Are Online Chess Ratings Accurate? What you must know

If you’re just starting to play chess online or if you’re thinking about taking it up, then you are probably aware of the fact that you can get online chess ratings supplied by the online group that you join. What you may be wondering is if these rankings are accurate and whether they mean very much anywhere except in the virtual world of chess? Well, the answer is complicated, but we’ve got to the bottom of it for you.

Are online chess ratings accurate? Online chess ratings are perfectly accurate within their own system assuming players are playing fairly over an extended period of time. The problem is that although they are accurate, online chess ratings are not the same as offline real world ratings.

Let’s look at what online chess ratings are actually worth, and how they compare to offline ratings.

Are online chess ratings accurate? - ChessPulse.com
Are online chess ratings accurate? – ChessPulse.com

What Is It That A Chess Rating Measures?

Before you jump up and say “skill at chess”, we’d like to point out that this is, of course, what you’d hope that a rating system would measure but sadly, this isn’t a measurable quantity (or at least, not an easily measurable one).

Firstly, there’s the problem of defining “skill at chess”, really the only skills we need to play chess are an understanding of the rules and most players have obtained that skill before they start worrying about ratings, rankings and the like.

Defining Skill At Chess For Ratings

So, what do we mean by “skill at chess”? Well, each ranking system defines this somewhat differently though they all appear to have some areas of overlap. They tend to define it as “skill at chess relative to the skills of all other players”.

This is actually a pretty reasonable way to measure someone’s skill when everyone has the basics completely covered – to compare them against each other.

The Challenge Of Large Group Comparisons In Chess Ratings

But every set of online chess players is made up of very different people (and given that there are a large number of online chess sites and most of them require some sort of fee for you to join their ranking system, people don’t tend to join too many different sites at once because it’s expensive and there are always more players on one site, why would you need to be on multiple sites?).

So, you’re not comparing apples to apples and oranges to oranges but one group of players against another different group. Now, it is possible that if you take a large enough group that when you average them out – they ought to be similar to all other large groups of chess players.


Groups Tend To Converge On Certain Patterns

But, in our experience, when you get a group of people together, all doing a similar activity – you tend to find that, over time, they tend to converge (a little or a lot) and in chess this means they tend to prefer certain openings, they tend to develop their pieces in a similar fashion, etc. and this may mean that the players on Chess.com have a very different set of skills to those on say Lichess.org.


So, Are Online Chess Ratings Accurate?

This is kind of where we came in. Yes, online chess ratings (assuming the playing pool is not contaminated with too many bots/cheats) ought to be very accurate indeed in terms of whatever that site is measuring.

Computers will be able to count wins, losses and draws and compare them in pretty much any algorithmic way they are told to against other players. They can analyze the moves within your play and see if they are of sufficient quality to fall into a certain rating and so on.

The Limitations Of Computer Grading Systems

But what the computers on any given online site cannot do (at least for now) is see the rating systems of other sites and make value judgements based on the pool of players using those sites and their individual rating systems.

So, if you play on Lichess.org, for example, their computer system can give you a super accurate rating when compared to other players on Lichess.org but what it can’t do is give you any indication, whatsoever, as to how you compare to players on Chess.com.

Real World Evidence Supports This Problem

This supposition has been borne out by endless attempts by players on individual online chess websites to come up with conversion tables translating your rating in one system to that in another. These have all turned out to be quite miserable failures. Almost certainly because you can’t readily compare apples to oranges – they’re both fruit, they’re both tasty and they’re both good but they’re different.

The same is true for chess players. Electing to play online with LIchess.org doesn’t by any kind of definition make you a better player than someone using Chess.com. That should be pretty obvious but as we already noted, over time, you will start to develop some traits of a Lichess.org player and this will make you different from the player with Chess.com traits.

How To Rate Yourself Against Another Player

The only way for you to find out how you really compare against each other is, well, to play against each other. There can be no disputing your respective skills if you play say, 10 games against each other, if one of you thrashes the other 10 to nothing, then they are clearly the better player. If, on the other hand, there are 10 draws in the mix, well, you’re of a similar standard.


This Problem Is FIDE’s Problem Too

We have discussed FIDE’s online rating system in depth in this article here and there we noted that one of the reasons that FIDE doesn’t see your online FIDE ranking as the same thing as your standard FIDE ranking is the problem of cheating. After all, it’s much easier to cheat online (even with all the anti-cheat technology FIDE employs along with most online chess sites, to be fair).

But that’s not the only problem. Once again, you have a select group of players who elect to play their online chess through FIDE’s My Arena system and as you would expect, over time, this group just like the groups on Lichess.org or Chess.com are going to become a more homogenous group of players.

Thus, even if you could eliminate all the cheating from online chess, it still might not be sensible for FIDE to automatically assume that they could translate your online ranking into your tournament ranking. This is because those playing in “real world” tournaments are also likely to have certain homogenous tendencies, but they are going to be different from those people playing online.

The only way to determine whether the online rating actually stands up in “real world” tournament play is to go and actually play in tournaments.


Does This Mean Online Ratings Are Worthless?

Absolutely not. Though you may not have any easy means of comparing one rating system to another, as we’ve already noted – within the group of players you regularly play online with, they are a very accurate way of determining your progress within the game and your group.

This is very useful information. As we said earlier, playing Lichess.org or Chess.com or FIDE My Arena doesn’t make you, by definition, a better or worse player than someone on another platform, it just makes you different and the only way to tell if you are better or worse than someone is to play them.


Conclusion

Are online chess ratings accurate? The debate on whether online chess ratings are accurate has raged for a long time and the answer is that online chess systems from Chess.com to Lichess.org to FIDE’s My Arena system are all very accurate at measure your ranking within that system. What they are absolutely no use for, at all, is giving you a ranking that can be translated into any other ranking system.

This shouldn’t (and, indeed, doesn’t) matter if you always play within a particular online environment. Your Lichess.org rating is far more important when playing on Lichess.org than any other rating scale as it’s a ranking that you gained by play within their system. It may matter, however, if you are looking to become a FIDE ranked player and if that’s the case, the only way to obtain your rank is to play in real life chess tournaments managed and run by FIDE.

Scroll to Top