How to realize ELO in a multiplayer game anyway?

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

The subject says it all. In one-on-one games ELO is easy and clear. But
in multiplayer games? If you score 0VP at a table did you loose against
all the others? Against the one with the GW (if any)? Against anybody
with a GW? And if you "win" (what is winning? scoring a VP? a GW?)? Are
additional points rewarded for going into the finals? Can your score go
down playing the finals e.g 5th place has a lower score than 6th
becasuse he lost the final (as it was the case with the old stupid system)?

 

[mummy]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

Johannes Walch wrote:
> The subject says it all.

Sort of. What does ELO stand for?

-Jon

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

Sten During wrote:
> 1) Don't. ELO is designed for duels. It's not, repeat, not, designed for
> multiplayer action where player number three could trash your game and
> gain nothing for it.

That is an argument against rating performance at all, by ELO or any
other method, in any multi-player game that has any possibility of king
making. So either you abandon rating performance, or you accept the
level of king making that is possible in VTES as part of the "error of
estimation".

> If, but only possibly if, VTES had given the VP to
> whoever ousted a player would ELO be adaptable to VTES as jumping the
> currently weakest player would normally always pay off, and then you'd
> really lose against all others.

Nah. You'd still have king making there, too.

> What people tend to forget is that duels always mirror a "last man
> standing" way of measuring points. Lose, draw or win. Remake all the
> rules and have first player ousted get 1 VP, second 2 VP etc and you
> end up with a multiplayer game resembling a duel. Poker is a good
> example.

In poker, every ousted player gets $0 (bankrupted out of the game). The
last man standing takes all. Unless you're talking about some external
reward system.

But poker also has forced risk (ante). VTES does not, so implementing
such a VP payout scheme would kill the game (by making it a waiting
game -- first person to make a move loses).

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

mummy wrote:
> Johannes Walch wrote:
> > The subject says it all.
>
> Sort of. What does ELO stand for?
>
> -Jon

The name of the band the guy was listening to when he created the
rating system.

Electric Light Orchestra.

http://www.elomusic.com/

enjoy

oscar

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

On Wed, 3 Aug 2005, mummy wrote:

>
> Johannes Walch wrote:
>> The subject says it all.
>
> Sort of. What does ELO stand for?

It's the name of the guy who invented the rating system.

http://en.wikipedia.org/wiki/ELO_rating_system

Matt Morgan

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

mummy wrote:

> Sort of. What does ELO stand for?

ELO is actually a guys name--the system was invented by Mr. Elo. It is the
scoring system used, primarily, for chess rating. When you join the
Federation or whatever, you get a starter rating (1000?). When you play
folks, your rating goes up or down. When you play someone with a higher
rating than you, you are expected to lose, so if you do lose, your rating
doesn't go down much, but if you win, your rating goes up a lot. On the
other hand, if you play someone with a lower rating than you, you are
expected to win, so if you win, your rating goes up a little bit, but if you
lose, your rating goes down a lot.

So if you are a newbie with a 1000 starter rating (I don't know off hand if
1000 is the correct starter number, but we'll assume for now) and you play
against a mid level guy with a 1200 rating, and you lose, your rating only
goes down a bit (say, 20 points or something), as you were expected to lose.
If you win, your rating goes up a larger percentage (say, 50 points), as you
were fighting uphill. If you sit down and play against someone with some
crazy, like, 2800 rating or something, if you lose, you lose very little,
but if you win, you win a great deal.


Peter D Bakija
pdb6@lightlink.com
http://www.lightlink.com/pdb6

"So in conclusion, our business plan is to sell hot,
easily spilled liquids to naked people."
-Brittni Meil

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

"Johannes Walch" <johannes.walch@vekn.de> wrote in message
news:dcq3pt$ol$1@news01.versatel.de...
> The subject says it all. In one-on-one games ELO is easy and clear. But
> in multiplayer games? If you score 0VP at a table did you loose against
> all the others? Against the one with the GW (if any)? Against anybody
> with a GW? And if you "win" (what is winning? scoring a VP? a GW?)?

