Brainstorm for a metric for the "interestingness" of a hit

Do you want to test the possibilities of this forum? Want to test out your new signature? You can do it here.

Moderators: Crazy Bob, Fons, avij, Phaseolus

User avatar
an-148
Euro-Master
Euro-Master
Posts: 4583
Joined: Wed Oct 11, 2006 5:04 pm
Location: summer: Ivanica, BiH ; winter: La Calamine (Liège) BE
Contact:

Re: Brainstorm for a metric for the "interestingness" of a h

Postby an-148 » Wed Feb 20, 2013 6:25 pm

I agree 100 % :D :D

User avatar
dserrano5
Euro-Master
Euro-Master
Posts: 3773
Joined: Sun Jan 31, 2010 11:59 am
Location: Madrid, ES

Re: Brainstorm for a metric for the "interestingness" of a h

Postby dserrano5 » Wed Feb 20, 2013 9:27 pm

tralla wrote:How will you explain to a new user that she'll not need to bother to register in Helmbrechts because all her hits will be moderated?


I'd say that uninteresting hits don't necessarily have to end up being moderated. They could be assigned a score from say 0 to 100, and the web interface could substitute the current interesting/moderated filter by a "hide hits having less than N points".

Just saying.

User avatar
tralla
Euro-Regular
Euro-Regular
Posts: 210
Joined: Wed Apr 27, 2005 5:23 pm

Re: Brainstorm for a metric for the "interestingness" of a h

Postby tralla » Thu Feb 21, 2013 2:46 pm

But who's to decide what's uninteresting? To me, my Finnish hits from the early days of EBTing are much more precious than the Egyptian hit where I may have carried that note back to Germany myself and therefore done most of the travelling for that note. And Finland is much more exotic for me than Egypt! I've been to Egypt several times but never to Finland. 8-)

Remember your first hit? The "Wow, this can really happen"? What if someone had told you "Yeah, nice, but there are people with *hundreds* of hits in this town, so your's doesn't really count"? I should use a different tactics if I wanted to keep people interested and motivated to stay with EBT.

And may I remind all that EBT's motto is "Follow your Euro notes in their tracks" (see banner ;-)) and not "Follow some of your Euro notes.. " or "Follow your Euro notes if we let you".

I should make an exception for obviously faked note entries if it is at all possible to identify these - *that* I'd gladly leave to the experts and cleaning women. Otherwise, keep in mind that there are thousands of active users out there abiding by the rules, entering their 3-6 notes every 3 days and happy to eventually score this hit. With that base, I firmly believe that EBT can cope with a small number of morons who need to up their egos by faking hits or locations or whatever...

End of sermon *g*. Tralla
Meine EBTST-Statistik (danke, dserrano5)
Meine nigmm-Statistik (danke, tt.01)
My little gallery of stamps and scribbles on Euro notes.

User avatar
CarlosManta
Euro-Master
Euro-Master
Posts: 3059
Joined: Mon Jan 07, 2008 6:11 pm
Location: Porto, Portugal
Contact:

Re: Brainstorm for a metric for the "interestingness" of a h

Postby CarlosManta » Thu Feb 21, 2013 3:24 pm

tralla wrote:But who's to decide what's uninteresting?

I guess the name of this topic is self explanatory, and answers this question.
We, as a community, recognize that not every Hit has the interestingness, and as a community are trying to brainstorm and see if we can come up with some metric that reflects what most of the community thinks.
For sure it will be imperfect, it will have its flaws, but the last proposals seem to be quite transparent and to include most parameters. Far better than what we have today.
This definition should be made public, and potentially will many hit moderation issues. If you look a the number of post in the moderation topic, it is quite obvious that this is a major issue.
In other words, after collecting the proposals, it will be the laws of mathematics and equations to decide.
:D

User avatar
an-148
Euro-Master
Euro-Master
Posts: 4583
Joined: Wed Oct 11, 2006 5:04 pm
Location: summer: Ivanica, BiH ; winter: La Calamine (Liège) BE
Contact:

Re: Brainstorm for a metric for the "interestingness" of a h

Postby an-148 » Thu Feb 21, 2013 4:22 pm

imperfect for imperfect, it's better to forget about "scaling" hits, because:

