Connection ErrorBad RequestThis page went staleSession ExpiredSign in requiredForbiddenNot FoundToo Many RequestsServer ErrorBad GatewayService UnavailableGateway TimeoutHTTP Version Not Supported
Session Expired
Your session has expired. Please log in to continue.
Sign in required
You need an account to do that. Sign in or create one — it only takes a minute.
This page went stale
Your security token was rejected, usually because the page sat open too long or you signed in elsewhere. Nothing was saved. Reload the page and try again.
Request Failed
We encountered an issue while processing your request.
Please check your internet connection and try again.
Error Details:
Share Link
Posts from Player rating for tournaments across all installments of Heroes.
I can compile all the results into a single table without any problems. And I can also award points. BUT! If the esteemed VDV_forever awards points using one system, and I use another, then nothing good will come of it. We need to standardize everything.
I am more inclined towards the option without aging, so that the rating always increases. Because tracking missed games is such a hassle. It's easier to identify and exclude one-day players from the final annual table.
I can compile all the results into a single table, no problem. And I can also award Grails. BUT! If the esteemed VDV_forever awards Grails using one system, and I use another, then nothing good will come of it. We need to standardize everything. I am more inclined towards the option without aging, so that the rating always increases. Because tracking missed entries is such a hassle. It's easier to identify and exclude one-day players from the final annual table.
1. I started with this, and it really eliminates a lot of headaches... But it won't work for the current rating because of the infinite growth of the rating among the older participants... A newcomer will never be able to catch up. Therefore, the current rating will have to be aged in any case. A cyclical rating (e.g., annual) is a simple cumulative total (without aging). Then it will be reset.
2. Grails must be awarded using a single system for everyone. This is mandatory. What specific system is up to you. In addition to the initial basic system for awarding Grails proposed by VDV_forever, you have been offered two more. The rest is up to you... :) Or maybe we wasted our time poring over all these numbers with Iron? :)
Well, that's wonderful. I praised your system, and you praised yours... :) :) So, we're both great. And let's end it for now...
No, I praised yours too :) I pointed out the obvious advantages and disadvantages from _my_ point of view. Why should I lie?AmberSoler;60964
"Elasticity (or inelasticity) of the system" - the ability to respond to changing external circumstances... An elastic system reacts with a delay, has a lower frequency and amplitude of oscillations, and therefore is more stable per unit of time. An inelastic system reacts instantly, immediately reflects the final changes, has a large amplitude and frequency of oscillations, and as a result, has less stability.
Oh, well, if we look at it from a high-level, scientific perspective, then I agree, let it be inelastic. I just showed an example where my system was more elastic than yours - with Grandpa Mazai, although it may be a specific case, but still, it means that in some moments it acquires the properties of elasticity. However, from the point of view of overtaking a leader who has been accumulating a rating for a year in a few months, I agree with you, your system is more elastic.
I will still try to run the rating using two systems for three games.
Added after 11 minutes AmberSoler;60969
1. I started with this, it really eliminates a lot of headaches... But it won't work for the current rating because of the infinite growth of the rating among the older participants... A newcomer will never be able to catch up. Therefore, the current rating will have to be aged in any case. A cyclic rating (for example, annual) is a simple cumulative total (without aging). Then it will be reset.
That's right, I agree with AmberSoler. Without aging, there is no point in my system.AmberSoler;60969
2. Grails must be awarded according to a single system for everyone. This is a must. Which system - you decide. In addition to the initial basic system for awarding VDV_forever, you have been offered two more. Further - it's up to you... :) Or maybe we wasted our time digging through all these numbers with Iron? :)
And if we create a common system, then your formula is very suitable, naturally without aging in 10% and without any summation from tournament to tournament. Although, can grails be obtained not only in OFF-line tournaments? In this regard, it may not be necessary to act according to a single scheme. Because there may be online tournaments and various activities on the portal and rewards from the administration.
Let me clarify a small point regarding the aging of ratings. It's the current rating that ages, not the annual one. Is the annual rating simply the sum of current ratings? If so, then the annual rating also ages. Wouldn't it be simpler to do this: the rating doesn't age, but it's reset every year. And the player's profile (or the final table) would display the previous year's result.
Could you clarify a small point regarding the aging system? Does the current rating age, or is it the annual rating that ages? Is the annual rating simply the sum of the current ratings? If so, then the annual rating also ages. Wouldn't it be simpler to do this: the rating doesn't age, but it resets every year. And the player's profile (or the final table) would display the results from the previous year.
1. We calculate the tournament rating using a complex formula.
2. The sum of tournament scores for the year (without aging) is the annual rating.
3. The sum of tournament scores (with the aging system applied) without being tied to a specific time is the current rating... The aging procedure
Let me clarify a small point regarding aging. It's the current rating that ages, not the annual one. Is the annual rating simply the sum of current ratings? If so, then the annual rating also ages. Wouldn't it be simpler to make it so that the rating doesn't age, but is reset each year? And the player's profile (or the final table) would display the previous year's result.