It's a good question. The original VEKN ELO system answered the question
by conceptually converting each multiplayer tournament game to a series
of two-player duels resolved by comparing the number of victory points
obtained by each of the two players and doing its rating calculation based
on who got more VPs. For example, say a five-player game between players
A, B, C, D, and E ended with player A getting 3 and 1/2 victory points,
players B and C getting 1/2 victory point, and players D and E getting
ousted without a point. This real game would be converted into ten
conceptual duels, rated simultaneously, with the following results:

A (3 1/2 VPs) defeated B (1/2 VP)
A (3 1/2 VPs) defeated C (1/2 VP)
A (3 1/2 VPs) defeated D (0 VPs)
A (3 1/2 VPs) defeated E (0 VPs)
B (1/2 VP) drew with C (1/2 VP)
B (1/2 VP) defeated D (0 VPs)
B (1/2 VP) defeated E (0 VPs)
C (1/2 VP) defeated D (0 VPs)
C (1/2 VP) defeated E (0 VPs)
D (0 VPs) drew with E (0 VPs)

A four-player real game would be reduced to six conceptual duels in
the same manner.

I don't consider this to be an issue in the implementation of ELO
system for VEKN. Other issues, stemming from the necessity of entering
tournament results simultaneously, are problematic but not this one.
Granted, its not very intuitive and it lends itself to a fair number
of drawn results (for instance when one player sweeps the table) but
I don't see any actual problems with that. Because of the number of
duels rated, it weights five-player games more heavily than four-
player games by a factor of four to three. To me, that seems
appropriate, although you could also add a fudge factor to four-
player games to balance the weightings if you wanted.

> Are
> additional points rewarded for going into the finals? Can your score go
> down playing the finals e.g 5th place has a lower score than 6th
> because he lost the final (as it was the case with the old stupid system)?

No additional points are rewarded for going into the finals. ELO is a
closed system - what one player gains, another player must lose. The
finals are rated the same as any other tournament game. But that doesn't
mean there's a penalty for going to the finals and losing. Because of
this close system nature, the expected outcome of playing any ELO rated
game should be zero. That's why ELO doesn't care how many tournaments
a player does or doesn't play in. If you're rated correctly, you
should have an equal chance of gaining points as losing points - in an
easy tournament, a hard tournament, a big tournament, a small
tournament, the finals of the Continental Championship, it doesn't
matter.

The better the players you play in any given game, the harder it should
be to do well in that game. But because their ratings are higher, this
will consequently get factored into the ratings calculations. So if
you did well, you would gain more points for beating better players.
But if you did poorly on the other hand, you would lose fewer points.
The *average* outcome - assuming all players are rated correctly at
the time they play - should be approximately zero.

Fred

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

LSJ wrote:
> Sten During wrote:
> > 1) Don't. ELO is designed for duels. It's not, repeat, not, designed for
> > multiplayer action where player number three could trash your game and
> > gain nothing for it.
>
> That is an argument against rating performance at all, by ELO or any
> other method, in any multi-player game that has any possibility of king
> making. So either you abandon rating performance, or you accept the
> level of king making that is possible in VTES as part of the "error of
> estimation".

I think you're right about the kingmaking, but I think Sten is right
that an Elo system requires some interpretation to apply to the scoring
system of multiplayer VTES. One logical way to do it might be to say
that you lose against the person who ousts you, and win against each
person you oust; your score doesn't change based on people at the table
that you didn't have an ousting relationship with. If you're not
ousted, perhaps either score a draw against each other remaining player
(if it's called on time) or score no change (if you withdrew)?

Another one might be "the old way", where you beat everyone with fewer
VPs than you at the end of a game, lose to everyone with more VPs than
you, and tie against everyone with the same number of VPs... but that's
rather less concrete in its claims of how meaningful your interaction
with the other players (who you might never have sat next to) was
during the game.