- like tralla, I consider that the vast majority of the users are very happy with all hits (maybe 2 or 3/years). This vast majority never goes on the forum and rarely click on some other link than "new note", "my bills","my hits"... and maybe their dotmap; they hardly understand why some hits appear and other don't. For random users, there wouldn't be any improvement to their "success" if we apply complicated rules to scale their hits.
- who wants this scaling?: an infime minority of power users who are passionate and for whom 95% of the time spent on internet concerns ebt. With the time spend on ebt, those user have gained a more precise overview of ebt than random users and developped their own idea about the interrestingness of a hit, depending mostly of their own situation: the problem is: 100 user (specialists!!), 100 different minds (also specialized!!): one consider the most important is the distance, another one the time, another one the border crossing, the hit-ratio of the country, the town, the village or the street where the hit occured (therefore you have sooooooooo much different formula's in this topic). NO FORMULA will ever be able to satisfy the power user themselve , not to speak about the feeling of random users !!
- why do some user want to scale the value of hits? I believe here is THE POINT : jealousy !!!!! jealousy about hits in Helmbrechts, Vienna, Klagenfurt or Kouvola !!

we are about to disturb (maybe definetely!) the user "lambda" to satisfy basic instincts of a handful of eternal unsatisfied users: that's not an improvement at all !

:cry: :cry:

User avatar
CarlosManta
Euro-Master
Euro-Master
Posts: 3059
Joined: Mon Jan 07, 2008 6:11 pm
Location: Porto, Portugal
Contact:

Re: Brainstorm for a metric for the "interestingness" of a h

Postby CarlosManta » Thu Feb 21, 2013 4:34 pm

an-148 wrote:- why do some user want to scale the value of hits? I believe here is THE POINT : jealousy !!!!! jealousy about hits in Helmbrechts, Vienna, Klagenfurt or Kouvola !!

I recommend you read this post on page 1 of this discussion. You will be clarified.
This game of accusations is something I am not interesting in participating. From this post onwards I will only write concerning the "metric", i.e., the topic being addressed here.

User avatar
an-148
Euro-Master
Euro-Master
Posts: 4583
Joined: Wed Oct 11, 2006 5:04 pm
Location: summer: Ivanica, BiH ; winter: La Calamine (Liège) BE
Contact:

Re: Brainstorm for a metric for the "interestingness" of a h

Postby an-148 » Thu Feb 21, 2013 5:02 pm

exactly because the situation described in your link, you'll see why it's IMPOSSIBLE to find an exact formula !!! (I really don't understand why you want to prove something using a fact that proves exactly the opposite :D :D )

accusations ??? I believe you dindn't read my comments correctly, or do I write so bad english? :cry:

accusations? where? I just try to warn against complexity which will banish random user, warn AND explain, demonstrate why this direction is dangerous for ebt.

Now, that is only my five cents: it's legitimate to share it, or?

Most probably, my opinion (and other's as well) will not be taken in account and the formula will be introduced: I have a "bad feeling" about it and simply want to share and try to explain why: end of story.

b.t.w.: I ask myself why every discussion here has to turn to vinegar !!! are ebters nervous people? it look like that :roll: :roll:

User avatar
dserrano5
Euro-Master
Euro-Master
Posts: 3773
Joined: Sun Jan 31, 2010 11:59 am
Location: Madrid, ES

Re: Brainstorm for a metric for the "interestingness" of a h

Postby dserrano5 » Thu Feb 21, 2013 5:18 pm

an-148 wrote:it's IMPOSSIBLE to find an exact formula !!!


Nobody said this formula/algorithm would be perfect. Of course it will be different to what we have now, and of course that means it will be better for some people and worse for some others, but I think an open, transparent method of defining "interestingess" of hits is an improvement over the opaque system we have today.

I only hope that the only hits removed from the listings will be the ones deemed totally uninteresting, ie those that are being moderated today. As long as a hit is minimally interesting, it should be shown, IMHO.

