Personally I find reviews of an online chess "couch" amusing.
If the "couch" does not have a high otb rating that all can see (cm fm online) does not count.
he/she should not be saying they are good. couches
@justme23 said in #31:
I know many strong chess players without a formal title who would make excellent coaches but a title is in a way insurance of a minimum expected level of knowledge. I agree about the second part that " he/she should not be saying they are good coaches" i believe that's the point of having a review system in place. No system could guarantee a 100 % accurate reviews and I don't think it's needed, frankly if 80% is correct the benefits will outweigh the shortcomings. By slowly improving the review system step by step the loud voices of those ( few ?)unhappy with how it works now would be reduced over time.
Say what? Any player could be a couch of someone that plays bad.
A good player does not need good reviews he/she has tournament wins.
@justme23 said in #33:
> Say what? Any player could be a couch of someone that plays bad.
> A good player does not need good reviews he/she has tournament wins.
Just because Magnus Carlsen is arguably the best chess player in the world does not mean he is the best coach. Coaching takes patience, understanding, and knowledge of tournaments and paths. There are a hundred things they could teach, but they teach things that are the most beneficial to the player.
@andro2 said in #19:
> Namely, every review someone writes for a coach first goes to the coach's assessment. The coach can then publish or NOT publish that review. Only then do these reviews become visible on the coach's profile.
This makes the entire review system rather pointless, dont u think?
With that knowledge in mind, taking down the review system was a good decision i would say.
Out of curiosity, how do big companies manage review systems, let's say Amazon has misleading reviews for certain products. Tons of misleading reviews on Chessable. Obviously, Amazon has much more money and technology then lichess. I think that every new review should be reviewed by a moderator before it is posted. I'm sure someone has mentioned this idea somewhere. Deleting all the reviews just harms the coach and doesn't really solve a problem. That being said, when you train with a coach, it's not guaranteed the coach is for you either way.
#16:
> The problem is that we don't have a good way to confirm that one person was coached by another.
> Think of the solution, not the problem.
>
> — Zeddicus Zu’l Zorander
No way to tell if reviews are legit? Really? I’d be surprised if that’s the case. I should think there are lots of ways to do it, without too much trouble. Perhaps I’m mistaken, but just bear with me. (And notice none of these methods are mutually exclusive; why not implement several, just to be safe?)
∗ ∗ ∗
1. Verified payments
Require that a payment be made through a form on this site from the student to the coach before that user can post a review, and/or ask for a valid transaction ID after the fact.
Obviously you would redirect to PayPal; but by going through a form here, or requesting that info afterward, you could have some sort of record that a transfer took place — and so verify that the student actually hired the coach. No personal info need be recorded, not even the amount; simply proof that the PayPal service was in fact used. I’m sure they already provide various kinds of confirmation to online retailers, so just let them do the work and receive their info here, et voilà: third-party verified student.
2. Associate the accounts
You know how you used to be able to see who was following whom? Well, that info still exists, it’s just no longer viewable on people’s profiles. So make a special kind of mutual “follow” for coaches and students. It needn’t show up on their profiles either, but it can tell the site: “These two users are in a teaching relationship.” Then only allow users who have such a relationship with the coach to post reviews.
By itself, this wouldn’t eliminate scammers, but it *would* eliminate spammers — e.g., no more “If you join my team, I’ll write you a nice review.” ;-) Plus, if some guy makes a second account just to review himself, it should be easier to figure out that’s what he’s done: the association between those accounts will make any overlapping data points — e.g. device, OS, browser, location, active hours — easy to spot. Even a VPN wouldn’t help much at that point. If one account is only used at the same time as the other, for example, that will show up. Anything suspicious and “No review for you!”
3. Check for shared studies
It should be easy to see if both accounts have ever edited a study together before.
4. Log study hours
If the two accounts were only in a study together for one minute, disregard that. If, OTOH, they’ve repeatedly studied together for an hour at a time, let that count towards the review-posting potential of the student account.
5. Offer secure studies
Make a special kind of study, where both the coach and student must “sign in” to it. Then simply check to see if the would-be reviewer has ever done this with the coach.
6. Delay reviews
Make it so that even after the payment is verified, accounts are in a relationship, studies have been shared, logged and signed, you still need to wait a month before you can post a review. :-P
This should nip most would-be scammers’ plans in the bud, since most of them won’t have that kind of patience. I mean, at this point you’d have to be pretty dedicated — wiring money to yourself, keeping more than one device in the same study for an hour at a time, etc. — and now you still need to wait before you can review yourself? Surely that weeds out most scammers. It would just be too much trouble for them to do all that, especially with more than one fake student. (And those who do anyway should be easy to catch: the associated accounts will be too obviously made for the purpose.) Meanwhile, it gives the real student plenty of time to edit that draft and be completely honest. ;-) As well as come up with examples of where some new principles were put to use. :-)
∗ ∗ ∗
And that’s just off the top of my head. ̄\_(ツ)_/ ̄ So what do you guys think? Might any of these measures help? (If not, why not? I’m interested to know what I’m missing here.) Or can anyone think of some other approaches? ’Cause I’m sure there must be more than the half dozen I just outlined. (Really, verified reviews aren’t a novel concept.)
When it comes down to it, all I’m saying is I doubt it’s true that “we don’t have a good way to confirm.” I think there are a plethora of ways you might do so. And since the more hoops one has to jump through, the harder it is to fake, I think you should go ahead and use all of them that you can. ̄\_(ツ)_/ ̄
@pawnedge said in #37: Thanks Pawnedge for many valid points, Let's say for every review the coach decide to publish he/she would have to upload at least one receipt of payment between the parties ( screenshot). Third party validation, i hope it's not PayPal since many coaches uses national mobile payment systems, in Sweden where i have most of my students we use "Swish" since it doesn't impose any fees in contrast to PayPal. I would say looking for shared studies combined with tracking time spent together in those studies would be enough. Even simpler to check headlines of studies, let's say my students name is Magnus Carlsen and we had a class for today i would name it Magnus Carlsen 15th June + today's theme. Some machine learning should be able to determine if there is enough correlation between Coach / student otherwise a bot could flag the review and admins would manually check it. Let's name the bot " Van Helsing review hunter " i like the idea of focusing on the solution and not just seeing problems. I also think the concept of collective punishment is wrong in this case thousands of coaches have been punished financially. Big changes har rarely any good, small steps leads to good results and just removing the reviews might have been a bit draconic. Regards Richard
@Pawnstructures said in #23:
> Dr hack, the solution implemented here was removing all reviews collectively slapping all coaches in the face. It's like seeing a fly for the first time and launching a inter continental ballistic missile at it with a nuclear warhead instead of asking a fly veteran who would probably suggest a more subtle approach such as a flyswatter. No warning issued, no dialogue. Just push the delete button and the problem is gone.
Lol that would be deleting the site:)
I understand your issue, i understand your message has been passed there is no point in continualy defending your opinion now.Move ahead.
@pawnedge said in #37:
> #16:
> 1. Verified payments
> Require that a payment be made through a form on this site from the student to the coach before that user can post a review, and/or ask for a valid transaction ID after the fact.
>
How would lichess administration verify payments? Lichess would not be party in transaction and hence has no means verifying the payment.