Yet another way could be to say that whoever gets the game win beats
everyone who didn't get the game win, and they all lose to him or her,
but aren't scored against each other. That might be the method most
closely tied to the normal definition of what winning and losing means
in VTES.

> > What people tend to forget is that duels always mirror a "last man
> > standing" way of measuring points. Lose, draw or win. Remake all the
> > rules and have first player ousted get 1 VP, second 2 VP etc and you
> > end up with a multiplayer game resembling a duel. Poker is a good
> > example.
>
> In poker, every ousted player gets $0 (bankrupted out of the game). The
> last man standing takes all. Unless you're talking about some external
> reward system.

In small games, sure, but in larger poker tournaments, surviving longer
normally means winning more money. If the tournament is large enough,
many places below first may be paid (e.g. at the World Series main
event this year, something like 560 people were paid, out of about 5600
entrants).

> But poker also has forced risk (ante). VTES does not, so implementing
> such a VP payout scheme would kill the game (by making it a waiting
> game -- first person to make a move loses).

Here, though, you're quite right, and it's an interesting point you
make, since the lack of forced risk in VTES is not infrequently a
problem in VTES games (especially tournament games, and even more
especially tournament finals). This is probably in part because VTES
has the Edge, which aids the person holding it, but no "Anti-Edge" to
force other players to either move the game along or be gradually
ousted.

It might be an interesting variant to add some kind of Anti-Edge to
VTES. Then again, it might turn out that an Anti-Edge would either be
too weak an influence (if its effects could be easily offset by, say,
hunting every turn with one vampire) or too strong (if it killed people
so fast as to make the game unfun and eliminate a lot of decks that
people enjoy playing)...


Josh

randomly speculating

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

"OrgPlay" <orgplay@white-wolf.com> wrote in message
news:1123086324.795057.321580@g43g2000cwa.googlegroups.com...
>
> mummy wrote:
>> Johannes Walch wrote:
>> > The subject says it all.
>>
>> Sort of. What does ELO stand for?
>>
>> -Jon
>
> The name of the band the guy was listening to when he created the
> rating system.
>
> Electric Light Orchestra.
>
> http://www.elomusic.com/
>

No wonder ELO formulae are so trippy...


DZ
AW

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

In message <_Z8Ie.55036$4o.854@fed1read06>, Frederick Scott
<nospam@no.spam.dot.com> writes:
>hat's why ELO doesn't care how many tournaments
>a player does or doesn't play in. If you're rated correctly, you
>should have an equal chance of gaining points as losing points - in an
>easy tournament, a hard tournament, a big tournament, a small
>tournament, the finals of the Continental Championship, it doesn't
>matter.

Could you explain what you mean by "ELO doesn't care how many
tournaments a player does or doesn't play in"?

In many sophisticated ELO systems, it requires rather a lot of games
played to get a high ranking. How many matches do you think a novice
chess player with a bad/starting rating would need to play in to hit
grand-master rating (assume he's playing against top-class players, if
that helps)? You don't suddenly go from novice to "Whoop, whoop, move
over Garry Kasparov" in chess ELO ratings if you beat him twice, or even
ten times.

--
James Coupe
PGP Key: 0x5D623D5D YOU ARE IN ERROR.
EBD690ECD7A1FB457CA2 NO-ONE IS SCREAMING.
13D7E668C3695D623D5D THANK YOU FOR YOUR COOPERATION.

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

"James Coupe" <james@zephyr.org.uk> wrote in message
news:G6QyYPYF3S8CFwRx@gratiano.zephyr.org.uk...
> In message <_Z8Ie.55036$4o.854@fed1read06>, Frederick Scott
> <nospam@no.spam.dot.com> writes:
>>That's why ELO doesn't care how many tournaments
>>a player does or doesn't play in. If you're rated correctly, you
>>should have an equal chance of gaining points as losing points - in an
>>easy tournament, a hard tournament, a big tournament, a small
>>tournament, the finals of the Continental Championship, it doesn't
>>matter.
>
> Could you explain what you mean by "ELO doesn't care how many
> tournaments a player does or doesn't play in"?
>
> In many sophisticated ELO systems, it requires rather a lot of games
> played to get a high ranking. How many matches do you think a novice
> chess player with a bad/starting rating would need to play in to hit
> grand-master rating (assume he's playing against top-class players, if
> that helps)? You don't suddenly go from novice to "Whoop, whoop, move
> over Garry Kasparov" in chess ELO ratings if you beat him twice, or even
> ten times.

