Wow, found a loophole in Lichess!

I recently made a post in this forum about a hypothetical situation where if I was rated 2600 but got a provisional rating of 1800 and started playing in casual pools I could easily beat 90% of the people in casual as you only play people rated +/-500 rating points. Now in this scenario, lichess knows that I have beaten players rated 2300 90/100 times which means it's clearly knows that I'm definitely rated higher than my provisional rating of 1800 but it would still pair me up with players +/- 500 rating points bracket for 1800, which I can even adjust to my liking from both ends, which means I can intentionally only play players lower than my rating. This, in rated games, is called sandbagging which means that you have intentionally deflated your rating to play players in lower brackets, and this is not allowed in Lichess according to their TOS, however in casual pools many players do this, they don't play rated games for years and years but are insanely active in casual pools.

In my honest opinion I don't think it's wrong to use this strategy even though I would never really do this as it's just not my thing to play like this but I wouldn't care if someone else was doing this so to get other people's opinion I posted on this forum and a few people pointed out this was a form of "trolling" and shouldn't be allowed here. This clearly is a loophole as lichess would probably know about this and is letting this happen.

Now before you say that it's just casual games who cares, I'd like to point out a few things.
1) Players who do this know exactly what they are doing, they want to play lower rated players to increase their win percentage for reasons unknown.
2) Many people, including me play casual games to improve their chess by playing higher rated players and then moving into rated to see how much we've improved but even here lichess sets a bar that no one can play people rated +500 points so there's always a limit and you more or less know what to expect but when you see a player with a provisional rating it's like playing anonymous chess against a player who could be rated 2800 for what you know so it's not a learning experience as if someone wanted to play a random guy with no rating wouldn't they just play anonymous section of lichess? It's a waste of time for someone who actually takes time out of their schedule to practice chess.
3) I know you might think that these players are also trying to improve their chess by playing casual with deflated provisional ratings, but what would you learn by playing lower rated players and even if they are playing higher rated players in casuals they never play a single rated games to get their rating established, so clearly their motive is not to improve their chess.
4) "Big deal people want to play lowers rated players let them do their thing, why do you care so much?" Well it's a little unfair to people with established ratings, as we can only go lower by 500 points so I guess having a provisional rating and sandbagging in casual has it's perks.

Now, I don't think this a huge problem as you only play a few of these type of players but it sort of takes the fun out of casual chess. I don't mind playing higher rated players but not knowing you're playing someone 600 points higher rated than you never helps you to play better, it becomes a game of hope chess where you hope you are actually playing someone your own strength.

Last time someone mentioned a few changes that can implemented where this can be improved. For instance, lichess could ask you to play atleast ten rated games before allowing you in casual pools if you have a provisional rating like it's in tournaments or it could just simply ask you that "would you like to include players with provisional ratings?" Or lichess could just simply notice that the player with a provisional rating is improving so they could get a 1800?+/-100 or something like that for their rating so when they play casual they get paired with higher rated players untill they actually are playing with players their own strength even if they have a lower provisional rating. There are many options that can sort out this issue.
Tell me what you think about this.

The solution is to play rated games and take each game as serious as possible but at the same stop caring about your rating. Switching off the display of player ratings (possible in the settings) can help those who are too attached to their rating.

@M0r1 said in #2:
> The solution is to play rated games and take each game as serious as possible but at the same stop caring about your rating. Switching off the display of player ratings (possible in the settings) can help those who are too attached to their rating.
Yeah, I sometimes turn off my rating settings too.
But sometimes I want to see how much I'm improving, also if you want to train for over the board chess you can't let ratings affect your chess. However I think in some situations your idea is very helpful, thanks for responding.

being rated provisional IS having anonymous mysterious rating (very akin to playing with no account).

it is having ESTABLISHED rating that's far below your actual strength that is what's considered sandbagging here.

however, an option to prevent getting paired with provisional players in the lobby (incl for casual games) would be something very nice to have in the lichess inventory!

(as it can get especially annoying in the case of chess960 on slow time controls: countless 1500?s will join your seek by accident, who cannot play, or will resign or disconnect just a few moves in, having wasted enormous amount of your time.. so grr)

That was a long post.

When its not rated, it doesnt matter.
If its rated, it doesnt matter either, as the lichess rating is like the monopoly money, but it would discourage lower rated players, so it is forbidden.

I do not think that casual players have issues playing stronger players, specially in the 1800+ bracket. Quite the opposite, they would like to get to play stronger players in an unfair match when their rating is not at stake/

Its a win-win scenario.

The problem with this "sandbagging" label is it becomes cloudy when you want to play thoroughly. Everyone is obsessed with vertical growth, they don't respect horizontal development. I have a rating now of 1759 after 4 games, but I know I am not 1759. I want to start at 1000 and work my way up incrementally (by 10-50 points). The response I get is, "You just don't understand Glicko and how the rating system works." I do, and I understand as you get higher rated the more you should follow this. However, if you are under 1500, there are fundamental parts of chess you need to work on. These are what the real 1750 players know inside and out.

The solution is the same as sandbagging, lose. It doesn't matter if you intentionally lose or you lose trying to make the best move. A loss is treated the same way. To rectify this, what I think they should do is use the second hump to determine sandbagging. This would be the second time someone loses a bunch of games. The advantage here is you have a previous history to compare. If someone like me now loses and loses to a point where the rating goes from 1759 to say 1300, you don't know if I was really sandbagging. In fact, I come in second often in the <1300 tournaments. There is always some player who performs at an exceedingly high rating performance. But my rating stays below 1300, and yet I get my accounts closed because of sandbagging assumptions.

If I legitimately won 1300+ and repeatedly did this with little to no effort, I would never join the <1300 tournaments. I would go to then <1500. Then <1700 and so on. But do I really want to join a <2000 now? Because I can't qualify now for even the <1700 tournaments. Yet, this is 400 higher than what I have struggled with in the past.

This illustrates how sandbagging is too ambiguous of an accusation because the powers at be can't see the true value of lower rated players. Think of it like this, watch kids (under 10) play ice hockey. These kids won't know how to stop and they crash into each other and trip and fall. How do you evaluate success in this case? This is similar to lower rated players. The one who screwed up the least becomes the winner. While you can say this with top GMs too, the players are like Olympic athletes. They have achieved a high standard. The kids playing hockey, and the low rated chess players are just trying to establish a standard.

This is the blind spot not recognized by chess sites. They should focus more on sandbagging from 1700 and upward. Allow players to establish rating points earlier instead of inflating their rating. This will then populate the server with more solid players as they get higher ratings instead of bouncing up and down triggering false positive sandbagging markers. And those players like me won't have their account wrongly closed.

Why would you want to start from 1000 if you already can play better? What is the point to be underrated? I see no good reason. The game doesn't change for different rating, it's only about you and your opponent.

I think, there may be statistical difference for people who play poorly intentionally. Just like it works for cheaters. Because my performance varies a lot and still no trouble.

Some people may be experimenting with garbage openings and ridiculous attacks, or they may be in a non competitive state of mind. Then when they want to play real chess go back to rated. Since multiple accounts are discouraged, casual chess is the way to play however you want.