Sandbagging Online Games

Agree with wiseguy, I've only played 2 online games so far, won both and my rating is 1450 which is waaaaaaaay over what I think it should be!

Maybe a 20 game minimum before the floor starts. 

Give me a chance to play in the 1000-1200 category (or possible lower, I don't know yet) please 

 It does not really matter to me if they choose a different number; again, I was only suggesting things and I figured a good number would be 200. Also, between Class A, B, C, D, E, Expert, 200 poinst is what determines each class (see below). If a 2000 decides to sandbag some games, s/he'll have to lose some games and get down to 1800; 200 points just did not seem like a big enough drop to make a difference in competition. It can be lower (150 or 100), or higher (250). I do suggest not going over 250 as the bigger the difference, then the worse it can be.

Here's something I found on the net...

United States Chess Federation ratings

The United States Chess Federation (USCF) uses its own classification of players: ^[6]

• 2400 and above: Senior Master
• 2200–2399: Master
• 2000–2199: Expert
• 1800–1999: Class A
• 1600–1799: Class B
• 1400–1599: Class C
• 1200–1399: Class D
• 1000–1199: Class E

 I think you just proved my point.

"Your honor, I rest my case!", your witness.

This is why I believe a ratings-floor regarding tournament rating should be done. It doesn't even have to apply to anything else if that is what the staff determines to be for the best, just tournaments.

To Wiseguy00 and Trant, this ratings-floor has nothing to do with the little guy having inflated ratings. It will eventually deflate as you play more games. It is to keep the Big-Fishes (or Sharks if you prefer) from gobbling up all the little/medium size fishes. So, I don't think you, and me, have anything to worry about. The only ones that would have an issue with this idea are those that chum the water, if you know what I mean.

I think that the velocity of games on chess.com, along with other variables such as the ability to pick your opponent and opponent timeouts may imply a higher variance for chess ratings here, so a number higher than 200 may as a result make more sense.

This is the part that I trust that the chess.com staff will do more analytically than I do in a forum posting. :-)

What about glory?

Sweet, meaningless, online glory.

It's because tournaments take

so

long.

Resigning (or actually, loosing) before the 4 moves has been done actually won´t loose you any rating at all, so if someone is resigning gamest against you on the first 1-3 moves, they aren't doing it to sandbag.

I have another theory - people like long-term struggles that tournaments provide. It's nice knowing that there's a final goal, to give the games that you're playing some context. And then after all that, you find that in your 1400-1600 rating section that you have one user who has since rocketed up to 2350, and is on pace to win it. You do a little more digging, and find that just before the tournament, he lost enough to get down from 1900 to 1550.

It's ok when you lose because the other guy just played better. It leaves a bad taste in your mouth when you feel like the other guy just gamed the system and really was going to win almost no matter what.

To Gonnosuke: These are my stats through 16 games playing opponents with an average rating of 1404. My highest is 1625 and I am now 1619 (this would mean that I cannot join a tournament lower than 1425 if the floor of -200 is implemented, which I would have to sandbag down to if that is what I wish to do) ~ Why is this such a bad thing to have done to me or anybody else for that matter?. It's not like it will only apply to some and not others; I know 16 games is not enough or a lot to accurately determine my ratings. I actually believe it has me approximately 150 points over what I really am, as I am being very realistic here. I invite anyone to go over my games and scrutinize it for anything that might be questionable; I play fair and square. I only ask for the competition that I play to do that same, which, in my opinion isn't asking for a lot.

I think "mass-confusion and outrage" are a little too much. Have a rating points next to your regular one that when you scroll over gives you the description of what it means. As with any changes, it will take some time for everyone to get use to it, but it will happen.

Rating for Hammerschlag

Games

 I couldn't agree more. I like the fight, the good fight. I don't mind losing as long as I know it was a fair fight. Don't challenge someone to a fist fight (boxing) then pull out a gun and start shooting. Gaming the system is the point, and I don't want that happening; not to me or anyone else. This protects "all" not just a few. So why would anyone be against this? Why would people (anyone) not want true competition, good competition, a good fight? It does not punish the majority and reward the few. It's an idea to keep it fair, even if there no unfairness going on. The gaming the system does exist, even if it only happens 1% of the time, that's 1% too many in my opinion.

Just to add ~ if one wants to get better, they need to play those of their level. The struggle in chess is what makes it better than other sports/games. The winning is just part of the process, a small part of the journey. This is why when finished playing a game, we start another one and play again.

Ignoring the sandbagging problem and addressing the floor idea...