Sure. Note the caveat I wrote at the start of the sentence: "If you're
rated correctly...". The formula starts novices at an average score (as
designated by whoever sets up the system) which, unless they happen to be
perfectly average players, is wrong. So as they play games, the point of
the formula is to correct your ratings towards what it should truly be.
As more and more games are played, the weight of the original dummy
average rating gradually decreases towards zero as at a speed determined
by the coefficients chosen by whoever set up the system. (If fact, those
coefficients regulate the whole issue of the value of recent games vs.
older games. That was what was wrong with VEKN's original implementation:
the coefficients placed far too much weight on recent games given the
amount of luck Jyhad involves.)

So the short answer is, the coefficients determine how long it takes to
move from your dummy average score towards your approximate true rating.
If you're a terrible player or a great player, it will take you longer to
get there than if you're an average player. So yes, in a technical sense
it DOES care how many games you've played - to a point. I wasn't trying
to maintain that it didn't; just ignoring that issue for simplification.
But once you've hit your approximate true rating, it no longer cares.
Or at least, it doesn't unless something causes an anomaly that changes
your ability a great deal in a short amount of time. ("HEY! STEALTH
AND BLEED REALLY WORK GOOD TOGETHER!!! I never realized that!" ;-) )

Fred

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

Peter D Bakija wrote:
> mummy wrote:
>
>
>>Sort of. What does ELO stand for?
>

> were fighting uphill. If you sit down and play against someone with some
> crazy, like, 2800 rating or something, if you lose, you lose very little,
> but if you win, you win a great deal.
>

All of this is correct, but one very important part of the ELO system is
that it forces the better player to win.
Let's assume (in chess, which I did play on tournament level at younger
days) that I have 1400 and meet someone at 1900. Should the match result
in a draw I gain some 15 points and my opponent would drop the same
amount.
Now, convert this to VTES, as it used to be. I build a cap 1 deck with
only Concealed Weapon, Saturday Night Special, Bum's Rush, Dragon's
Breath Rounds and forget my Effective Management because I don't
understand the engine needed for this kind of deck. Just before the
third prelim I get this nasty suspicion that for an unknown reason
I won't be able to oust before I'm ousted, so I spend the game rushing
upstream hoping I'll oust every predator and win (which won't happen).
The likely event is that I'll oust my predator, more or less no matter
what he she plays, and after that I'll be ousted myself. If said
predator was much higher ranked than I was I'll "steal" rating from
him/her.
Needless to say this was bad

Sten During

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

LSJ wrote:
> Sten During wrote:
>
>>1) Don't. ELO is designed for duels. It's not, repeat, not, designed for
>>multiplayer action where player number three could trash your game and
>>gain nothing for it.
>
>
> That is an argument against rating performance at all, by ELO or any
> other method, in any multi-player game that has any possibility of king
> making. So either you abandon rating performance, or you accept the
> level of king making that is possible in VTES as part of the "error of
> estimation".
>

Agreed, but the current system doesn't punish a draw forced upon the
player with a higher rating (0 VP versus 0 VP gains neither player,
bur nor does it lower the rating of the better player).

>
>>If, but only possibly if, VTES had given the VP to
>>whoever ousted a player would ELO be adaptable to VTES as jumping the
>>currently weakest player would normally always pay off, and then you'd
>>really lose against all others.
>
>
> Nah. You'd still have king making there, too.