I don't even know what the Annual Rating is. I can say this about my own:
1. There's a player who hasn't played anywhere yet - their rating is 0.
2. The player plays in a tournament. Now they have a rating that they earned from the tournament, Rtournament.
3. The player plays in another tournament. Now Rtournament becomes Rold. And for this tournament, they earn Rtournament again.
4. Now the player has a new rating, Rnew, which is calculated by adding 2/3*Rold and Rtournament.
Rnew = 2/3*Rold + Rtournament
2/3 - This is the aging factor. The aging factor could also be 1/10.
5. The player goes to a tournament to earn Rtournament again. After the tournament, Rnew is now called Rold, and now go to step 4 (cycle).
Added 4 minutes later AmberSoler;60982
1. We calculate the tournament rating using a complex formula.
2. The sum of tournament ratings for the year (without aging) is the annual rating.
3. The sum of tournament ratings (with aging applied) without being tied to time is the current rating... The aging procedure, applied constantly, negates old results, and therefore, in the end, the current rating only reflects the latest achievements in descending order...
I don't understand what you mean by "in descending order"? The Current Rating can both increase and decrease, but if the player is constantly winning and gaining points, there is a ceiling, so there's no point in resetting it. It can also decrease on its own. I'm saying this to AlexPank.
Okay, thank you. I think I understand now. So, the current rating is calculated continuously, after each tournament, and we apply an aging process to it. The annual rating is displayed at the end of the year, simply by summing the rating from each tournament. But then it turns out that the leaders of the annual rating will never be overtaken. Maybe we should reset the annual rating and not apply aging to the current rating?
AlexeyPank;60998That's it, thanks. I think I understand. So we calculate the current rating constantly, after each tournament, applying an aging process to it. The annual one is output at the end of the year, simply by summing the ratings from each of the tournaments. But then it turns out that no one will ever catch up to the leaders of the annual rating. Maybe we should reset the annual one and not age the current one after all?Everyone looked at the annual rating and threw it in the trash :) While the current one lives forever and is always aging. Next year, just add those values again for everyone after each tournament. You have to understand: each player's current rating can fluctuate from 0 to the ceiling—which in our case seems to be ideally 1000—even without a reset. And it can drop not only because of the aging process, but also if, for example, a player performs significantly worse in the current tournament compared to the leader. Ok? P.S. The annual rating doesn't have to be thrown in the trash; it can be put in the Hall of Fame!
It reflects the results in decreasing order... The most significant result is from the last tournament - 1.0. The result from the second-to-last tournament is less significant because only 0.9 remains, the third-to-last tournament is even less significant (0.81), then - 0.729; 0.6561... and so on, in decreasing order.
It's interesting who should be explaining this to whom...?
It reflects the results in descending order... The most significant result is from the latest tournament - 1.0. The result from the second-to-last tournament is less significant because only 0.9 remains, the third-to-last tournament is even less significant (0.81), then 0.729; 0.6561... and so on, in descending order. I wonder, who should be explaining this to whom?..
:) So, why remember or know how it's arranged? And can you determine, by looking at the current rating, how many of the most recent significant results are currently included? I don't think so. And if you can, then why bother?
:) So, why remember or know how things are arranged there? And can you determine, by looking at the current rating, how many of the most recent and significant results are stored there? I don't think so. And if you can, then why?
It contains ALL the results of ALL the tournaments in which the player participated. I was actually explaining the calculation mechanism to Alexey, not to you...
Sage! I'll take the third one, maybe even the first. What will you take? I won't be able to handle more. I need to manage two systems simultaneously. Which system will you choose? I can't decide yet, so I'll figure it out in practice.
Finally, I got to HW today. I guess I'll have to take everything else, because, besides us, there are no other people who want to calculate the rating yet. P.S. I'm still thinking about which system to use. Maybe I'll do the same as you: I'll figure it out in practice.
It contains ALL results of ALL tournaments that the player participated in. I wasn't actually answering you, but explaining the calculation mechanism to Alexey...
No, well, it's not like I know everything. See, aging constantly eats away a certain part, and as a result, it can become 0.05, or even zero. And then you build it up again. In short, if zero is reached — the player will have to be wiped from the table.
Added after 2 minutes Sage;61026
Finally got around to HW today :) . It looks like I'll have to take everything else, since besides us, there are currently no other people wanting to calculate the rating. P.S. As for which system — I'm still thinking. Maybe I'll do as you did: figure it out in practice.
And why negotiate with anyone? I've already agreed with VDV and same-named Prophet. But who is handling the other game versions?
AlexeyPank
