Rating Disparity Ratio

  • 98 posts
  • Page 6 of 7
dough_boy wrote:
The real reason that veterans don't play live, or newbies is that I can lose 35 rating, and gain only a fraction of that. All it takes is bad luck on a drop, bad luck on dice, or a bad player that throws the game.

I have now played 3 live games.

1. 1st one, had two separate 10v3's, lost 9, killed 1
2. 2nd one, I had even dice, other player couldn't miss, including a 5 on 8 and they went 8-1
3. 3rd one, pretty sure the two other players were the same, or knew each other, because they only attacked me, or made weird moves like not blocking their bonus during the other players turns.

So this rule is just stupid as limiting to 50, I still will NEVER play another live game, let alone against 3 or less people.
B4rny wrote:
Hi,

If even the (very) high ranked players say this new rule is unnecessary, please get rid of it.
Or work it down to Captains, Lieutenants, ... to not lose more than 30, 40 points per game lost. Now it's just the 'lucky' few, benefitting form the new rule.

So my opinion: get rid of the new rule, or get some system that works for every rank, with a maximal amount of points to lose.
sfclimbers wrote:
A limit sounds fair to me. I'm not big enough to suffer from it. But, I can see clearly that I would be against such large losses if/when I ever got there.
The_Bishop wrote:
Well, I didn't want to add anything anymore to this discussion because basically I have already said all I had to say. However after one year perhaps we got some more data to analyse for those who care.

-- The dominating 12 list for 16 May 2020 --rating --games played
1  ...  Dominator @slackbatter63162513
2  ...  General @AlexCheckMate5870623
3  ...  Lieutenant General @hooboy11532616696
4  ...  Lieutenant Colonel @Knightmare749581211
5  ...  Brigadier General @vikingo13374929517
6  ...  Major General @alphax21149002058
..
-- The dominating 12 list for 04 Jun 2021 --rating --games played
1  ...  General @AlexCheckMate7026753
2  ...  General @Knightmare766072237
3  ...  General @slackbatter60862764
4  ...  General @vikingo13375701624
5  ...  Brigadier General @TP_knighty46324591
6  ...  Brigadier General @Deepdaleduck44914809
..
Vikingo1337 in the last year has grown less than 800 points in spite the new rule, AlexCheckMate has grown by about 1200 points and the current Dominator, Joey-Knightmare-7, around 1600. The three of them play against noobs and so are being helped by the new rule, but actually only the latter plays many games, the other 2 only play a few team trap-games per month and that's all.

So let's divide by the total number of games played in the last year, this will clear things:

Viki.  +772 / 107 =  +7.21 points per game (really a great ratio, congrats!)
Alex.  +1156 / 130 =  +8.89 points per game (wow he wins always!! hahaha)
JK7  +1649 / 1026 =  +1.61 points per game (nice, finally something realistic)

I feel like none of them really needs or deserves any special rating help, for just playing against noobs. That's exactly what they did already before the rule change and likely what they would still keep doing either way. Plus -- one more thing, quite expected actually -- as you can see last year the difference between 1st and 6th was of 1416 rating points, while now is of 2535, the disparity rule is causing more disparity!

Please read this also if you have time:
https://dominating12.com/forums/2/general-discussion/3373/no-cap-on-points-lost
«God doesn't play dice with the World» ~ Albert Einstein
bluebonnet wrote:
Hmm,
We had the same arguments going on where I came from. How to get top ranks to play lower ranks. I was always of the opinion they play the game to suit their style and if they care a lot about their points, they become more careful about the games they choose. Not all, but definitely some.

 Implemented something similar as well I think. Just different numbers, same concept. I don't think I ever saw data on it though.

But 1 thing is definetley different here that I think needs added to bishops calcs I think. You get points back for suicides from what I was reading. Only time you got points back at my old digs was if there was a game glitch. Which unfortunately happened way too often. They may have done it for a suicide, but I don't remember ever seeing it and there was no rule on it. But I found suicides to be extremely rare in the games I played.

So the question is, how often do suicides occur? I assume they happen more with lower ranks... So does it affect the top d12 more for the those that do play lower ranks? Or was it like my old place where you may have heard about 1 suicide in a thousand games? (pulling number from my posterior)
bluebonnet is online.
AlexCheckMate wrote:
The_Bishop
Well, I didn't want to add anything anymore to this discussion because basically I have already said all I had to say. However after one year perhaps we got some more data to analyse for those who care.

-- The dominating 12 list for 16 May 2020 --rating --games played
1  ...  Dominator @slackbatter63162513
2  ...  General @AlexCheckMate5870623
3  ...  Lieutenant General @hooboy11532616696
4  ...  Lieutenant Colonel @Knightmare749581211
5  ...  Brigadier General @vikingo13374929517
6  ...  Major General @alphax21149002058
..
-- The dominating 12 list for 04 Jun 2021 --rating --games played
1  ...  General @AlexCheckMate7026753
2  ...  General @Knightmare766072237
3  ...  General @slackbatter60862764
4  ...  General @vikingo13375701624
5  ...  Brigadier General @TP_knighty46324591
6  ...  Brigadier General @Deepdaleduck44914809
..
Vikingo1337 in the last year has grown less than 800 points in spite the new rule, AlexCheckMate has grown by about 1200 points and the current Dominator, Joey-Knightmare-7, around 1600. The three of them play against noobs and so are being helped by the new rule, but actually only the latter plays many games, the other 2 only play a few team trap-games per month and that's all.