User avatar
CarlosManta
Euro-Master
Euro-Master
Posts: 3059
Joined: Mon Jan 07, 2008 6:11 pm
Location: Porto, Portugal
Contact:

Re: Brainstorm for a metric for the "interestingness" of a h

Postby CarlosManta » Thu Feb 21, 2013 9:26 pm

dserrano5 wrote:I only hope that the only hits removed from the listings

Why would any Hit be removed from the listings at all?
One of the adavantages of having a score is exactly to avoid the need to remove or moderate Hits.
Each Hit would have a score associated, and all Hits would be kept.
Notice that the Hit score is complementary to the Hit list or the Hit count.
It is like the coins in your pocket. Maybe we both have 5 coins, but I may have 5 5-cents coins, and hence 25 cents, and you have 5 20-cent coins, and hence 1 Euro. But we still have 5 coins each.
So you see, there is even no need for the ultra-conservatists to show all this alarm. Everyone will keep all the Hits, even the ones moderated today.
What would (or hopefully) happen is that additionally a score is automatically given to each Hit.
Thus you wouel get in addition to your number of Hits, your Hit Score.
This is common in many fields. Research papers have an impact score, webpages have a PageRank (complimentary to pageviews), and so on.
Moreover, if you look at you list, you could sort by date, days traveled, kms travelled, but also by score!

The way I see it it is really about creating additional information, not about removing or taking away any information that already exists.
If you ask me, it is really hard to understand the arguments of this ultra-conservatists. All that exists today is kept, except the moderation that they constantly moan about.

So it is not a surprise at all that this idea of creating a metric came from users in Germany or Vienna. It is obvious that users with many Hits are the most interested in finding the "interesting" Hits in their lists.

User avatar
dserrano5
Euro-Master
Euro-Master
Posts: 3773
Joined: Sun Jan 31, 2010 11:59 am
Location: Madrid, ES

Re: Brainstorm for a metric for the "interestingness" of a h

Postby dserrano5 » Thu Feb 21, 2013 11:53 pm

CarlosManta wrote:
dserrano5 wrote:I only hope that the only hits removed from the listings

Why would any Hit be removed from the listings at all?


Not knowing what decisions are being/will be taken, one can only guess that there's a greater-than-zero probability that some hits could be hidden (e.g. those with an interestingness <10%). I was only stating, from a totally agnostic point of view, that I hope that that isn't the case.

Let me add, too, that I'm not alarmed and I'm not an ultra-conservatist. I'm for the introduction of this formula.

User avatar
tralla
Euro-Regular
Euro-Regular
Posts: 210
Joined: Wed Apr 27, 2005 5:23 pm

Re: Brainstorm for a metric for the "interestingness" of a h

Postby tralla » Fri Feb 22, 2013 8:59 pm

Wow, this is the first time I've been called an ultra-conservatist. Should I feel offended now or am I just growing old? :lol:

I agree completely that today's moderation system is opaque but I hesitate to believe that the formulas presented in this thread will be less opaque to the majority of our friends at EBT who are not rocket scientists. Hence my elegant and intuitive suggestion that
1 hit = 1 hit

While I find CarlosMantas analogies somewhat off the mark, I cannot help adding my own - what's the English word for a system where a small minority of powerful persons enforce their rules on a large number of people?

Maybe, if "all" of you super-users are so keen on "your" universal metric, you can provide a means for each user to turn that feature on or off in her personal profile (default would be off, of course). That way, every user would have a choice/vote - now that would be in the spirit of a democratic society, wouldn't it?

Vive liberté, égalité, fraternité
tralla

... ultra-conservatist. This is so cool. *very big grin*
Meine EBTST-Statistik (danke, dserrano5)
Meine nigmm-Statistik (danke, tt.01)
My little gallery of stamps and scribbles on Euro notes.

User avatar
nossi
Euro-Master
Euro-Master
Posts: 1388
Joined: Thu Jul 15, 2004 9:41 pm
Location: 1220 Wien

Re: Brainstorm for a metric for the "interestingness" of a h

Postby nossi » Fri Feb 22, 2013 9:29 pm

I think, there is a basic problem which can't be solved. for maybe 90% of all EBTlers statistc and specially ranking issues are not interesting. They insert some notes a week and are happy if they have a hit. They also dont make much difference between their hits. One is as good as the other.

