[Attempt at] New PBEM rating system - continued

[Solver]

Oops, missed a lot due to connection being down for almost a day, but Mathe, it looks promising... now just wait till I have a closer look at it .

[blackice]

Math tell me why you don't see the flaw? Could it be you have not play tested this? I think so... If so email me your results or post them here if you have play tested this...

I keep saying this but I am sure you will find out soon enough we have been there did that, the old system, the elo system. They are flawed require lots of rules are complicated. Very few liked it, remember the community voted out the first one and ruled out the elo.
So insinuations no it's called play testing and lots of it. Try it, we have... Maybe you should take the time...

You and Solver simply are not listening or do not care what people like Troll have to say. What he is saying is what the majority have already said.

I challenge you to play test this and put up a simple formula so the average person can also play test it.

Until then you have a formula that may or may not work, you say it does but I have not seen any proof it does under normal game conditions.

Anyone else?

[blackice]

For instance as I said CLEARLY:

in a 3 player game, A(dam), B(rian) and C(hris) are playing.

If Adam starts a game on tundra and chris on mountains and rivers etc... Who would you expect to win the game?

Was it skill?

If you are Adam and a newbie like Paul comes into your game who would you expect to win?

Was it skill?

You simply can not base a rankings system solely based on skill in a game such as CTP. It is ludicrous at best. I simply can not figure out why some people simply do not get this very important point.

Based on this one point your system is flawed, you will have people dropping games like flies within the first turn 9. We have already seen this take the hit and start another game.

Sooner or later you have several ace land games and you will be in top stop based on land not skill.

I also can not stress enough we have been there done that do you understand that? Your idea is not new it has been done and play tested it does not work...

At least in the 10% system the only person to take a fair sized loss is the loser. All other players gain or take a minor hit. That equals desire to play the game lets look at your examples:



in a 3 player game, A(dam), B(rian) and C(hris) are playing.
the ratings before the game:
A : 1050
B : 1020
C : 980
ABC : +4 / -1 / -3 // SUM = 0
A : 1054
B : 1019
C : 977
ACB : +4 / -11 / +7 // SUM = 0 ( Why would C lose 11 points for beating a better B? Why would B gain more than A for being last, why would B gain for losing to a lesser C? )
A : 1054
B : 1027
C : 969
BAC : +9 / -6 / -3 // SUM = 0
A : 1044
B : 1029
C : 977
BCA : +9 / +7 / -16 // SUM = 0
A : 1034
B : 1029
C : 987
CAB : +17 / -6 / -11 // SUM = 0
A : 1044
B : 1009
C : 997
CBA : +17 / - 16 / -1 // SUM = 0 ( Why would A lose less here for being beaten by both lessor opponents? Then the above example where he beat B? At that note why would B lose a whopping 10 more points here when unlike the above example he beat the top ranked player, flawed.)
1. A : 1049
2. B : 1004
3. C : 997
ACB : +4 / -11 / +7 // SUM = 0
CBA : +17 / - 16 / -1 // SUM = 0
CAB : +17 / -6 / -11 // SUM = 0
Between ACB, CAB and CBA I really think you should have another look at this part of your system, rules, the other part can wait...

[checkMate]

I don't think those number could be correct.

If the order is XYZ

X will always gain and will gain more the Y Y may loose or gain. Z will definitely loose. Y will always gain less then X and will gain or will loose less than Z. Anything else would be rediculous.

[Kralj Matjaz]

[Mathemagician]

the point you dont seem to get, blackice, is that the f**king scores are not reassessed every *9 turn.
the new rating simply replaces the old rating, so i really dont see where you suddenly take these values from and seem to base another rated *9 turn on these somehow.

A : 1054
B : 1019
C : 977

A : 1054
B : 1027
C : 969

A : 1044
B : 1029
C : 977

A : 1034
B : 1029
C : 987

A : 1044
B : 1009
C : 997

1. A : 1049
2. B : 1004
3. C : 997

this system may not be the simplest, blackice, but its sure simpler than you like it to be.

[Mathemagician]

Originally posted by blackice
Math tell me why you don't see the flaw?

alright, so i tell you why i dont see a flaw that you claim exists ? no problem !!!

Very few liked it, remember the community voted out the first one and ruled out the elo.

judging from Kralj's post i conclude that the community was how many people out of the real community ?

So insinuations no it's called play testing and lots of it. Try it, we have... Maybe you should take the time...

pretty pointless to tell me to do playtesting (which you know is a lot of work), when you could far more easily just tell me what the so-called flaw is, so i could see it without putting 100 times the work in it and give up, which seems to be what you'd prefer.

admitted, i havent tested this system with 'real' results.
but that is the reason why i am discussing it here and also why i made the suggestion that we start the season with different systems and then the players make a simple vote which they like most.
they can see the results instead of understanding the whole system. this is what elections are usually based on.

You and Solver simply are not listening or do not care what people like Troll have to say. What he is saying is what the majority have already said.

i listen a lot, you talk a lot, yet still we do not seem to understand each other's problems. what a cruel world.

I challenge you to play test this and put up a simple formula so the average person can also play test it.

..or the average person directly sees the result in a sample table they can look at.
what does the average person say ?
how about this...
each average person who likes gives some theoretical test results, here in the forum, and each of our systems is applied to these test results and each of us then posts a sample table from our own system using that test data...

Anyone else?

sure as a certain hot place there is someone else !

[Mathemagician]

Originally posted by blackice
You simply can not base a rankings system solely based on skill in a game such as CTP. It is ludicrous at best.

you can, because
a) dealing with bad luck is part of skill
b) bad luck will be eliminated from the results the more games are included.

