Home » Forum Home » Development

Topic: Testers needed - set-based ratings
Replies: 35   Views: 118,619   Pages: 3   Last Post: Feb 15, 2009, 10:41 PM by: baddsixx9

Search Forum

Back to Topic List Topics: [ Previous | Next ]
Replies: 35   Views: 118,619   Pages: 3   [ Previous | 1 2 3 ]
gaylen

Posts: 112
Registered: Dec 16, 2001
From: cottage grove, oregon
Home page
Re: Testers needed - set-based ratings
Posted: Feb 13, 2009, 11:38 PM

I just checked out the test site and let myself in since the lights were on. No one was home there...probably because its like going to a new hangout where no one is hanging out.

I never knew about this test site until today.

Maybe a good way to advertise it and get player's arses over there would be to put a link/button in the 'game room' that would transport players over to the test site's 'game room' through the click of a link.

I would make a transporter button with an image of the enterprise transporter/stage on it. If you are a donor then your 'loaded image' appears and fades out [programing using an opacity slider] as a transporting effect and then fades back in, in the same room setting, but now in the opposite twin site, i.e., pente.org, or test.pente.org.

Maybe a computer voice could announce 'id' departures and arrivals with airport terminal sounding acoustics.

I would probably have an internet jukebox in the game room
that only donor transporters could use for one song while transporting, or 10 cents for others. Only one country and one western song could be played per day and other types of music might have to be segregated.

zoeyk

Posts: 2,233
Registered: Mar 4, 2007
From: San Francisco
Age: 45
Home page
Re: Testers needed - set-based ratings
Posted: Feb 14, 2009, 12:04 AM

there is no draw,....with out draw button someone must lose by Time Out if board is full and cant move.
the last stone placed wins. in C6 white is final stone.
i think is same in pente. but in pente i dont think the board can fill complete with out 5 captures. this seems impossible to ever happen.

Scire hostis animum - Intelligere ludum - Nosce te ipsum - Prima moventur conciliat - Nolite errare
richardiii

Posts: 387
Registered: Dec 8, 2002
From: Huntsville, Alabama
Age: 67
Re: Testers needed - set-based ratings
Posted: Feb 14, 2009, 6:42 PM

I really struggle with the set-ratings concept.

What if Jim (2000) and Bob (840) play a set?

Jim wins his P2 game but loses his P1 game. A draw right?

Jim remains at 2000 and Bob remains at 840.

But don't you think for Bob to win even a single game vs Jim that he should have a dramatic gain to his ratings? No? Well what if this went on for 10 sets in a row?

I am sorry but the only advantage i c in set based ratings is an unfair advantage for the higher ranked player when he slips and loses a game vs a lower ranked player. He comes out scott free for his poor play while the lower ranked player gets snubbed for his monumental achievement.

When i play a 1400 player 10 games i generally win all my games, unless i make lazy mistake and slip in a game, as I sometimes do. Well that lose may not mean much to me, but to that 1400 player it means a great deal indeed. But with this system that player will now be denied any sense of accomplishment or any real chance for success at all, as it is very unlikely that a 1400 player will ever take both a P1 and a P2 game from me in the same set.

The set based system seems to serve only as an "Oops" buffer for the higher ranked players.

The set based system may at least seem fairer to evenly rated players, but the present system reflects that already. Any evenly rated players who split 10 games generally wind up with their ratings largely unaffected.

I am not trying to deflate Peter's efforts in this whole matter, but I can't participate in something I c as discouraging and unfair.

I know all this is an effort to even the odds vs the P1 advantage. Why not simply make a P2 win twice the value of a P1 win, and a P1 lose twice the negative value of a P2 lose?

baddsixx9

Posts: 111
Registered: Nov 14, 2008
Home page
Re: Testers needed - set-based ratings
Posted: Feb 15, 2009, 12:23 AM

Well crap since you put it that way I see your point R3..........there's a baby in the bathwater here.

watsu

Posts: 1,464
Registered: Dec 16, 2001
Home page
Re: Testers needed - set-based ratings
Posted: Feb 15, 2009, 10:10 AM

If Jim and Bob were the only two Pente players playing each other in set based games, their differences in ratings in comparison to the way they split sets would indeed be odd in R3's scenario. However, most players don't play only a single opponent- and if they do, their ratings are fairly meaningless as far as the larger Pente community goes. In this situation, either Jim is grossly overrated or Bob is grossly underrated (or perhaps some middle ground). The lack of a rating change in that particular example is unlikely to carry over to all of Jim and Bob's other games against other players, so Bob should be able to move up and Jim down (if necessary) simply by playing a larger sampling of Pente players.

Retired from TB Pente, but still playing live games & exploring variants like D, poof and boat
baddsixx9

Posts: 111
Registered: Nov 14, 2008
Home page
Re: Testers needed - set-based ratings
Posted: Feb 15, 2009, 10:41 PM

Can't say Im picking up what you're puttin' down there Watsu........... Pictures, Images and Photos
I just know it'd be regrettable to lose Rich's support for the system here and for no recognition to be given to a lower rated player for capitalizing on a mistake made by an advanced opponent.....

Replies: 35   Views: 118,619   Pages: 3   [ Previous | 1 2 3 ]
Back to Topic List
Topics: [ Previous | Next ]


Powered by Jive Software