Power-User and some other people like to compare to each other. For them EBT is a game with competition and therefor they have an other sights of hits. The most important thing for them is to have a fair and equal chance in this competition. But this is not so easy to build, because the basics are very different. For this users a formula, which make a balance between citys whre you can get easily a hit (like Helmbrechts, Helsinki or Vienna) and others, where it is hard to find one (like Bologna) is very important.

Is it possible to find a way, which satisfies both sides? Maybe if we have two statistics? One for each hit counts 1, and the new form, where hits have different importance.

Here we could have a new filter: Important (not "interesting"!) hits!
Liebe Grüße aus Wien - nossi -

User avatar
CarlosManta
Euro-Master
Euro-Master
Posts: 3059
Joined: Mon Jan 07, 2008 6:11 pm
Location: Porto, Portugal
Contact:

Re: Brainstorm for a metric for the "interestingness" of a h

Postby CarlosManta » Mon Feb 25, 2013 2:01 am

tralla wrote:Maybe, if "all" of you super-users are so keen on "your" universal metric, you can provide a means for each user to turn that feature on or off in her personal profile (default would be off, of course). That way, every user would have a choice/vote - now that would be in the spirit of a democratic society, wouldn't it?

I have a hard time understanding where this idea of "imposition" comes from. It is just an additional piece of information. Does not even interfere or change what is already there.
However, I have no objection what so ever in having the score as an optional feature, or even with the feature turned off as default.
(Optional in the sense of not being shown, it would still be calculated in the background, for statistics, rankings, etc)

User avatar
avij
Forum Moderator
Forum Moderator
Posts: 5615
Joined: Mon May 27, 2002 10:45 pm
Location: Helsinki Finland
Contact:

Re: Brainstorm for a metric for the "interestingness" of a h

Postby avij » Mon Feb 25, 2013 8:16 pm

Moved from the other topic:

lmviterbo wrote:Also, it would be superneat if it's possible to include historical meaningfulness to hits without changing interestingness rating over time. Here's an example:

avij wrote:How many hits the users have in common? If it's their first hit, it's more interesting than if it was their 500th.

Let's suppose I have my first hit with avij. This would be rated 25, for instance. When we have our second hit, it would be rated, say, 15. Now the first hit shouldn't change its rating because of a second hit. In fact, I guess this is exactly what you are thinking but I wanted to clarify it.

As to bundle hits, I can't really say that in a bundle of 30 hits the first that was entered is more meaningful than the last one, just a split second later. Bundle hits might need a special treatment so that they all keep, for example, an average between the rating the first one would receive and the rating for the last one.

Yes, one design goal is that once a score has been calculated, it shouldn't change. However, it must also be possible to recalculate the score if, for example, the user has entered an erroneous location for the note and fixes it later on. The system should know all the score calculation parameters that were in effect when the note was entered. Another reason for the recalculation is if/when we change the score calculation algorithm and/or weights.

Bundle hits are a problem. Some bundles whose notes all have the same entry places & times are easy, but the "grouping" becomes much more difficult when there are differences in the hits among the bundle notes. If you have brilliant ideas for solving this grouping problem, I'd be interested in hearing about those.
Money makes the world go round. We track how the money goes round the world.
EBT Tech WG leader. Do not PM me if your question is not related to Tech WG or the association.

User avatar
lmviterbo
Euro-Master
Euro-Master
Posts: 4158
Joined: Thu Aug 21, 2003 5:23 pm
Location: Lisboa, Portugal
Contact:

Re: Brainstorm for a metric for the "interestingness" of a h

Postby lmviterbo » Mon Feb 25, 2013 10:19 pm

avij wrote:Bundle hits are a problem. Some bundles whose notes all have the same entry places & times are easy, but the "grouping" becomes much more difficult when there are differences in the hits among the bundle notes. If you have brilliant ideas for solving this grouping problem, I'd be interested in hearing about those.
Sorry, avij, either my English is rusty or I have sleep deprivation (which is true today, in fact). I'm not getting it. I was talking about notes that are entered by someone in a row and produce passive hits in a row (to the same person in the same place and date). I don't see how notes that are entered at a single time and found at different times or places might be more or less interesting than any random set of notes entered twice at different times. (But I also don't know if that's what you meant.)