Based on this one point your system is flawed, you will have people dropping games like flies within the first turn 9.

simple workaround:

a) dropping a game means last place.
b) make clear that only someone who is serious shoud start a game and deal with the problems they get. i, for my part, never drop a game due to bad land, let alone a multiplayer game.

also, in my system, only the most actual *9 turn counts, no matter how many *9 turns you might have been on the bottom of the rankings before.

Sooner or later you have several ace land games and you will be in top stop based on land not skill.

not if you only count the most actual *9 turn. cause then all the dropped games will drag you down and the few leading games wont help you.

gotta leave now.
more on the other points later...

[Solver]

Pretty good thinking, Math .

[quinns]

I think both rating systems have been described in depth. It is time to vote, in my humble opinion!

Solver, I think much would be "solved" in this regard by creating two polls:

Poll 1. Would you like to be part of a multiplayer rating system for CTP?
A) Yes
B) No

Poll 2. Which CTP rating system would you prefer:
A) The 10% Wager System described by Blackice;
B) The Quadratic Formula System described by Mathemagician;
C) I don't care which system is used.

[Mathemagician]

could we possibly add my own system here ?
the quadratic system id rather call "the Solver/quinns" attempt.

[Mathemagician]

also, what about the idea to start with different systems ?

e.g.
Poll 2
Would you like to
a) start the rankings, using different systems and narrow down to one several weeks into the season ?
b) view some sample rankings based on sample results first and then decide on a system ?
c) decide on a system right away ?

and Poll 3 would be:
"IF a system is used right away, which should it be ?"
a) 10% wager system of blackice
b) quadratic formula system of Solver/quinns
c) Mathemagician's brilliant non-exponential ELO based system
d) where's the challenge ladder thread again ?

[Mathemagician]

in Poll 1 i vote YES anyhow

[Solver]

Several systems is a BIG pain in the ass.

Half three in the night, can't think straight... let's see tomorrow, I'll start at least the Poll #1.

G'Night.

[blackice]

I don't think those number could be correct.

Copied and pasted from Math's example, which I point out his example is flawed, so yes I agree with you the numbers do not seem right. Can you explain that Math?

so i really dont see where you suddenly take these values from and seem to base another rated *9 turn on these somehow.

The "values are copied and pasted from your example... and your example is flawed, if that is the base of your system it is flawed..I simply took your start ranking and then applied the possible A,B,C finish's you gave examples for.

so, for several finishing orders:

ABC : +4 / -1 / -3 // SUM = 0
ACB : +4 / -11 / +7 // SUM = 0 c loses for beating b? b gaims for losing to c and a?
BAC : +9 / -6 / -3 // SUM = 0
BCA : +9 / +7 / -16 // SUM = 0
CAB : +17 / -6 / -11 // SUM = 0 a loses for beating b?
CBA : +17 / - 16 / -1 // SUM = 0 b loses for beating a?

Now tell me those numbers posted are wrong again? You posted them based on your system...I just applied the final math and whoa it's flawed...
Have another look...Tell me in simple terms your posted this what did I or you miss?

Kralj post's mean about as much to me as used toilet paper, I like everyone and think everyone is a good person until they prove to me i'm wrong ...Kralj just keeps providing reasons to prove me wrong...I would think by now others too...

b) bad luck will be eliminated from the results the more games are included.) dealing with bad luck is part of skill

So if you only have time for one or two games? If you go back you will find this was the single most shouted complaint of the 10% system. Right old timers...You know it as well as I do...

I don't think you quite get the gravity of land in this game...Or do you hummm. Now toss in rookies to the mix oh the fun of applied non-skill.

10% system makes allowance for this yours does not and you simply dismiss it, not good.

Here is another one we have seen, you can only challenge people up or down three from your ranking...You have top playing top and rookies playing rookies, bad news been there the elite system.

If not that then why in heavens sake would a top player play a rookie? Time for more rules, been there too it will not end pages of rules to cover all the angles. The 10% needs no such rules. Simpler then you care to admit...

a) dropping a game means last place..also, in my system, only the most actual *9 turn counts, no matter how many *9 turns you might have been on the bottom of the rankings before.

For one 9th turn or for the entire ranking? Seems rather sever if over all ranking drops to last for a dropped game, 10% system needs no such rule... So the joy of playing a game on tundra being in last for the entire game as a rookie, I can not think of a better way to punish one's self or to drop the league all together you?
Been there seen that...

Land, Math, accounts for the better part of a win no matter what your skill level... In Multiplayer and PBEM challenges you know as well as I do land is extremely important. This comment of yours is a whitewash of reality.

Why play a back to back same map challenge games if land means nothing? Why play the same map in the world challenge if land means nothing vs skill? Why start games over again in MP if land means nothing vs skill.
That exclusion of land as a major factor in this game vs skill is pure BS and you know it. I realize this system is yours and you would like to show your skills, which you have done. But do not hardcore it through by whitewashing what every single "SKILLED" player here knows...LAND IS A HUGE PART OF ANY WIN and or loss..Be real...

Flaws as pointed out in my last post, I took your starting ranking then applied the posible A, B, C finishes you gave examples to. I posted the comments did I miss something or are you just ignoring it? YOU GAVE THE MATH I JUST POSTED IT...it's flawed.. LOOK AGAIN ÌööÌ...

Now what problem do you have with the 10% system? I for one would like to hear it?

As far as a poll goes we maybe should consider just calling them:

a) 10% wager system
b) quadratic formula
c) non-exponential ELO based system

Give examples of the system I.E. The Formula's let the players decide, mind you been there did that the quad lost to the 10%. But hey lets do it again...and again...

We should keep it impartial if not it attracts childishness like what we have seen from... well take a guess think sore loser..., all personal with no content no community in mind posting what so ever...

No time for that if we plan on keeping this an impartial fact based, common sense, reality vote. That which should be in the best interest of the community.