So let's divide by the total number of games played in the last year, this will clear things:

Viki.  +772 / 107 =  +7.21 points per game (really a great ratio, congrats!)
Alex.  +1156 / 130 =  +8.89 points per game (wow he wins always!! hahaha)
JK7  +1649 / 1026 =  +1.61 points per game (nice, finally something realistic)

I feel like none of them really needs or deserves any special rating help, for just playing against noobs. That's exactly what they did already before the rule change and likely what they would still keep doing either way. Plus -- one more thing, quite expected actually -- as you can see last year the difference between 1st and 6th was of 1416 rating points, while now is of 2535, the disparity rule is causing more disparity!

Please read this also if you have time:
https://dominating12.com/forums/2/general-discussion/3373/no-cap-on-points-lost

Hey there The_Bishop,

I'll react in two parts; first part in a spoiler, as I'll mainly just be replying to you, second just regular, as it'll be more on topic.

addressing The_Bishop (click to show)

Seems logical that someone (you) brought this up again around now; it's indeed been a year and data can be analysed, if desired. Not quite sure what the data is supposed to unveil though... I guess there's 2 possible outcomes?

1) the rule has been applied to more games than expected
2) the rule has been applied to less games than expected

if 1 => guess the rule is effective? there's more mingling of rating in games (as wanted?)
if 2 => guess the rule isn't all that effective? (it doesn't "hurt" as much as you think it does? no harm in keeping it? might even want to intensify it to see whether that'll bring the intended effect (surely not what you wish to achieve, The_Bishop?))

As for a mathematical response. If your main issue is with rating of people rising; the largest cause for the rise isn't this rule - it's the introduction of more new players (all joining with 1000 rating - an idea could be to deduct 1000 rating of everyone at a specified time and let new players join with 0 rating instead; but I can understand that it's much more visually pleasing to start on 1000 and that this may keep more people interested in staying on this website).



Anyhow, I can also see how my posts in this thread could be seen as rather biased due to my rating - so be it.

-Alex
“Gravitation cannot be held responsible for people falling in love. How on earth can you explain in terms of chemistry and physics so important a biological phenomenon as first love? Put your hand on a stove for a minute and it seems like an hour. Sit with that special girl for an hour and it seems like a minute. That's relativity.”

― Albert Einstein
The_Bishop wrote:
@Bluebonnet, yes I feel what you mean. Many parameters are there actually and that's why it's hard to write a conclusion. In fact I just wanted to show the data we have access to, in order to get a better idea of how things progressed in the last year. I think to know from which site you come and I know the formula in use there... Well, it's different from here, so it can't be really compared. I think our system is better actually and doesn't really need any caps.

@Alex, I'm just showing some data to consider. Of course your name is there: you have the highest rating, you are so the most favourized by this rule and you showed agreement and appreciation for it. I don't care how often the rule applies because that dosen't determine by itself if the rule is good or bad. And there is no expected percentage of application, so I feel like you are really thinking at a wrong datum. Taking off 1000 points by everyone I assume it's a joke, it would cause some bugs. Yes, I have a new rating proposal -- concerning 2p games only -- but I can't really find the time to post it.
«God doesn't play dice with the World» ~ Albert Einstein
slackbatter wrote:
Some of my personal data:

From 25 May 2020 - 24 May 2021 (first date is the first finished game of mine with the new rule):

I had 241 finished games with a Win or Loss (draws excluded).

I lost exactly 50 points in 63 of those games.

It would take a lot of effort to figure out how many points I should of lost in each game, but I think it pretty conservative to say that I have at least 300 points I wouldn't have had otherwise and possibly as many as 500. I was above 6000 rating for the majority of that year so it would not be too usual to lose over 60 points in a game without the rule.

I played fewer games than most previous years for me but otherwise didn't change my game choice practices in any conscious way. All games I create do not have any ratings limits, but I join plenty of games that do.
vikingo1337 wrote:
I suppose I should be flattered that anyone would spend this much of his free time going through my finished games on a Risk site. Personally I prefer golf, but each to his own.

Regardless, despite Bishop's impressive number-crunching, and it really is impressive, the most important question is still left unanswered:

Have more games been played between high-ranked players and low-ranked players after the introduction of the cap? We still don't know, but that was the reason for introducing the cap in the first place. Not to decide who benefited from it, or whether or not that was to Bishop's liking.

The hypothesis behind the cap was that high-ranked players did not play low-ranked players very often, as they stood to lose a significant number of rating if the game was lost, whereas they stood to gain very few rating if they won. Even with the cap, that is still the case today – as pointed out by Alex.

