Crossed 1600 Blitz !!!!!

Was Preggo banned? I had assumed that he just closed the account.

Congrats

This same guy said almost the same thing to me except he said it’s the average rating of the oppenents you beat and the average of the oponents that beat you. I asked what happens if I beat a lower rated opponent that I randomly got matched wit? Now by that logic I am actually worse by winning a game I am supposed to win hahaha.  

Looking at someones stats, such as average opponent, can give clues about if they're currently over or underrated, but that's just one piece of the puzzle. Your view of ratings is very simplistic, and frankly, it seems you don't understand them at all. For example you can accurately have a 1500 rating even if all you do is play 1400 rated opponents. This is possible because, simply put: the math allows it.

But it's true people can have certain practices that will pump up (or deflate) their rating. One of these things is playing a lot of rematches. To get an accurate rating, among other things you should play random people, and play often.

Other stats to look at is a person's most frequently played opponent, peak rating, and their rating history graph, but mostly if a person plays often, and random opponents, then their rating is very close to being accurate (which just means accurately predicting future performance).

Yeah, that's me.

And no, preggo wasn't banned.

It says on the page whether or not a person has been banned or closed it themselves.

Thx for the encouraging words guys. Things are a little rough in real life right now so I am not playing well and have sunk to around 1550 again.

Could go even lower.  However I have noticed a trend that when I hit a high elo I sink around 150 points after that, yo-yo between that range for a while and then get a run in a few months and break the new target.

So hopefully I would hit a new high sooner than earlier.

Having said that I do agree that a truer way to look at your elo is the avg elo of the players you have defeated in the recent games.

That being said I do agree that the true elo is the avg elo of the players you have defeated in the recent past. However the elo against your name is also a measure and I was pleased to have crossed 1600 finally.

Yeah, that's getting a little closer.

If you average the ratings of the players you beat and the players who beat you... then all you're doing is averaging the ratings of all your opponents.

In that case you could increase your "true" rating by simply playing higher rated players. Win or lose your average will go up.

So along with your opponent's average rating, you need to know the win percentage... and that's basically what a rating formal is. Two pieces of information: 1) The difference in rating between two players and 2) the predicted win percentage.

That's how you get tables like this

Rating diff = rating difference
H = expected win percentage of higher rated player
L = expected win percentage of lower rated player

In rating systems, if you preform better than what you're expected to, your rating is adjusted upwards.

If you preform worse than expected, it's adjusted downwards.

That's why you can have an accurate rating by exclusively playing people rated higher or lower than yourself. For example from the table we see if you consistently score 64% against players rated 1400, then you're rated 1500. That's your real rating (although we're assuming those 1400s are playing all sorts of opponents and that their ratings are accurate).

The first few opponents of a new account are always lower rated. Because I only have a few games it messes up the averages.

But look at what I said. I said one of the key things for an accurate rating is playing often. If you see a user like me with only a few games, and all played on 1 day, then of course it's not a reliable rating... it might happen to be accurate, but you can't trust that it is.

That's why you shouldn't trust my rating, not some silly fixation about average opponent rating.

Although if, after 50 games, I was anything below 1900, it would definitely be wrong. It would mean I was having a bad day, or possibly playing drunk as I usually do

So, after 50 games, would I be rated 2000? Maybe not, but I would hope so. My peak (in both bullet and blitz) is 2100.

Oooh, you're saying my real rating is 1700?

Ok, if the bar is that low, I have nothing to be afraid of

But I don't like to lose, I like to win

That's why I play unrated games a lot, because you get to play a lot of noobs.

But ok, I'll play a few hours later today.

I can't promise I'll do 50 in a row, because that's probably something like 4 hours at least, and I'm not going to commit to 4 hours

This is why when I check someone’s rating I consider their glicko rd as well. 

+1

Ok, so I'm just about drunk enough to start playing.

For reference here's my current stats (picture below).

Through the magic of not understanding math, @bishbash1 asserts that my "true" rating (whatever that's supposed to mean) is 1744. He can be so sure even though I've only played 11 games. Good for him.

Hopefully I'm able to play for a few hours and we'll see how that works out.

 congratulations! i have had a similar feeling recently.. i switched to 10 minute controls and every now and then i can tag a 2000 rated player... i know they aren't playing their best or anything.. and its just online blitz.. but still.. a nice feeling