also perhaps some missing feature information about the random mixing. One an put lower bound and upperbounds, and pick among a random list of ratings filtered by that. So there are strates. And psychologically (chess cognition and learning, and fun level) there seems to be a +- 200 layering of the mixing. My serial opponents don't need the rating after a while, as the games speak for themselves. but for new pairings, choosing beyond that is either boring or lots for work, and I put it some one magnanimous player that kept enduring my games, but it was a lot of conscious work at my level of experience (compensating missing experience and its intuition equates to lots of conscious variation candidate exploration. (good for learning that computation is not the way....).
Technically not strates at the population level. Rating being continuous (approx. in silica), it might be a strate for me at current time, but the center of that is all over the place at the population level. bound range mixing for the individual in the population. but it all connects.
@dboing said in #39:
But I don't know the proportion of cheaters necessary to change the population basis of such competitive random pairings dominated rating system. I think it only matters because of personal identification. And that is not applicable.
It is an ethical issue. It is one thing to claim unknown unvariables it is something very different if deliberately misleading the player base either as a policing or independent action.
>
> But as random paring expected/ing player, I am not sure that the crude averaging over many skills rating system is flawed or distorted as such. It is a crude promise of game difficulty level. I don't expect it to be a destiny or self-esteem value.
See this is the point. The rating itself has a meaning attributed by the players not the site. They do have agency after all and if the site keeps distorting this with technology then the player base will walk away. And when they do walk its because of unsatisfaction with the product as a value experience. It is after all the players who voluntarily give their time to the site for the lived experience. Clearly infinite random pairings create information loops that distort reality for participants that much should be obvious by now. Please don't shoot the messenger.
I don't see as clearly your assertions. But I will ponder on the notion of "agency" which I don't really understand and could be what I am missing.
I am not understanding also the product notion very well. I view the population of players as being the site, not the technology, which i understand in abstract form (mathematically), not to induce the loops you might be referring too.
Yes competition is loopy already in itself. competitive ratings are loopy by design in the sense that it is based on the pool of players. Even engine ratings are like that. no assurance of unbiased exploration of chess. I think an external reference system based on isolated measurable skills would be more satisfying from the feedback loop possible effects.
but the loopiness in not the monopoly of online chess. Actually onine chess using glicko, is better defined measure aware of time series of games (not relevant here).
The technology point is not clear to me.
@EmaciatedSpaniard said in #40:
There is no merit in a certain rating.
Please....It is the whole basis upon which chess players measure each others abilities. It is a bit like saying there is no merit in the difference between doctors.
It is a crude measure. Traditions have that tendency to render conventions appearing as unmovable truths.
Even gravity, for which we only know how it behaves when 2 masses are evolving, we don't know what that is. But talk about it long enough, and you will think you know what gravity is. As if it existed without 2 masses involved. Well we can assigna a field to it mathematically, but the only experience of it, or measure is when there is interaction. We have a construct that works, and can be held a placeholder for the next many reproducible tests.
Maybe wrong analogy, because of the precision attainable from the existing to date reproducible properties ad infinitum (we hope, that there is a law of gravity, i am not making a good point).
anyway. to physicists, using the word gravity often enough gives it reality. While it should not (maybe). Mathematically they know that Newton's law is always about 2 masses. I think that gravitons might be the answer to that. But they did not need the graviton, to make gravity palatable. usage frequency has that tendency in our minds (for any word being communicated).
@dboing said in #45:
> It is a crude measure.
Not really. The gravity analogy does not work at all. There are actually limits both physically and mentally that human beings can reach. Discussing the 1 0 pool on lichess it has been pointed out that a 2000 type player should not consistently be able to defend at 2600 strength, sure they might in a few games, but not consistently. If they do then they are really playing at 2600. There are far to many players in the 1 0 pool who play with speed, defending and attacking with precision compared to their rating.
In my opinion,as a speculation, if you ran a test on a data set of 1 0 Lichess players form 3 years ago and then compared them to now you would see some glaring discrepancies. And I would be willing to bet that those differences would be purely technological in nature ie nothing to do with human skill.
@Firegoat7 said in #46:
> Not really. The gravity analogy does not work at all. There are actually limits both physically and mentally that human beings can reach. Discussing the 1 0 pool on lichess it has been pointed out that a 2000 type player should not consistently be able to defend at 2600 strength, sure they might in a few games, but not consistently. If they do then they are really playing at 2600. There are far to many players in the 1 0 pool who play with speed, defending and attacking with precision compared to their rating.
>
> In my opinion,as a speculation, if you ran a test on a data set of 1 0 Lichess players form 3 years ago and then compared them to now you would see some glaring discrepancies. And I would be willing to bet that those differences would be purely technological in nature ie nothing to do with human skill.
The question of population rating evolution on the site has been raised a few times that i recall. But the population size has also risen a lot. And only invoke technological something (the site or the players) as the only explanation is a choice.
Who said that chess population skills with increasing demographics was at steady state. Have we completely uncovered the whole opening tree, and have all individuals been equally experience in all the trails of that forrest.
Also, do all capable athletes who may also have other make a living traits adopt an competition athlete carreer with all the sacrifices it might entail.
The gravity analogy was a psychological comment about how we get attached to words that we repeat often in exchanges.
That attachment may not be harmful to understanding, like in gravity (it even might support intuitive thinking), but OTB ratings have always used pairing based measures. The demographics are a way different scales, and the proprotion on non-tiered pairings is way higher.
Also the initial conditions (1300 Fide ELO or 1500 lichess glicko) vary. The populations are not really comparable. And I don't think one can claim that chess player population is at steady state. There might be increased awareness of chess....
I am saying that we don't know, and being adamant about one hypothesis only seems like putting blinders and might coming from some moral ground before the facts. I am also guessing that. might be wrong. I don't have your perspective (background).
afterthoughts.
Also, may the comparison is flawed. online chess is not a big tournament. It is a community of chess players playing games that are recorded whether they would make it in a tiered tournament or not. The aim (despite the lobby inflationary proportion of ladder stuff), is not to isolate one best player, but that all player get to play the chess they enjoy. I wonder how much of the lichess population is actually into the ladders...
I guess I was right to point to the psychology of words like rating. Stuff might be hidden behind. assumptions to their purpose. I don't care about pyramidal validity. There might be technological bumps, and some bot population, does not change the quality of chess I would get at my level. The bots better be at the level that are appropriate. I think the hidden assumption is purpose of rating. OTB seems to be weeding out and isolate. tournament. and meta tournaments. all pyramidal. Again, here there is no who's who. in spite of the leaderboards thingies, those are motivator extras... (taking too much place i think).
I don't know I asked this a similar question a couple of weeks ago. A 1600+ beat me the other day I'm a hundred lower and he gained 16, yes 16 points for beating someone rated a hundred lower that's never happened to me I can only think he's just come off question mark xxx I don't care but if it's a conspiracy theory as a schizophrenic like myself can believe a hundred rated player gaining 16 points against me makes me question??? Wtfigo??
There were people in the forum tonight who said they knew how to manipulate your rating to whatever you wanted on some link ,did you see that li chess arbitrators?? xxx
This topic has been archived and can no longer be replied to.