Ratings; Upward trend?

Do you think the ratings are inflating? If so what are some theories as to why?

any observations would be nice.

http://www.personalchesstraining.com/rhp_statistics_may_2008.html

http://www.personalchesstraining.com/rhp_statistics_september_2008.html

Yes in some categories, no in others.

I'd be surprised if there is ratings inflation on this site. The starting rating for newcomers is quite low (1200) and the biggest gain/loss of points from the pool is when an engine user gets booted off the site after reaching a 2000+ rating.

I think it's quite easy to prove a rating inflation, or a rating deflation.
The average of all active players shouldn't rise or fall during a period of time.
If it does then (I think, in order to avoid inflation/deflation of rating) the rating system should be revised in one way or another.

Originally posted by tomtom232
Do you think the ratings are inflating? If so what are some theories as to why?

any observations would be nice.

Mine is suffering from the credit crunch 🙁

Originally posted by Fat Lady
I'd be surprised if there is ratings inflation on this site. The starting rating for newcomers is quite low (1200) and the biggest gain/loss of points from the pool is when an engine user gets booted off the site after reaching a 2000+ rating.

The amount of rating points that a newcomer brings to the system is their rating at the point that they lose their provisional status - adjusted for the net change they have caused in the rating of their opponents over the first 20 games.

this is covered in the 'Ratings Inflation and Deflation' section of http://en.wikipedia.org/wiki/Elo_rating_system.

"Practical approaches

Because of the significant difference in timing of when inflation and deflation occur, and in order to combat deflation, most implementations of Elo ratings have a mechanism for injecting points into the system in order to maintain relative ratings over time. FIDE has two inflationary mechanisms. First, performances below a "ratings floor" are not tracked, so a player with true skill below the floor can only be unrated or overrated, never correctly rated. Second, established and higher-rated players have a lower K-factor.[7] There is no theoretical reason why these should provide a proper balance to an otherwise deflationary scheme; perhaps they over-correct and result in net inflation beyond the playing population's increase in absolute skill. On the other hand, there is no obviously superior alternative. In particular, on-line game rating systems have seemed to suffer at least as many inflation/deflation headaches as FIDE, despite alternative stabilization mechanisms."

K-factors are also covered in the article(they adjust the sensitivity to winning and losing in terms of the number of points exchanged-paraphrased from wiki).

Originally posted by FabianFnas
I think it's quite easy to prove a rating inflation, or a rating deflation.
The average of all active players shouldn't rise or fall during a period of time.
If it does then (I think, in order to avoid inflation/deflation of rating) the rating system should be revised in one way or another.

the average will rise until the pool 'matures'. and even after that it'll fluctuate with the public interest in chess. a 'fischer boom' brings a wave of low rated newcomers, who will reach their final strength in 10-15 years or something like that. there will never be stability, the pool is in constant slow change.

The first draft of the RHP Statistics program just gathered data and presented the results with a graph. To keep track of this information I put it in a database table. Now, all you see is the raw numbers, but I'll add the graphs, trend analysis, harmonic smoothing, and moving averages to this program later. BTW, the old graphs are still available.

http://personalchesstraining.com/main.php?request=rhpStats

There is no trend in the data. The mean and median are both flat. The distribution is skewed left of center.

There have been many students involved in our project that have increased their ratings, but these are individual scores. The constant influx of new members with low ratings keeps the overall flat.

With the new database table we will be able to plot these numbers in a time series where we can determine trend, cycle, seasonal movement, and randomness.