@Darthpika and Austino: Yeah, I know exactly what you guys are talking about. I took a break from pokemon for a few years and came back last year at states. At states and regionals my combined record was 6-6. But at nats, I was 8-1 in swiss and lost in t128. I remember I was pretty close to getting an invite by only actually preparing for 1 tournament. At the same time, my little brother worked all season, going into regs with one of the best ratings in NA. He was 6-1 in swiss, won the t8 round, and then lost in t4 to a good player. He managed to lose points on the day.
To many points for cities
- Thread starter Jaeger
- Start date
- Status
NoPoke
Active Member
I want it to be a discussion too.
It is always better to have more points than fewer. We both know that so I leave that unsaid as it is obvious. So how can I say that a high rating is a disadvantage? The disadvantage is two fold. If you have spent time at many events establishing a high rating and you then play against a lot of players who have not played many events then elo will be using your relatively well established rating against their recently reset rating. Meet a strong new entrant a couple of times and almost all of the previous hard won rating can dissapear. This is not good, in effect a late entrant who does well benefits from all your previous hard work: all those early events were little more than extra points for you to give away to the stronglate entrants. The other disadvantage comes from elo treating all games as decided by skill alone. This is just wrong and again rewards the lower rated player excessively when they win.
At this point I won't have convinced you that there is a disadvatage to having a high rating as it is still self evident that it is always better to have a high rating than a low one. In order to convince you you have to shift perspective and look at the pair of players taking part in the match. Elo is based upon lots of pairs of comparisons. If the lower rated (new) player has an advantage in how elo treats them in the comparison, then as the system is symetric the higher rated player is at a relative disadvantage. As there is almost always a higher rated player in the comparison then there is almost always a relative disadvantage for the higher rated player.
Summary:
From an individual perspective there is no disadvantage to a high rating. But hte individual perspective is the wrong perspective.
The instant you look at a pair of players the relative disadvantage is there and it is always against the higher rated player.
The rating of any individual is of no importance: only their relative position to all the other players matters. The top X players get invites, they are being assessed relative to the cohort.
---------- Post added 02/02/2011 at 07:17 PM ----------
===
Darth Pika, if the players you wish to catch up do better than you at future tournaments then you will not catch them up. If you do better than them then you will catch them up. Continue to do better and it doesn't take long for you to overtake them. A 100 point lead is not unassailable at the end of Cities, but you do actually have to demonstate at tournaments that you deserve to catch up. They worked for those points and you have to work for yours too. No easy buckets.
It is always better to have more points than fewer. We both know that so I leave that unsaid as it is obvious. So how can I say that a high rating is a disadvantage? The disadvantage is two fold. If you have spent time at many events establishing a high rating and you then play against a lot of players who have not played many events then elo will be using your relatively well established rating against their recently reset rating. Meet a strong new entrant a couple of times and almost all of the previous hard won rating can dissapear. This is not good, in effect a late entrant who does well benefits from all your previous hard work: all those early events were little more than extra points for you to give away to the stronglate entrants. The other disadvantage comes from elo treating all games as decided by skill alone. This is just wrong and again rewards the lower rated player excessively when they win.
At this point I won't have convinced you that there is a disadvatage to having a high rating as it is still self evident that it is always better to have a high rating than a low one. In order to convince you you have to shift perspective and look at the pair of players taking part in the match. Elo is based upon lots of pairs of comparisons. If the lower rated (new) player has an advantage in how elo treats them in the comparison, then as the system is symetric the higher rated player is at a relative disadvantage. As there is almost always a higher rated player in the comparison then there is almost always a relative disadvantage for the higher rated player.
Summary:
From an individual perspective there is no disadvantage to a high rating. But hte individual perspective is the wrong perspective.
The instant you look at a pair of players the relative disadvantage is there and it is always against the higher rated player.
The rating of any individual is of no importance: only their relative position to all the other players matters. The top X players get invites, they are being assessed relative to the cohort.
---------- Post added 02/02/2011 at 07:17 PM ----------
===
Darth Pika, if the players you wish to catch up do better than you at future tournaments then you will not catch them up. If you do better than them then you will catch them up. Continue to do better and it doesn't take long for you to overtake them. A 100 point lead is not unassailable at the end of Cities, but you do actually have to demonstate at tournaments that you deserve to catch up. They worked for those points and you have to work for yours too. No easy buckets.
Last edited:
Austino
New Member
7-1 states week 1
5-2 DQ states week 2
9-2 at regionals (one swiss loss 6th round)
4-3 Nats (technically 6-3 into top cut since I got two byes from regs) swiss
You want an answer to this? Read the next quote.
While someone with a low(er) rating would have gained a LOT of points with that record. No need for an individual to go X-0 to gain a lot of points at bigger events, I used that as a comparison only, not suggesting it is easy, just a comparison.
And yet Lotads brother even losing points would still have been ahead of the lower ranked player who gained on the day.
About my brother, I talked to him a little but ago and the info I posted was a little wrong. He actually did have a single digit gain in points for that tourny, but he did lose 7 points in a states in which he was 5-1 in swiss (losing 6th round) and losing in top cut. But still, that's 5-2 with losses to 2 good players (obviously). I think my point is still valid.
NoPoke
Active Member
@Jaeger: so how would you have liked LoTad's brother to have been treated?
there are lots of possibilities but most (all?) of them look worse....
LoTad's brother performed very close to expected by elo and the system left his points unchanged. Lower rated players with the same record will have gained points with a similar performance on the day.
there are lots of possibilities but most (all?) of them look worse....
- all players on the same record gain/lose the same?
- no one has any points before nationals? (LoTad's states/regioinals record was solid)
- high rated players who lose should be penalised even more?
- high rated players can't lose points at nationals?
- that the only way to qualify for worlds is by winning a states/regionals/nationals
- or other suggestions?
LoTad's brother performed very close to expected by elo and the system left his points unchanged. Lower rated players with the same record will have gained points with a similar performance on the day.
- Status