Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Sign In Register

Quick Links


Categories

In this Discussion

› › Identifying Top Contributors to LET - Page 2
New on LowEndTalk? Please Register and read our Community Rules.

All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.

Identifying Top Contributors to LET

2

Comments

  • raindog308raindog308 Administrator, Veteran

    @Lee said: Yeah, go use that shit on someone else for likes, doesn't work on me.

    ...what, facts?

    image

  • raindog308raindog308 Administrator, Veteran

    @default said: I want an ugly vote. I am ugly, so I will most likely win.

    Not with that adorable avatar!

  • jsgjsg Member, Resident Benchmarker

    @raindog308

    Either ...
    ... "the community" decides by voting, in which case this whole discussion isn't needed. Simply count "Thanks", the "votes" of the users (but then memes and what not will count as "valuable").

    -or-

    At the end of the day @jbiloh, you, and the team will have to define what they want LET to be because that also defines what is valuable here on LET and what is not.

    No definition of 'valuable' ~ no "who is valuable" and, way worse, no positive guide lines for us users.

  • bulbasaurbulbasaur Member
    edited October 2021

    @raindog308 said: Yes, your formulae are very interesting and I will look at those deeper.

    Thanks for the kind words, if you do happen to implement it based on what I wrote I'd suggest two changes:

    • Consider calculating scores only if the comment/thread has been thanked at least once (otherwise, someone may resort to making many useless comments and "win" by brute force).
    • As it stands, the current formula for comment scores overestimates the usefulness of meme comments, I'd suggest calculating it as comment_thanks / 5 + log_10(comment_char_count) which does a better job in my testing.

    Updated scores to help you get some idea about how it works:

    Perhaps you should even calculate the final score as 10 * (thread_score + comment_score) to bring one digit after the decimal point into consideration, as you can see above a meme comment got ranked almost the same as a thoughtful comment by @seriesn. (I know providers aren't ranked, but those comments are some of the most useful I could find.)

    Thanked by 2chocolateshirt adly
  • FranciscoFrancisco Top Host, Host Rep, Veteran
    edited October 2021

    @stevewatson301 said: A thoughtful comment by @Francisco about cociu's situation:

    33 thanks / 5 + log_10(2556 chars) = 10.007560849486362

    🤴

    Francisco

  • NekkiNekki Veteran

    Excellence in forum posting cannot be calculated by statistics and formulas.

    Thanked by 4TimboJones bulbasaur BlaZe Falzo
  • jsgjsg Member, Resident Benchmarker

    @stevewatson301

    While I value your approach of trying to find a sensible formula I'd like to note two points

    • your formula punishes length and favours short comments. 'Thanks / x + log10(comment_length)' translates to a longer comment getting a lower score. I find this important because it strongly favours short (~ rule of thumb: low value for the community) posts
    • (not specific to your approach) any kind of considering length of posts is "bent" by (quite common) posts containing a long quote plus a short actual comment. So the question (@raindog308) arises whether his numbers can see length of quote vs length of actual comment.
    Thanked by 2AlwaysSkint bulbasaur
  • NekkiNekki Veteran

    @jsg said:

    While I value your approach of trying to find a sensible formula I'd like to note two points

    • your formula punishes length and favours short comments.

    Longer posts are boring - factual.

    Thanked by 8skorupion jackb adly Andrews TimboJones lentro bulbasaur BlaZe
  • SirFoxySirFoxy Member

    peenus

    Thanked by 2Nekki dahartigan
  • raindog308raindog308 Administrator, Veteran

    @Nekki said: Excellence in forum posting cannot be calculated by statistics and formulas.

    Are you sure?

    +--------------+-----------------+
| Name         | all_time_thanks |
+--------------+-----------------+
| jar          |           23266 |
| Nekki        |           20629 |
| raindog308   |           17771 |
| AnthonySmith |           14418 |
| Francisco    |           11869 |
+--------------+-----------------+
  • NekkiNekki Veteran

    @raindog308 said:

    @Nekki said: Excellence in forum posting cannot be calculated by statistics and formulas.

    Are you sure?

    > +--------------+-----------------+
> | Name         | all_time_thanks |
> +--------------+-----------------+
> | jar          |           23266 |
> | Nekki        |           20629 |
> | raindog308   |           17771 |
> | AnthonySmith |           14418 |
> | Francisco    |           11869 |
> +--------------+-----------------+
>

    Now deduct all the thanks from @netomx.

  • jsgjsg Member, Resident Benchmarker

    @Nekki said:

    @jsg said:

    While I value your approach of trying to find a sensible formula I'd like to note two points

    • your formula punishes length and favours short comments.

