|
« TAP TO LOG IN
|
GameKnot related: New: rating floor
« Back to forum| From | Message | ||
|---|---|---|---|
|
 New: rating floorYour rating can never fall below your rating floor. The idea behind it is that if you time-out in a lot of games due to some unforeseen circumstances (and the automatic emergency postponement isn't enough to save you), you rating won't drop too much. And more importantly, when you do hopefully get back to playing, you won't have to play hundreds of games just to break even, and your opponents won't be unpleasantly surprised when your playing skills don't match your temporarily low rating. We've been getting a number of complaints from players who fail to notice that their opponent's rating is uncharacteristically low due to a recent drop because of a large number of timeouts, so they feel cheated. Which is understandable, but you can't just dismiss it with "sorry buddy, your own fault, you should've checked their profile page before joining the game". The rating floors should help with this (to a degree). Comments? Objections? Suggestions? |
||
|
kingdawar 02-Nov-10, 13:47 |
Deleted by kingdawar on 02-Nov-10, 13:52.
|
||
|
kingdawar 02-Nov-10, 13:52 |
And you have to consider a lot of hypothetical situations, like player A who has had a strong high, followed by a bad streak donating loads of points to player B for free and then player B can return the favour (both A and B would "benefit") If this is an idea to keep timing out players involved with the site for somewhat longer, then why not. Maybe the top rating from the "past x years" could be used instead of all-time high. |
||
|
Regarding your concern about "donating" points back and forth -- there is certainly more than one way to cheat the rating system, but most of the time it is easily detectable, and then you lose your account altogether. It's a good idea about limiting the highest rating to past X years, or maybe something like lowering everyone's floor 50 points once a year. We'll consider it, thank you. |
||
|
easy19 02-Nov-10, 14:21 |
lets see i took my own rating thingy My Average chess rating: 2163 - 300 = 1863 nearest 10 would be 1860 ? My Highest chess rating: 2295 -300 =1995 nearest 10 would be 2000 ? (my Average opponent rating is 1755 and is quit low this could be taken in consideration's in some way.) my current rating is 2225 so if i now start timing out all my games i would drop 225 rating points. and if my average was taken i would drop 365 points - so it seems that the proposed rating floor is quit Fair for me. |
||
|
kingdawar 02-Nov-10, 14:59 |
Easy question |
||
|
easy19 02-Nov-10, 15:22 |
Heinz (means 2100 top score but current far below 1800) i have no way to look this up but i think GK staff can get this answer for us.. if i have to predict i think Les then 2% of the total player base will benefit from the new rating floor by abusing it if implanted.. |
||
|
ganstaman 02-Nov-10, 16:53 |
Would this cause a rating inflation over-all, and how noticeable would that be? |
||
|
Sounds good to me I don't think it will "encourage" anyone to time out more than they already do. I don't think it would cause a rating inflation. Most games that would be played after the floor was hit (during a downward spiral) would be short games that were automatically started (like mini-t's). In those cases, the winner does not get points anyway, so this is more like a prevention of points lost in the system rather than inflation. Looking forward to hearing more opinions on this proposal..... |
||
|
tactical_abyss 02-Nov-10, 19:01 |
By the same token however,I would also have to admit that NOT changing the rating system is also ok with me!GK is not the USCF and there are differences with different kinds of tourneys that the USCF has including of course,OTB.Since GK is a smaller knit family of players,I can also see that having our own type of independent rating system in many cases SHOULD be different,even largely different that the USCF rating scale. I for one actually do believe that any player SHOULD really do a thorough check of a players profile page and past game history before simply accepting or submitting challenges,so there is no "suprise" or complaint as stated above,due to time-outs ect ect.Its not like it a big job or anything to check a few past game histories and players really should get alittle MORE involved in"knowing"their opponent alittle better before simply going into the game "blind". Just like when I research a potential opponents past game strengths and weakness's.If my potential upcoming opponent has won 98% of 100 Sicilian games as white and his rating is close to mine,you can be assured that I will not be jumping into a Sicilian game with this opponent,but probably go with a different style opening rather than simply commit suicide! For if I didn't do alittle research on my opponents past games,and played the Sicilian and lost...I would feel that it IS my fault for not taking alittle more time to have done some research,so I will not complain,nor do I have a right to! But thats me,not the next player,of course! So either way,it dosen't matter to me with any rating change calculations,although if it DOSEN't change,thats also ok with me! |
||
|
New Rating Floor |
||
|
black_cat_hamlet 03-Nov-10, 02:24 |
It makes perfect sense So would this apply to people who have already dropped below this limit as well? |
||
|
maca 03-Nov-10, 11:17 |
Rating floor I'm also concerned over the idea that a number of people playing currently well below their maximum rating would experience a sudden rating boost at the moment the change were to be implemented. This type of bumps might be rather large, and would be very unnatural features in said players' rating history. I suggest that we eliminate such boosts by setting the reference rating for the floor EITHER to the player's maximum rating, OR to 300 rating points above their current rating, whichever figure is SMALLER. The latter option would lead into the player's floor rating being their exact current rating. Naturally, the reference rating should not remain the same all the time, and should be refreshed periodically to keep up as the player's skill level changes. The suggested idea of making it 'top rating of the last x years' should work fine, as long as the aforementioned correction for the reference is also taken into account whenever the floor rating is being re-considered. In case of GK, the time-frame for these updates, x, could well be one year in my opinion. It's not necessary to make it too long, because most of the players here won't probably commit into the site for periods as long as several years. Regards, MaCa. |
||
|
baronderkilt 03-Nov-10, 15:31 |
Good Idea ... It is like taking Kasparov or Anand and suddenly removing them from the Top 100 list. IMO the best implementation might be to keep the 300 or even 400 as a MAX LIFETIME loss (which could be done by programming, and if a player suffered a massive stroke or something to where they Really lost more playing strength than that and stayed at the 400 floor for a year ... they could write in to you GK for a manual adjustment request to reduce them further ... or just reset their Floor so they could continue to decline to a true rating up to 400 more. In practical term tho, this probably never happens) So again to say, a 400 pt Lifetime Max Loss to lifetime floor. But limit loss in any one Year to 100 points. This is much less unfair to the opponents they then face. And better incentive for the player to climb back to where they should be ?! Or even to 100 points in any period of 6 months, or 3 or 4 months. That's my thoughts GK ... but as they say; I'm fine without whatever you do on it. Sign Me, BaronGot-3-Games-This-Year-But-Enjoy-Them-As-Much-as-others-enjoy-100 }8-) |
||
|
von_gooms_gambit 04-Nov-10, 13:44 |
Deleted by von_gooms_gambit on 07-May-13, 23:38.
|
||
|
grege79 09-Nov-10, 01:28 |
|
||
|
I'm all for a floor.... ratingsgraphlookslikeanEKG |