Yes, but to a lesser degree. Multiplayer games rewarding the one who
kills the weakest player, or at least punishes players in the order
they are ousted tend to make players more eager to go for a kill
(and to a degree to gang up on the strongest player)

>
>
>>What people tend to forget is that duels always mirror a "last man
>>standing" way of measuring points. Lose, draw or win. Remake all the
>>rules and have first player ousted get 1 VP, second 2 VP etc and you
>>end up with a multiplayer game resembling a duel. Poker is a good
>>example.
>
>
> In poker, every ousted player gets $0 (bankrupted out of the game). The
> last man standing takes all. Unless you're talking about some external
> reward system.
>
> But poker also has forced risk (ante). VTES does not, so implementing
> such a VP payout scheme would kill the game (by making it a waiting
> game -- first person to make a move loses).
>

All agreed. I was thinking of the big Texas Hold 'Em tornaments for the
simple reason that rating is geared towards tournament play.
Each player gets a position, and on each table that position is
determined by reverse order of getting ousted.

I should also clarify one thing. Nothing I write here implies any kind
of wishes for the rules of VTES to change. I'm merely making examples
using VTES because people here are familiar with the game.

ELO is in my view not suitable for multiplayer games, and even less
so for VTES because of its mechanics.

Sten During

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

Johannes Walch wrote:
> Frederick Scott wrote:
> > But once you've hit your approximate true rating, it no longer cares.
>
> The problem is that for V:TES there can never be the "approximate true
> rating". In chess only one skill is required: making your moves better
> than the opponent.

I suspect that one skill could be broken down into many smaller skills
like you do with V:TES.

> V:TES consists of many skills: playing the cards
> without errors, "reading the table", dealmaking, anticipating the
> metagame, building a good deck. In chess you have a very static setup,
> each game starts the same (there is only one little factor - who is
> beginning) and the variables is the opponent. In V:TES you have a very
> dynamic setup, the variables are in no specific order: your deck, your
> opponents decks, your opponents, the relative position of your
> opponents, the metagame, whether your playing round 1 or round 3 (in
> round1 everybody goes for the GW in round3 often enough people try to
> get some more VPs) and probably some more. This leads to two
> conclusions: on the one hand your performance in a game is often not
> determind by your skill but by luck.

Agreed that V:TES has a lot of variables outside of skill. However time
has a dampening effect on them. Eventually, the luck disappears and you
are left with something that reflects your skill.

-Robert

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

OrgPlay wrote:
> mummy wrote:
>
>>Johannes Walch wrote:
>>
>>>The subject says it all.
>>
>>Sort of. What does ELO stand for?
>>
>>-Jon
>
>
> The name of the band the guy was listening to when he created the
> rating system.
>
> Electric Light Orchestra.
>
> http://www.elomusic.com/

Do I have to mention that the "E" in Elo, Keit Emerson, sometimes used
to stick knives in between the keys while playing his keyboard in live
performances? Doesn´t this say all about the ELO system

--
johannes walch

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

Frederick Scott wrote:
> But once you've hit your approximate true rating, it no longer cares.

The problem is that for V:TES there can never be the "approximate true
rating". In chess only one skill is required: making your moves better
than the opponent. V:TES consists of many skills: playing the cards
without errors, "reading the table", dealmaking, anticipating the
metagame, building a good deck. In chess you have a very static setup,
each game starts the same (there is only one little factor - who is
beginning) and the variables is the opponent. In V:TES you have a very
dynamic setup, the variables are in no specific order: your deck, your
opponents decks, your opponents, the relative position of your
opponents, the metagame, whether your playing round 1 or round 3 (in
round1 everybody goes for the GW in round3 often enough people try to
get some more VPs) and probably some more. This leads to two
conclusions: on the one hand your performance in a game is often not
determind by your skill but by luck. If you get dealed out, no chance,
no matter how good you are. On the other hand your skill level is
usually different depending on the afformentioned parameters. Therefore
your rating might be correct for the current metagame (e.g many
intercept decks resulting in many deals) but for the next metagame (e.g
lots of S&B, hardly any deals) it might be wrong, because you are a good
player and a brilliant dealmaker.