    Longer posts are boring - factual.

    For some, yes, for others not. Plus there are things that can and things that can not be put in a 3-liner, especially not for someone whose native language isn't English (but maybe that's just me being untalented with other languages than mine).

    But anyway my remark was just a neutral observation. Whether short or long posts are valued higher is up to @jbiloh and the team.

  • FranciscoFrancisco Top Host, Host Rep, Veteran

    @raindog308 said:

    @Nekki said: Excellence in forum posting cannot be calculated by statistics and formulas.

    Are you sure?

    > +--------------+-----------------+
> | Name         | all_time_thanks |
> +--------------+-----------------+
> | jar          |           23266 |
> | Nekki        |           20629 |
> | raindog308   |           17771 |
> | AnthonySmith |           14418 |
> | Francisco    |           11869 |
> +--------------+-----------------+
>

    All (former) admins except me. Neat.

    Francisco

  • NekkiNekki Veteran

    @jsg said:

    @Nekki said:

    @jsg said:

    While I value your approach of trying to find a sensible formula I'd like to note two points

    • your formula punishes length and favours short comments.

    Longer posts are boring - factual.

    For some, yes, for others not.

    Have a fucking day off mate.

    Thanked by 1TimboJones
  • NekkiNekki Veteran

    @Francisco said:

    @raindog308 said:

    @Nekki said: Excellence in forum posting cannot be calculated by statistics and formulas.

    Are you sure?

    > > +--------------+-----------------+
> > | Name         | all_time_thanks |
> > +--------------+-----------------+
> > | jar          |           23266 |
> > | Nekki        |           20629 |
> > | raindog308   |           17771 |
> > | AnthonySmith |           14418 |
> > | Francisco    |           11869 |
> > +--------------+-----------------+
> >

    All (former) admins except me. Neat.

    Francisco

    So by default you win, except because people arselick providers like they arselick admins, you’re also disqualified, but less disqualified than the formers admins, so technically you still win.

  • FranciscoFrancisco Top Host, Host Rep, Veteran

    @Nekki said: So by default you win, except because people arselick providers like they arselick admins, you’re also disqualified, but less disqualified than the formers admins, so technically you still win.

    DEFAULT! DEFAULT! DEFAULT!

    Francisco

  • rcxbrcxb Member

    LET beware! This is how the purge starts...

    Thanked by 2AlwaysSkint bulbasaur
  • raindog308raindog308 Administrator, Veteran

    @Nekki said:

    @raindog308 said:

    @Nekki said: Excellence in forum posting cannot be calculated by statistics and formulas.

    Are you sure?

    > > +--------------+-----------------+
> > | Name         | all_time_thanks |
> > +--------------+-----------------+
> > | jar          |           23266 |
> > | Nekki        |           20629 |
> > | raindog308   |           17771 |
> > | AnthonySmith |           14418 |
> > | Francisco    |           11869 |
> > +--------------+-----------------+
> >

    Now deduct all the thanks from @netomx.

    Same query, with all netmox-given thanks filtered. You're still quality:

    +--------------+-----------------+
| Name         | all_time_thanks |
+--------------+-----------------+
| jar          |           21795 |
| Nekki        |           18993 |
| raindog308   |           17027 |
| AnthonySmith |           13916 |
| Francisco    |           11527 |
+--------------+-----------------+

    Of netomx's thanks, @Nekki was the biggest beneficiary (1636), followed by jar (1471) and me (744). @netomx is also @Nekki's biggest thanks-giver (3x more than the next person, @Pwner at 501).

  • NekkiNekki Veteran

    Ok, now deduct all the thanks given when people were moderators/admins.

    Thanked by 1default
  • raindog308raindog308 Administrator, Veteran

    @Nekki said: Ok, now deduct all the thanks given when people were moderators/admins.

    Although looking at the schema suggests that role changes are logged, they are in fact not logged, so only a user's current role is known. It's not possible to go back and say user X had role Y from YYYY-MM-DD to YYYY-MM-DD. We can only say user X has it or doesn't have it now.

  • NekkiNekki Veteran
    edited October 2021

    @raindog308 said:

    @Nekki said: Ok, now deduct all the thanks given when people were moderators/admins.

    Although looking at the schema suggests that role changes are logged, they are in fact not logged, so only a user's current role is known. It's not possible to go back and say user X had role Y from YYYY-MM-DD to YYYY-MM-DD. We can only say user X has it or doesn't have it now.

    Therefore, you have to disqualify everyone except @Francisco as staff votes should be void.

