Announcement

**Mathemagician** · July 17, 2003, 03:43

ahhh, i finally could post this.
actually, it is already a day old by now and i think ive got the excel stuff understood and solved quite okay (though not tested yet).
still, a guru would help.

**blackice** · July 17, 2003, 06:15

I still maintain CTP is not chess nor is it Magic it is unique and would require a rankings system that addresses all the variables of this unique game, chess and magic simply do not compare...

**Mathemagician** · July 17, 2003, 08:12

my system is meant to incorporate both the current (NOT past) in-game rankings of the players on the power graph of each game they're playing, and their score (=skill estimate) in the overall ranking table.

no system will seem perfect to everyone, but i think this ranking would have more significance than quinns' ladder right now..

**Solver** · July 17, 2003, 08:30

A question to you all - who can count the changes that the various ratings systems have undergone, ever since the original idea came here?

I must say, though, that the MP community here has became much more interesting after the ratings were introduced. The Challenge Ladder is quite interesting though, IMHO.

**Mathemagician** · July 17, 2003, 08:33

Originally posted by Solver
A question to you all - who can count the changes that the various ratings systems have undergone, ever since the original idea came here?

mine wouldnt undergo any changes as it is perfect the way it is.

**Solver** · July 17, 2003, 08:51

Now that's a pretty self-confident comment. I don't believe that anything can be perfect. Ouch, just noticed the smilie

.

Seriously, though, I think our ratings have been through a very high number of changes. Where are the current PBEM ratings anyway, eh?

**Mathemagician** · July 17, 2003, 08:54

Originally posted by Solver
Now that's a pretty self-confident comment. I don't believe that anything can be perfect. Ouch, just noticed the smilie

.

seriously, i think that by the time i got this system done, it would be quite good. thats serious !

Seriously, though, I think our ratings have been through a very high number of changes. Where are the current PBEM ratings anyway, eh?

nowhere ?
the question is where they could be !

**Solver** · July 17, 2003, 09:01

Good, yes, possibly, but not perfect

.

Well, there might very well be a reason why the PBEM ratings are no longer here the way they were before. You are relatively new here, so you don't remember how big and good they were, say, at a point 2 years ago.

Both the WTC ratings used, as well as the more simple 10% formulae were good, and yet none of these rating systems have a big visible presence now... there has to be a reason, doesn't there?

I believe that, one of the main reason for that were the time constraints. With about 50 active rated players engaged in a few dozen PBEM games, there was quite a deal of results to proceed under the calculations to get the new ratings - that takes time. And while a few people have, at different points in time, expressed a desire to write a program that would simplify the job, unfortunately, it never happened.

I'll take this chance and yet again say big thanks to Quinns, Gavrushka, Blackice and other people who had the devotion to manage the ratings.

It might be my nostalgia, but this is what I would see as the best way:

Returning the PBEM ratings as they originally were, with minor tweaks possibly, using the old tennis formulae (that are rather simple, actually), but this time finally using a decent program to keep track of players and do the calculations, so that all the job would just take a few minutes, feeding ratings to the program and getting the result.

**blackice** · July 17, 2003, 09:02

Toss out a formula so we can see.

Do to personality conflicts, mass hysteria and partisan politics the current system was folded.

IMHO it was not perfect but did address PBEM as a unique game far better than the old one. Actually I was not the only one to think that way.

I main problem was finding the time to do it. I have since found a program that would automate the calculations.

back to finding the time....

**Solver** · July 17, 2003, 09:08

Pete, what program have you found?

And Blackice also just proves my point - the rating system was, erm, postponed or cancelled mainly not because of its drawbacks as a system - the main problems were time and some personal conflicts among the members. Both the tennis and 10% formulae worked fine, the former being my personal preferrence.

If only the few remaining active CtP players could stand out without conflicts...

**Mathemagician** · July 17, 2003, 09:14

Originally posted by blackice
Toss out a formula so we can see.

are you addressing me here ?

IMHO it was not perfect but did address PBEM as a unique game far better than the old one. Actually I was not the only one to think that way.

i noticed a lot of people around here just say 'the old' and 'the new' system. could you please be more precise, e.g. by mentioning a key aspect of the system you are referring to ?

I main problem was finding the time to do it. I have since found a program that would automate the calculations.

which one ?
i believe excel could do as well...

**blackice** · July 17, 2003, 09:15

Returning the PBEM ratings as they originally were, with minor tweaks possibly, using the old tennis formulae

CTP is not tennis either... The system could not be fixed it was tried.

The new system on the other hand was new and needed a few minor tweeks. It was a proven system made for CTP.

**Solver** · July 17, 2003, 09:24

Mathe, Excel definitely doesn't cut it. Yes, it's a decrease in time, but still, you have to manually keep track of the players, manually enter their ratings, and such. The idea for a perfect program would be just one where you select names of players, order them by their power graph settings, then the program fetches their current ratings from the database, does the calculations, and there you go.

Blackice - the tennis system was fine. CtP is not tennis, but the formulae were very well designed to work with other games, too. I remember that some people who were not too good with maths got scared of the formulas, but they seemed very easy to me once I took a look. The system had drawbacks, but they were mainly fixed. Although, I remember that you were probably the strongest one of those against this system.

The "new" system was also good, being designed to simplify the formulas somewhat, and make it easier, although it still took time. While it was balanced enough and rather good... I did at times feel it was too simple.

Mathe, the old Tennis system merely used same formulas as Tennis does to calculate ratings adjustements by powergraph, where the key aspect was the difference in ratings between players. That is, a high rated player got little by beating a much lower rated players, but a low rated player got a lot by beating a highly rated player. For the record, very similar formulae are used in many RTS matchmaking systems.

The new system was more dependant on the player count in a game, and it was, well, simpler, depending largely on certain constant numbers, but I don't exactly remember off the top off my head.

**blackice** · July 17, 2003, 09:24

are you addressing me here ?

yes

Some of us have

I think Math's suggestion some time back of limiting the 10% or making it static per game would have fixed the systems one minor flaw. Mind you I did not see it as a flaw. The more games you play the more rankings you should have.

The change was contemplated to hopefully bring the community together, back to partisan politics....

Excel can do it not problem but it still requires manual input by a member or several volunteers. This program is incorporated into the HTML or PHP coding.

The players enter them then it automatically adjusts the rankings.

All the players do then is post the information for confirmation reasons.

Personally I have aways agreed with combining PG and score as a more realistic approach to CTP.

Announcement

[Attempt at] New PBEM rating system - continued

[Attempt at] New PBEM rating system - continued

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment