i think that a % is just another number and a trend line is just a graph . . .you'll know you are getting better when you are . .but then . some days you'll make such stupid mistakes you'll actually know you are getting worse . .. but then . . anyway . .the point is . .that the numbers will go up and down . .all of our numbers do . . .have fun . .relax . enjoy . . . try this - create a "treasures" folder, and a "learning" folder . .. keep your noteworthy games in places where you can go and look at them . . .the best evidence is in your play . not in your number . . but that's just my opinion . .. everyone will disagree with you as you will soon see
Originally posted by incandenza Your % wins could go up just by choosing to play very weak players, but that doesn't mean you're improving.
Your rating also could go up just by choosing to play much weaker players. It won't grow as fast, but it will grow.
If you play pretty equally matched players generally and you see a general increase then that should be a fairly good indication. That's essentially what I look for, but the rating, of course, isn't perfect.
Originally posted by PsychoPawn Your rating also could go up just by choosing to play much weaker players. It won't grow as fast, but it will grow.
I don't know the details of how the ratings are implemented on this site (other than looking at the formulas), but that shouldn't be possible if they're implemented correctly. The amount you gain should round down to 0 with a sufficient rating difference.
If you win against any player exactly the % of games you're expected to win, your rating shouldn't change.
Your rating only goes up when you win at a higher rate than the calculated win expectancy.
In fact that is the whole principle that makes the ratings meaningful. They predict an exact win percentage between two players. When the reality is that the win percentage is different than that, the ratings change.
Originally posted by incandenza I don't know the details of how the ratings are implemented on this site (other than looking at the formulas), but that shouldn't be possible if they're implemented correctly. The amount you gain should round down to 0 with a sufficient rating difference.
If you win against any player exactly the % of games you're expected to win, your rating shouldn' ...[text shortened]... s. When the reality is that the win percentage is different than that, the ratings change.
You likely won't win exactly the % of games you're expected to win.
I think in theory you wouldn't be able to gain rating by playing lower ranked players often, but in practice I think it's definitely possible.
Technically, that's true about any rating level though since you are expected to lose more against higher ranked players.
I agree that the ratings are somewhat meaningful, however, they are only meaningful to a certain degree.
i'm not a guy, but it's ok. anyway, i think i understand the rating system pretty good. my only point is that in the last month my rating has gone from 1600 to 1200 something and back up to 1659. so, would you be saying that i got worse and then better again that fast? no, i doubt it. i bet you would agree that we can have trends in our play that might be affected by a lot of other factors that don't have anything to do with getting stronger or weaker. here a just a few factors: maybe i tried a new line of openings and didnt know them so well and dropped points. how about once i got weaker i got matched with a lot of players who were 300 points or so below my "true" strength. OR . .gee . . .maybe it was because i got kind of crazy and played 280 games at the same time and lost some focus and concentration . . .gee, ya think?
Well, I was referring more to the other two posters (although they might not be guys either). Anyway, the phrase "you guys" is kind of gender-neutral anyway.
The rating can only measure your actual performance in games, not some kind of theoretical best performance if you had played the right opening and you weren't tired and you didn't have too many games in progress, etc. So I would say, yes, in a sense you did get worse and then get better; not in terms of your true ability, maybe, but you were setting up circumstances where you would lose more.
The rating is still reflecting your performance as accurately as possible. I don't really think you can fault the rating system for actual swings in your performance.