    Whoever, I would also suggest that thanks given to a provider are compromised, because bumlicky-licky.

    Therefore you need to find the member with the highest thanks count that has never been a staff member or a provider. That person should be crowned the greatest poster in the history of LET.

  • tinyweaseltinyweasel Member

    @raindog308 said: a badge system would exist

    LowEndWebhostingTalk.

  • defaultdefault Veteran

  • FranciscoFrancisco Top Host, Host Rep, Veteran

    @Nekki said: Therefore you need to find the member with the highest thanks count that has never been a staff member or a provider. That person should be crowned the greatest poster in the history of LET.

    So @deank i'm guessing?

    Francisco

  • raindog308raindog308 Administrator, Veteran

    @Nekki said: Therefore you need to find the member with the highest thanks count that has never been a staff member or a provider. That person should be crowned the greatest poster in the history of LET.

    As stated that is possible, but if you add this condition:

    @Nekki said: staff votes should be void

    ...then it is not.

    I can sort-descending who has the most thanks and go down the list until I find someone who wasn't a staff member or a provider.

    But I can't adjust the thanks count to filter out thanks given by staff, other than those who are current staff, because there's no record of when someone was or wasn't staff. I don't necessarily agree that staff thanks should be void, though - I don't think people mods run around mechanically thanking people for no reason.

  • FranciscoFrancisco Top Host, Host Rep, Veteran

    @raindog308 said: But I can't adjust the thanks count to filter out thanks given by staff, other than those who are current staff, because there's no record of when someone was or wasn't staff. I don't necessarily agree that staff thanks should be void, though - I don't think people mods run around mechanically thanking people for no reason.

    I don't see there being much bootlicking for mods/admins. Not sure why mods/admins should be disqualified.

    Jarland deserves every thanks he's ever earned.

    Francisco

    Thanked by 1adly
  • deankdeank Member, Troll

    @Francisco said:

    @Nekki said: Therefore you need to find the member with the highest thanks count that has never been a staff member or a provider. That person should be crowned the greatest poster in the history of LET.

    So @deank i'm guessing?

    Francisco

    My thanked count is low, only 10127.

  • raindog308raindog308 Administrator, Veteran
    edited October 2021

    @deank said: My thanked count is low, only 10127.

    ...which is in fact the highest thanks count other than mods and providers. Only @jar, @Nekki, me, @AnthonySmith, @Francisco, @Amitz, and @FAT32 are higher.

    I would say mods/admins' thanks are legit. They get some small boost because they announce rules, etc. but I think most of @FAT32's, for example, are from his extraordinary work during black friday, @Nekki's from his legendary personality, @Amitz from his sex appeal, etc.

    +--------------+--------------------+
| Name         | ReceivedThankCount |
+--------------+--------------------+
| jar          |              23261 |
| Nekki        |              20631 |
| raindog308   |              17779 |
| AnthonySmith |              14437 |
| Francisco    |              11875 |
| Amitz        |              10728 |
| FAT32        |              10392 |
| deank        |              10127 |

    In typical Vanilla fashion, there is a total kept in the users table and then a log table I can sum, and they number generated from each differs slightly...possibly deleted comments I'm guessing. But only by a few numbers.

    On the other hand, the most gracious person is @netomx, who has given 26,031 thanks, nearly double the next two people (@vimalware at 14,482 and @uptime at 14,320). The only person who is on both the top thanks-received list and the top thanks-given list is @jar (#5 at 9,874 given), who is a sort of gratitude switch-hitter.

    Well I guess you can all tell who here used to be a DBA.

    Thanked by 1netomx
  • defaultdefault Veteran

    @raindog308 said:

    @deank said: My thanked count is low, only 10127.

    ...which is in fact the highest thanks count other than mods and providers.

    Thanked by 1deank
  • raindog308raindog308 Administrator, Veteran

    I also have to share the news that LET faces a grim Y2K-type situation.

    The "ReceivedThankCount" column is implemented as a "smallint(5) unsigned". Why, I have no idea. Because Vanilla.

    In short (pun intended), if a user ever gets to 65,535 thanks, their thanks will max out and they will become unthankable.

    Fortunately, this just generates a warning in MySQL and does not wrap around to 0.

    I can't think of a reason why that column couldn't be updated to a larger integer with an ALTER TABLE. Such drastic action may be needed as we approach the Thankspocalypse. The T64K bug.

    Sure, you're saying, if people are barely at 20,000, they won't get to 65K in my lifetime...well, my friend, that's the same kind of thinking that is burning up the planet. Shame on you. Think of your grandchildren.

    Thanked by 1TimboJones
Sign In or Register to comment.