That makes ELO very unsuitable for a multiplayer game like V:TES.

--
johannes walch

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

Johannes Walch wrote:

> Do I have to mention that the "E" in Elo, Keit Emerson, sometimes used
> to stick knives in between the keys while playing his keyboard in live
> performances? Doesn´t this say all about the ELO system

Keith Emerson was in Emerson Lake and Palmer (ELP). ELO was the Electric
Light Orchestra, fronted by Jeff Lynde (sp?) who went on to be in the
Traveling Willburys with Dylan or Orbison. And all sorts of other stuff.

-Peter, who has a PHD in dorky 70's art rock.

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

"Johannes Walch" <johannes.walch@vekn.de> wrote in message
news:dcscg6$g7s$1@news01.versatel.de...
> Frederick Scott wrote:
>> But once you've hit your approximate true rating, it no longer cares.
>
> The problem is that for V:TES there can never be the "approximate true rating". In chess only one skill is required: making your
> moves better than the opponent. V:TES consists of many skills: playing the cards without errors, "reading the table", dealmaking,
> anticipating the metagame, building a good deck. In chess you have a very static setup, each game starts the same (there is only
> one little factor - who is beginning) and the variables is the opponent. In V:TES you have a very dynamic setup, the variables are
> in no specific order: your deck, your opponents decks, your opponents, the relative position of your opponents, the metagame,
> whether your playing round 1 or round 3 (in round1 everybody goes for the GW in round3 often enough people try to get some more
> VPs) and probably some more. This leads to two conclusions: on the one hand your performance in a game is often not determind by
> your skill but by luck. If you get dealed out, no chance, no matter how good you are. On the other hand your skill level is
> usually different depending on the afformentioned parameters. Therefore your rating might be correct for the current metagame (e.g
> many intercept decks resulting in many deals) but for the next metagame (e.g lots of S&B, hardly any deals) it might be wrong,
> because you are a good player and a brilliant dealmaker.
>
> That makes ELO very unsuitable for a multiplayer game like V:TES.

I agree with many of the points you made above. But your conclusion
is just as valid for any other rating procedures as for ELO. So if ELO
doesn't work for these reasons, nothing elso could either.

And it is an overstated conclusion. Were it so hard for ELO to rate
players, I assert it would also be hard to intuitively know whether some
players were clearly better than others. But it's not hard to detect
player skill on an intuitive level, so I assert that skill will also
show itself over time in an ELO rating as well - as long as the
coefficients used to define the relationship between recent results
and older results are set correctly.

Fred

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

Robert Goudie wrote:
> Johannes Walch wrote:
> > Frederick Scott wrote:
> > > But once you've hit your approximate true rating, it no longer cares.
> >
> > The problem is that for V:TES there can never be the "approximate true
> > rating". In chess only one skill is required: making your moves better
> > than the opponent.

Indeed, chess involves (very broadly clasified) at least two playing
styles (tactical, like Kasparov) and strategic (like Karpov), plus
several dozen of different openings (all of them with at least half a
dozen of minor variants), plus different involvement for the players if
they are in the opening, the middle-game or the ending (check Amazon
and you'll probably find lots of books written specifically for each
stage of the game).

Besides, pro chess players, have a pack of analyst helping them
designing strategies to beat their opponents, anticipate openings they
might use, etc, etc, et...

K.

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

Kalessin wrote:
> Robert Goudie wrote:
> > Johannes Walch wrote:
> > > Frederick Scott wrote:
> > > > But once you've hit your approximate true rating, it no longer cares.
> > >
> > > The problem is that for V:TES there can never be the "approximate true
> > > rating". In chess only one skill is required: making your moves better
> > > than the opponent.

> > Goudie: I suspect that one skill could be broken down into many smaller
> > skills like you do with V:TES.

