Posts from Player rating for tournaments across all installments of Heroes. Auto-translated

Possibly. Is there a rating for individual versions?

The results are interesting, but debatable. I personally didn't expect to see myself in such a position. In fact, I should be lower, much lower. Apparently, even this rating isn't a perfect solution.

But why?
Since this is a cumulative rating, the more games you play, the higher your rating... it's all correct... you're the last of those who played 6 games... I don't see any problem... because the overall rating, combined with the average rating, shows who's who... :D
Sincerely.

And yet, the results are debatable. Let's try to discuss them. In my opinion, the leaders are somewhat disadvantaged in this rating.

We have already discussed that the average rating also gives incorrect results. Now, let's say someone plays one game, gets 100 points, and stops playing. They will always be in first place :) compared to those who play every time but get 1st, 2nd, or 3rd place. Is that fair?

That's why ratings need to be updated to reduce the ratings of players who no longer play or play infrequently.

What if the map allows you to get close to the leader? Why should the winner suffer if 10 people fall behind him by 1 day? He chose the best Walkthrough. And those who are behind him will get almost the same.

As I can see from the rating, the number of games greatly affects the result, much more than with the oscillator and my bonus.

I wish I could hear Amber's opinion, but he has disappeared somewhere.

You took a phrase out of my message... but you should have read the whole sentence... I wrote: "The overall rating combined with the average rating shows who is who..."
That is, the overall rating in itself doesn't show anything... the same goes for the average... but if they are both considered together, it immediately becomes clear who is who... ;)

Still, I'm against aging...
It's possible to deal with inactive players in another way...

What if the map allows you to get close to the leader?

Why should the winner suffer if 10 people are one day behind him? He chose the best playthrough. And those behind him will get almost the same.

As I see from the rating, the number of games greatly affects the result, much more than with the oscillator and my bonus.
I'd like to hear Amber's opinion, but he's gone somewhere.

It's very simple.. I suggested (way back) and am suggesting now to maintain a cyclical rating.. like with the clans...
i.e., once a year (for example: January 1), the current Player Rating of all players will be reset... and an entry will appear in the user profile:
"Rating for the year 20XX = K, number of games played N, average rating for the year = M"
Respectfully.

So what??? Those who played once and got 100, what will they get in a year according to your system? 100 – I guessed correctly :) And will they outperform all the active participants who didn't take first place in any of the Tournaments? In the next year, there will again be people with 100 and only one tournament played. And the active players will again be left out in the cold :)

Moreover, according to your scheme, the active players will have worse records in their profiles than the players who successfully played only ONE Tournament this year.
What do you say to that?

Why will they bypass the activists? After all, their total rating will be much lower than the activists'...

I don't understand why the activists' profile entry would be worse? :confused:
Total rating is much higher! Number of games is too! Well, the average rating is lower, of course...
but let's say we compare two entries:
1) Rating for year 200x = 100, number of games played 1, average rating for the year = 100
2) Rating for year 200x = 729, number of games played 9, average rating for the year = 81
which one would you want to see in your profile?

I would dream of seeing the second entry in my profile... :D
But, of course, all people are different, and their tastes vary...

Now for a bit of philosophy:
Under this system, past achievements aren't "forgotten"... i.e., even if someone won one tournament, that's already HISTORY! And that must be remembered.

In the case of aging, I don't argue, one-day wonders won't stay on the pedestal for long... but they will "hang around" the player list for quite a while. Because the lower their rating becomes, the slower it will drop... And until when? Probably until the rating falls below 1... right?
Calculate how long a one-day leader will hang around the list with 10% aging? I got 46 times... tell me, do you really need this?

From tournament to tournament, this player will slowly die... causing you trouble and introducing confusion into the rating..

By "confusion in the rating," I mean it will be completely unclear who is an active player (but not very successful) and who is a one-day wonder, since both will have completely identical numbers...

Summary: Under one system, everyone will be assigned their real rating and it will remain in history; under the other, we will only be able to see the leader, and regarding the rest, it will be unclear who is an active player (but not very successful) and who is a one-day wonder.

Well then, it turns out you agree that the Overall Rating is higher in status than the Average Rating???

Well, I would have liked something else :) I would like an overall rating of 1000 games at level 10, and an average of 100 :D
1) Rating for the year 200x = 100, number of games played = 1, average rating for the year = 100
2) Rating for the year 200x = 300, number of games played = 10, average rating for the year = 30
3) Rating for the year 200x = 500, number of games played = 5, average rating for the year = 100
Now compare 1) and 3) - The average rating says nothing. Nothing at all. Player 1 played 1 tournament, player 3 played 5 - the average rating is the same. And player 2 will say that he is better than player 1 - as if he had tried playing 10 tournaments, I would be watching him :)

Well, then I want to see my alternative rating in my profile too :)

They insisted on 10% :). During the Kanigema tournament, I even promised not to age the rating until the tournament was over. By the way, how many rounds will there be? I thought there would be 5. In general, after the Kanigema tournament, I will reduce everyone's rating by half :) Immediately by 50%. Then those who played few tournaments will drop in the ranking.

But they will also hang around with you. Then someone will return a year later and say, but why am I starting from scratch? My rating remained. And this will cause you a lot of trouble too.

Well, I think I can easily add the number of games next to each one, all my data is stored in tables - then I can copy these data to you. Only I won't calculate the average rating. And let everyone analytically think and compare players' results as they do with the average rating.

And for what reason, if I may ask, are you aging the rating by 50% right away? Just because your left foot felt like it?

This approach is unacceptable. There should be clear rules that must be followed... and less improvisation...
By the way, how are the tournaments that take place during the ВК Tournament aged? Specifically, the "Second Skeleton Gathering Tournament"?
Are they not aged? That's not right. They should be aged according to your method... only the ВК Tournament should not be aged!

But once a year, the rating is reset anyway... or rather, it is transferred to the Hall of Fame for year n. So, I don't see any problems here...

Excellent! All that's left is to cancel the aging and add an average... :D
i.e., we are on the right track...
But we initially wanted to manage with just one number... but it seems that's not working out

It's just that each rating has different tasks, and it's difficult to compare them.