If you, Bishop, want to establish who benefits most from the cap, then that is a different discussion entirely in my opinion. But if you do, and it seems likely, then you should look at the high-ranked players who have LOST most games since the cap was introduced – not just against noobs, but against other players as well. Those players are the ones who have benefited most from the change – not the ones who kept on winning.

But first you must find out whether more or fewer games have been played between high-ranked players and low-ranked players since the cap was introduced. THAT is what we've been testing out for the past year now. And if the conclusion is that more of those games have been played, well, then certainly that is a good thing, right?
"The brave man well shall fight and win, though dull his blade may be."
~Fafnismal 28
slackbatter wrote:
So tell us vikingo, have you played, or at least been more willing to play, more games against low ranked players?

The cap has made zero difference for me. I doubt there's any way to evaluate this other than for those benefiting from the cap (there are only a few of us) to say if it has changed their behavior when joining games.
God_of_War wrote:
Cant' the rating system just be reset annually so everyone starts over? 

There is already a "historic" saving of the highest rank achieved? Is that not good enough or do we need to make sure we are kept "above" the rift-raft of players? I don't remember one person's rating over another, but I remember the Name of the player to determine if they have low, medium or high skill in this game.

Thoughts on an annual reset? Like a Ladder that is climbed? A new Season? Maybe it will bring back a lot of players to attempt to make the climb?
Hi there!
vikingo1337 wrote:
You wrote this in your previous post, slackbatter:

"I think it pretty conservative to say that I have at least 300 points I wouldn't have had otherwise and possibly as many as 500."

So the cap indeed has made a difference for you. Whether it has changed your behaviour in terms of which games you join is another matter. Only by pinpointing the players who have lost the most games with the cap, and comparing their behaviour in terms of the players they faced and the games they joined, is it possible to determine the effects of the cap.

I think I have played more games with newcomers than before, yes, as an experiment. My experience with the cap, as I mentioned earlier, is that it doesn't work as well as it is supposed to. If you keep winning all – or the vast majority of – your games, you will not feel much different with the cap in place, even if you join more games than before. But if you lose those extra games you join, then you most definitely will benefit from the cap.

That being said, I still believe that it is a more fair rating system with the cap than without it for the simple reason that it may have stopped certain players, myself included, from refusing to play newbies at all. This suggestion was not created by me, mind you, but it certainly resonated with me after having lost 70 rating in a few games, where I stood to gain maybe 10 or 20. So in those rare cases I'd say that the cap is effective in that it reduces the injustice of the win-loss disparity. Just not as effective as it could have been.

Another thing that occurred to me a while back is that there also seems to be an upper rating cap in place in terms of how many rating you can win in a single game: 150. For my part, I haven't seen anyone win more than 150 rating at least. So it makes sense that there is a lower cap, too. But perhaps it should be lowered even further for it to have the desired effect.

As for your suggestion, God of War, I'm fairly sure it would lead to players leaving this site and not coming back. I also think it's off-topic.
"The brave man well shall fight and win, though dull his blade may be."
~Fafnismal 28
vikingo1337 wrote:
And yes, bluebonnet. Suicides happen from time to time on this site. Certainly more often than every 1,000th game.

Naturally it affects those players who have more to lose than those who don't. Not just in a rating sense, but also in terms of the overall game experience. Experienced players who put in more time than the average player on strategising and communicating can have all their work undone in the blink of an eye by a suicide, and they usually don't appreciate that. More often than not, the balance of power is also destabilised by a suicide, leading to a random winner and an abrupt ending to the game.

In contrast, a suicide does not appear to be a big thing to at least some new players, seeing as the game tends to be more low-involvement for them. It's my impression that players who do it often either don't know, or care, that suiciding is against the rules.
"The brave man well shall fight and win, though dull his blade may be."
~Fafnismal 28
God_of_War wrote:
vikingo1337
You wrote this in your previous post, slackbatter:
As for your suggestion, God of War, I'm fairly sure it would lead to players leaving this site and not coming back. I also think it's off-topic.

I'm not sure what word I would have for those who "quit" due to a restart every so often. If they are so good at this game, than surely they would all be able to prove it by climbing the rating system over and over again. I don't think there is an argument there. People either enjoy playing risk or they truly enjoy knowing that others are beneath them and always will be.

But you are correct, probably is off-topic, although it would solve the issue in evening the field for everyone where you couldn't complain about caps, or playing lessor rated players and losing your shirt on rating if you lose.

Carry on.
Hi there!
kwikool wrote:
one of the other things i've noticed is that many of the high level players play team games. they set the game up with 2 high level players and then open it up to anyone. those joining can be newbies and have no idea how to play a team game. or if a seasoned player joins he is generally hooked up with a unseasoned player . this give the person setting up the game a huge advantage.  i dont blame them for taking advantage of the "system" but it is a way yo play newbies at an unfair advantage. and its hard to get your partner together to join the team game at the same time....if there was a better way to select teams or reserve the other team member for those who join after the initial setup, it would make this more competitive.