> Indeed, chess involves (very broadly clasified) at least two playing
> styles (tactical, like Kasparov) and strategic (like Karpov), plus
> several dozen of different openings (all of them with at least half a
> dozen of minor variants), plus different involvement for the players if
> they are in the opening, the middle-game or the ending (check Amazon
> and you'll probably find lots of books written specifically for each
> stage of the game).

I think Google fouled your attributions. I said that Chess could be
broken down into smaller skills (and not Johannes' comment that was
attributed to me). The above version may be fixed (barring more Google
mishaps!).

-Robert

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

Robert Goudie wrote:
> I think Google fouled your attributions. I said that Chess could be
> broken down into smaller skills (and not Johannes' comment that was
> attributed to me). The above version may be fixed (barring more Google
> mishaps!).
>
> -Robert

Yep... you're right, sorry

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

Peter D Bakija wrote:
> Keith Emerson was in Emerson Lake and Palmer (ELP). ELO was the Electric
> Light Orchestra, fronted by Jeff Lynde (sp?) who went on to be in the
> Traveling Willburys with Dylan or Orbison. And all sorts of other stuff.
>
> -Peter, who has a PHD in dorky 70's art rock.

Sorry, I was totally confused for a moment .....

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

Johannes Walch wrote:

> Sorry, I was totally confused for a moment .....

Understandable--thay are both ponderous and absurd. But ELO is a bit more
entertaining in the "goofy 70's radio single art rock" kinda way where ELP
are bit more preposterous in the "we did an entire album of Musgorsky
classical covers, but with electric guitars and a moog" kinda way.


Peter D Bakija
pdb6@lightlink.com
http://www.lightlink.com/pdb6

"So in conclusion, our business plan is to sell hot,
easily spilled liquids to naked people."
-Brittni Meil

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

Frederick Scott wrote:
> I agree with many of the points you made above. But your conclusion
> is just as valid for any other rating procedures as for ELO. So if ELO
> doesn't work for these reasons, nothing elso could either.

I disagree. ELO rates negative performance. The current system doesn´t
rate negative performance as long as you attend enough tournaments.
Quite a difference. I agree with you that the current system does not
reflect the skill of a player very well who does not attend enough
tournaments. But with any sort of rating you need a sufficient sample
size to judge.

I dislike the idea very much that playing a tournament with a very
experimental deck can influence your rating negatively esp. when you
loose against a poor rated player (not necessarily a bad player!). Of
course it does affect your rating if you don´t play enough tournaments,
but when you attend fewer than 8 tournaments in 18 month you better
bring a good deck ;-)

--
johannes walch

 

[Guest]

Archived from groups: rec.games.trading-cards.jyhad (More info?)

"Johannes Walch" <johannes.walch@vekn.de> wrote in message
news:dcv99p$dbs$1@news01.versatel.de...
> Frederick Scott wrote:
>> I agree with many of the points you made above. But your conclusion
>> is just as valid for any other rating procedures as for ELO. So if ELO
>> doesn't work for these reasons, nothing else could either.
>
> I disagree. ELO rates negative performance. The current system doesn´t
> rate negative performance as long as you attend enough tournaments.
> Quite a difference.

I don't know what you mean by saying it doesn't "rate negative performances".
It rates everything. It may not use some performances if they're not on
your top eight list but a player doing poorly certainly pays opportunity
cost for it if nothing else: his score _could_ raise and it doesn't. This
is a cost.

In any event, you also don't explain how you figured out that because
the current system may not rate some tournaments that all that stuff
that bothered you about ELO makes no difference in the current system.
That makes no sense. All that stuff creates good scores and it creates
lousy scores. You need to explain the distinction you're making in
much more concrete terms.

> I dislike the idea very much that playing a tournament with a very
> experimental deck can influence your rating negatively esp. when you
> loose against a poor rated player (not necessarily a bad player!).

Sure. But you're naive if you thinking playing experimental decks makes
no difference in the current system - even if you play more than
eight tournaments.

Fred