16 Oct 05
Originally posted by PeakiteDepends on how many games they play and how fast they play them. My rating has only hit 1700 once in over a year, so if I had to be limited by a 90 day highest rating, it would pretty much eliminate my chance of winning any bands that come up at that time. Not everyone is going to shift up to higher bands, it's the few unlucky ones who will get stuck with it.
If everyone is shifted up, you won't be playing people who are rated more highly - they'll be in the next band up as well.
17 Oct 05
Originally posted by marinakatombThere are examples where this rule isn't sufficient:
[b]2. Any player entering a banded tournament should have completed a minimum of 50 games. It is far too easy for new players to enter banded tournaments before they have reached their true rating which means 2100 strength players can enter 1300-1350 tournaments which is just stupid in the extreme and happens FAR too often![/b]
Tournament 774 - http://www.redhotpawn.com/profile/playerprofile.php?uid=162924
This is a 1600-1699-tournament where scrappie now has won one of the groups and has advanced to the second round. He has now completed 70 games with a rating of 1914, but may not yet have reached his rating peak. He may very well have joined the band in good faith, but he will be 2-300 rating points higher then the others when round two starts.
What can be done about this? Some alternatives:
1) Do nothing, and accept that good players MUST start at 1200 and pass you on the way up.
2) Automatic dismissal. When a player reaches a rating X points higher than the upper band they are removed from the tournament. Any finished games in active round is set to lost, but with no recalculation of rating. All remaining games are stopped with the points awarded to their opponent - but with no calculation of rating. (no need to punish these people (or award their opponents) rating-wise - we just want them out of the tournament)
3) Tournament moderators. They should more or less follow the rules above, but give the climbers a chance to defend themselves - and thereby open for exceptions from the rules.
17 Oct 05
Originally posted by ViAI was all in favour of tournament moderators as well. I still am. But some people seem to disagree with this. Personally, i'd happily give up a little time to weed through the banded tournaments. Personally i think that Russ would do well to have someone else start and maintain these tournaments. A lot of the time, a player has exceeded the band by 50 odd points before it's even started! With a little common sense, these tournaments could be much better maintained than they are. Russ obviously doesn't have time to do this himself, we have volunteers doing other things on the site, why not this?
There are examples where this rule isn't sufficient:
Tournament 774 - http://www.redhotpawn.com/profile/playerprofile.php?uid=162924
This is a 1600-1699-tournament where scrappie now has won one of the groups and has advanced to the second round. He has now completed 70 games with a rating of 1914, but may not yet have reached his rating peak. He may very ...[text shortened]... ive the climbers a chance to defend themselves - and thereby open for exceptions from the rules.
- Joined
- 07 Jun 05
- Moves
- 5301
17 Oct 05
Originally posted by marinakatombYour suggestion is flawed in a number of ways.
[b]1. [...]
Please only enter this tournament if you genuinely are this strength.
2. Any player entering a banded tournament should have completed a minimum of 50 games. [...]
3. Players are only eligible for a tournament by their highest rating in the last 90 days[...]
Thank you.[/b]
1. If I cannot enter banded tournaments at my current rating, then I cannot enter at all. I have to wait until my rating stabilizes. I cannot even enter at what I expect my rating to be, as my measurements from OTB, or other sites, do not match my expected rating here:
Originally posted by ark13 in the other thread
Now I'm a 1500 OTB .... I entered a 1600-1649 banded
Yet ark13's rating has been as high as 1901 - Note I am not critcising you ark13, you entered in good faith that you had reached your peak. I am just using you as an example that OTB ratings do not match this site.
So given that I know my OTB rating, I have to wait until my rating exceeds that by 300 points before entering a tourney? I played in an OTB swiss tourney at the weekend. My performance would have put my rating in the middle of the next section - probably improved by playing slow games here. Does that mean I should have been kicked out of that tourney too?
2. Part of the reason I subscribed was to play tourneys. I will join one when I have more time. I play slowly, as this is a correspondance site, and I don't want to lose by blundering. I do my blundering on a blitz site. Here I want to lose by being outplayed. Given that I want to play slowy, 50 games seems unreasonable, unless you want to refund my subscription until I get there.
And how do you propose real life 2100 rated players get their RHP rating up to 2100, other than by playing games. If RHP ratings are overinflated, as my response to point 1 suggests, then actually, more strong players should wipe out 1300 tournaments on their way up so that those in them have more realistic ratings.
You seem to object to strong players with low ratings playing in these tournaments. While I can see the point, given that strong players are more likely to want to play someone of their own level, then you would have to propose some system for them to enter higher rated banded tournaments. This seems just as open to abuse. The only real alternative is to let them play and progress as quickly as they can. Time limits and number of game limits just get in the way.
Yours in Opposition.
20 Oct 05
Originally posted by gezzaPerhaps players should be allowed to enter banded tournaments up to one band higher than their rating would ordinarily allow. This way players can opt for a more challenging tournament, or enter a higher band if they suspect they are underated.
