read the formulas and understand them before commenting. I don't mind the comments... really, I don't ... I just get kind of tired of explaining the same thing over and over again. The same thing (described above) applies to the total number of games played. Someone could again abuse this "proposed" MORE GAMES BONUS rule, and get a very high rating just because they have loads of extra time on their hands to play many games... and not because of their ability. We discussed this at great length when we first started this rating system.

And once again, (and again and again and again and again) THE MAXIMUM NUMBER OF POINTS A PLAYER CAN DROP AGAINST ONE PLAYER IS ONE-HALF OF A RATINGS POINT. THE MAXIMUM NUMBER OF POINTS A PLAYER CAN GAIN AGAINST ONE PLAYER IS ONE-HALF OF A RATINGS POINT.

Therefore, YES, if you play 5 very weak players and defeat them all you might gain 0.2 points. Then if you lost to a player that was about your equal, you would lose 0.25 points... and YES you would then have A LOWER RATING FOR YOUR FIVE WINS AND ONE LOSS. Why is this so hard to understand? It would be like Pete Sampras playing some school boys in tennis and defeating all 20 of them. Then losing one match to Gustavo Kuerten and then saying, "Hey, that's not fair!! I beat all of those elementary school boys... TWENTY OF THE THEM... and only LOST ONCE to Gustavo ... ONLY ONCE ... WHY DID MY RATING GO DOWN??!!"

Why can't I get this concept across to you Blackice? Am I really that bad in explaining how this works? Is anybody else out there confused about this subject?

P.S. Oh, and thanks for the good words Lung!

[blackice]

It would be like Pete Sampras playing some school boys in tennis and defeating all 20 of them. Then losing one match to Gustavo Kuerten and then saying, "Hey, that's not fair!! I beat all of those elementary school boys... TWENTY OF THE THEM... and only LOST ONCE to Gustavo ... ONLY ONCE ... WHY DID MY RATING GO DOWN??!!"

Obviously then it would be an advantage (smart for your ranking) only playing Gustavo as the other games 20/1 in this case just become meaningless for rankings anyways as per Klair's example.

Or fun as in mine playing a wide variety of ranked players without suffering the yoyo effect

I do get it Quinn maybe better than you think as with most all systems it promotes selective play and a hierachry.

Now what about the penalty 420 as an example if, when he comes back does he start at square one? Is there a time limit for him? He has been removed from the rankings that would suggest he would start at square one?

Just think its kind of weird with one person doing well in one game being at the top of the ratings.

Maybe not for long though

Any rate lets not muck this thread up it is becomimg hard to follow the game win reports. Another thread has been started to expand upon ranking strategy...not the validity of it. That has been covered again and again and again. It is pointless and a waste of threads.

Good job Quinn BTW keep up the good work

[quinns]

I think I see what you are saying now, Blackice. Maybe I was a little harsh in my EXPLANATION . Thanks for the comments.

Regarding quitting and then returning.... that is a good question. Actually, Blackice is a better case in point, because his rating is now in the 15's (even though he is unrated), due to multiple changes from rated to unrated status. And Blackice is probably more like a 22 or 23 player rather than a 15. That's a good question. If we let Blackice come back in at 21 (deity initial rating) then some players that have beaten deity, but are now less than 21, would just quit and then return to get their rating of 21 back. You see the problem. Quitting really hurts the system, but there should be some fair way to do it. Any suggestions?

[Lung]

Lung's FAST! pbem powergraph - turn 209

Lung - Australians
Stavros - Germans
TheBirdMan - Romans
Paul - Dutch
St. Jon - DEAD Indians

St. Jon seems to have committed Hara Kiri and eliminated himself from Lung's Fast pbem, although no one's owning up to it

[Gavrushka]

Quinns just a thought for quitting players....... IF a player quits then their finishing rating is recorded, and when they decide to rejoin they rejoin at there old rating less say double the inactivity penalty for the period they were out - but not more than 1 or 1.5 points or at least 90% of their rating when they quitted - They also carry a Black Marker, so if they quit again, say within six months they lose triple the above, and if they quit a third time they are forced to where their genitalia on a necklace....

[blackice]

I wish they would fix this thing these double posts and time outs are getting annoying it took 4 minstes to edit this thing

[blackice]

Heh I don't mind you using me as an example but It is not the best example... I only cover a single reason for not wanting to be ranked there are better ones like 420...

Gavrushka that offers no incentive, personally if one quits for personal reasons, family, job why then should they be punished continously? The whole idea is participation that is why I used 420 as an example.

One way to ensure a steady start and less quiting would be to allow players until the 9th turn to decide if they would like to be ranked in any given game. Right now most quit because of bad land and no hope to win via that or life. So as it stands they get hit in the ranks each 9th turn and get a penalty for wanting to curb their continued losses. I can't blame anyone for wanting out of a game they can not win personally I would find that boring. I would prefer to start a new one too can you say hari kari...

I would think a time limit would be more fair, with a decreasing to start from square one approach. If they start back up again within 4 months the old score applies, each month after that a decrease in the penalty, a year a fresh start. This would offer incentive to rejoin. it would also be more fair to people who quit for reasons beyond their control IE: a life

Let's face it most games here are with newbies or the middle to low zones so your crawl to the top is going to take forever. I mean how many of us are going to play Klair lol we need more incentive for players not more penalities...

One more thing has anybody given more thought to the way the data is gathered for the ranking system ie: the civ score instead of the pg?

It is more accurate it takes many more real game factors into concideration then the pg. It may also help to curb quiting as terrain is taken into concideration in it. Kinda like two birds with one stone don't you think?

The original problem with it was it was a pain to do, with this whole system based on honesty I see no reason not to use it. We have moderators in most games now who could check from time to time to ensure honesty. Each player could forward their score to the moderator for compiling with the next turn. And heavy penalities for fudging.

Nope this is not changing the ranking system just what data is used for the ranking system.