This was mentioned on another thread, except the suggestion was the floor would be based on your average rating over a certain number of games, rather than your highest rating. So if you stayed above 1600 for however many games, your floor could be 1400, as an example.

 A certain number of games? Like the last 25 games (if applicable - or all your games if you have not reached 100). The sandbagged games would fall under that and would be counted. It wouldn't have as much of an effect...again, it was the idea that I am pushing for, not the exact way I mentioned; so if that is what the staff determines is the best way to implement a "floor" then so be it. It's still the idea...

Instead of a floor, here's a better idea:

Extend the tournament criteria to include some additional variables. This would address the request at hand while not affecting every tournament. Enable this logic:

"I want to allow players in the rating range 1400-1600, who have completed at least 30 games on chess.com, and whose rating in the past 3 months has not been higher than 1800".

Now - good luck getting enough people that satisfy such criteria, but that's not really the subject of this, now is it? :-)

Hmm, there is the problem that you can have wild variations in your own rating. For example, it is nowhere near impossible to go on a great run and go up 300 points (to 2000) say, even though your usual rating is 1700. Then with 200 points you'd be forced to go 1800 and up even though your usual rating oscillates between 1600 and 1800.

There is another way, perhaps.

It's possible to work out your expected rating from your results. For example if you score 70% against 1600, then the formula says your rating is 1800 (say). So you could have a floor a certain number below this rating performance over the last (however many) games.

Certainly simply having a high rating coming out of a tournament isn't enough evidence to say that sandbagging has occurred -- one would need to inspect the activity prior to a tournament as well and even then there might be a number of legitimate explanations for something that appears to be sandbagging.

It's important to note that not only is there is a selection bias towards better players advancing in tournaments, the very process of advancing puts upwards pressure on their rating.

I don't know whether sandbagging is actually a problem or not, I simply don't have enough tournament experience to know, but a couple of reasons I like the approach is because it's simple and also because it's preventive as opposed to punitive.  I think this is precisely what should be prescribed for something for which the root cause or motivation is so subjective and indeterminate in each case -- ultimately this is the same rationale for why we have a time per move statistic available to us.

 I don't have any problem with this idea either as I have not had any for the other suggestions. Again, it's not the way to do it that's my concern, it's the ratings-floor that I am concerned about. How it's implemented will be up to the big guys that are in charge; however they decide it is best done.

~ Just my thinking, if someone goes on a great run, against opponents of the same caliber, then most likely they have made improvements in their game that now allows them to keep up with better competition. On the other hand, if a person picks and choose opponents well below their level and goes on a run, then I think they are just going for wins because that makes them feel better although that will in the long run hurt them, especially heading into a tournament.

 I don't really see why 200 is too low; ok, so what would be a good/decent number in your opinion? BTW; this floor idea will not "bar" anyone from any tournament, if this is your true concern. It will however make tournaments more exciting as the competition will be well matched; what's wrong with facing good competition? Tournaments will be undecided until the finals and thus offer exciting games, instead of 2500 joining a 1800 tournament and rolling over the competition; where's the excitement in that?

Why would someone ~2700 want to play a tournament for 1500-2000 anyway? What would that prove? What kind of competition would a 1500 or 1600 offer a 2700? (I know you didn't say that the floor should be 700 or any other number, but what would be good?).

Here's an idea to add to the rating-floor idea (for tournament purpose); take the 10 highest rated player that has been defeated by said person and apply that stats to the rating-floor. How the math will be applied I don't really know, but if there's any mathematicians out there willing to maybe work out the details/formula then that would be awesome.

e.g.

So someone, let's say is 2000, and the average for the top 10-players they have defeated is ~1750, then that statistic would be considered into the ratings-floor. Meaning this person might be choosing to play lower rated players to inflate their ratings (and maybe ego) so their rating-floor might be lower than says someone like (see below)...

On the other hand, let's say another person (1625) has the average top 10-players defeated of ~1900, then that would be totally different from the above example. Although this stat would still be used in the calculation of the ratings-floor just the same. This does not necessarily mean that this individual had sandbag down to 1625, but you can see this person can defeat 1900s (and even higher as this is an average) on a regular basis, thus a tournament for under-1650 would be something that this person would dominate. No excitement here...

I'm still eagerly waiting to hear from a staff, as I am curious as to what they think of the idea of ratings-floor for tournaments.

 Check out my first game finished. I think you might be wrong here, Tricklev.

I do like the idea. It has my vote.

I have not done any tournaments yet but I think that would be good for me to do.