Your suggestion is flawed in a number of ways.
1. If I cannot enter banded tournaments at my current rating, then I cannot enter at all. I have to wait until my rating stabilizes. I cannot even enter at what I expect my rating to be, as my measurements from OTB, or other sites, do not match my expected rating here:
[text shortened] ...
Restriction:
- Players may not use this to enter more than one tournament from a single group of rating banded tournaments.
20 Oct 05
Originally posted by ViAi believe scrappie had not completed 50 games when this tournament started.
There are examples where this rule isn't sufficient:
Tournament 774 - http://www.redhotpawn.com/profile/playerprofile.php?uid=162924
This is a 1600-1699-tournament where scrappie now has won one of the groups and has advanced to the second round. He has now completed 70 games with a rating of 1914, but may not yet have reached his rating peak. He may very ...[text shortened]... ive the climbers a chance to defend themselves - and thereby open for exceptions from the rules.
20 Oct 05
Originally posted by marinakatombWhile I consider that your ideas have good intentions there are drawbacks.
[Please only enter this tournament if you genuinely are this strength. Players are requested to respect the banding of this tournament, and not enter lower bands than their true strength.
Obvious you might think, but as things currently stand, players appear to feel it is acceptable to beat up on lower ranked players in these tournaments! It is not acce ...[text shortened]... problem because the banded tournaments are not currently fulfilling their purpose.
Thank you.[/b]
Advantages of having a low rating:
Easier banded tourneys.
Easier clan matches.
I'm sure, myself included, there are players who resign in games they are losing to save their rating from increasing too much.
We need incentives to get your rating higher not keep it low.
The faster you play the higher the chance a massive rating spikes. Your suggestion would penalise those players who play fast and hit a lucky streak. I've also had situations where two players rated 2000 have resigned in the same week in both (black and white) tournie hardcore games leading to a 120 point rise in my rating. this effectively rules me out of banded tournies for the next 30 days. If what you were saying came true - it would be 90 days.
My solution is to have an average all time rating. This rating would be used for all tournaments.
21 Oct 05
1 edit
Originally posted by marinakatombThat's the stupidest post I've seen on this subject thus far.
Anyone worried about going over their band by one point, eg, a person hits 1601 for the first time and then drops below that level agian but is stuck playing 1600+ tournaments for 90 days, sorry, that's just the way it's got to be! I will no doubt suffer from this myself but something MUST be done about this problem because the banded tournaments are not currently fulfilling their purpose.
Thank you.
Case in point:
Someone like me, who is a decent 1300, maybe 1400 player, will be stuck playing in tournaments I can never win because a while ago I had a few good victories and some time-outs against higher rated players?
Your proposal will stop some of the abuse, but will also penalise lower rated players who had rating spikes.
Ludicrous.
21 Oct 05
1 edit
Originally posted by invigorateExactly. This would be the only good solution. I've been proposing just this for months now.
My solution is to have an average all time rating. This rating would be used for all tournaments.
No new player can realistically know what their rating would be on here. They wouldn't know the player strength here. Also, tournaments was one of the big incentives for me to join - if anyone told me I couldn't join tournaments for a few months because my rating needed to be established first - I probably wouldn't have subscribed.
The only 'abuse' we can realistically hope stop is players leaving the site / mass resigning etc. and then coming back at a very low rating (dustnrogers...)
I believe an average rating can be used in conjunction with the current rating system. If we can have a system weighted by time spent in a certain rating range, it will solve many problems.
Example:
A player plays here for a year and steadily builds up a highest rating of 1700. For 7 months of the year he had reached his peak band between 1600-1699. The player leaves the site and his rating falls to 1100 in 2 months time.
Now if we had an average rating system using weights, this player's average rating could be calculated as this for example:
14 months total time.
1 month in each band (incremented by 100) up to 1499
2 months inside 1500-1599
7 months inside 1600-1699
2 month downward spike 1700-1100
Multiply the time spent in each band by the average of the band:
1/14*1250 + 1/14*1350 + 1/14*1450 + 2/14*1550 + 7/14*1650 + 2/14*1400 = 1536
This will be the banded rating of the player.
Problem solved.
I've said before, I'm not a statistician - so my reasoning may not be perfectly accurate, but I think a similar solution would be very effective in stopping any rating 'manipulation' or 'abuse'...
21 Oct 05
You'll never fairly solve this problem. You can only limit it a bit.
1) You must be non-provisionally rated. Not a big fix, but it helps the lower bands.
2) If your rating exceeds the maximum rating by 100 + 100*Rnd# during the tournament, you will be removed from the tournament at the end of your current round and the second place person in your group will advance. If it is the final round, the second place person will be declared the winner.
So in a 1325-1349 tourney, you would be removed if during a round you exceeded the max:
Round 1) 200
Round 2) 300
Round 3) 400
Why the sliding scale? Tournaments are long. This allows for improvement of the player.
Just something to bat around.