Scale points won by number of opponents in game

[Metsfanmax]

Not "rounds" but "moves" are more important here.

10 rounds 1vs1-game = 19-20 moves, but 10 rounds 8-p game = 73-80 moves. Let's say speed games need in average 2 minute per move, so 1vs1-game takes in average 40 minute but 8p-game needs 160 minutes.

On one aspect of this I completely agree: speed games are quite a bit longer with more players. I don't think this is a concern for 24-hour games if you're a premium member though. But I don't think that's a reason to award more points. It's just something you knowingly sign up for when you join a speed game with a large number of players.

I don't agree that the number of total moves made by all players has a unique bearing on how many points you should earn, because in either case the individual who won took the same number of turns (which presumably is proportional to the effort expended on the game).

Hm... from this perception I can agree with you in some degree : 8p-games aren't worth 7x 1vs1-games, but 8p-game can't have the same value with 1vs1-game. Just find a better formula

Frankly any formula we come up with will be arbitrary, which is why I'm reluctant to support anything other than the OP. However, if I could coalesce support around the alternate formula I've presented:

Points Won = (loser score / winner score) * (20 / sqrt(# of opponents))

Then I would likely push for that to be implemented if no one could find a better formula. In fact I'll change the OP right now so that we can have a discussion on that. Will everyone agree that an eight-player game is worth 2.6 times as many points as a 1v1? No. But it's better.

[Night Strike]

Do you think that a person who wins a 12 player game deserves 11 times as many points as a person who wins a 2 player game?

That only happens if everyone in every game are at the exact same score when the game finishes. That rarely happens, so there are times when you get more than 20 in a 1v1 game and other times that you get less than 220 in a 12 player game. And yes, if you beat 11 more people in a game, then you should have the potential to earn 11 times as many points from that game.

[Metsfanmax]

Do you think that a person who wins a 12 player game deserves 11 times as many points as a person who wins a 2 player game?

That only happens if everyone in every game are at the exact same score when the game finishes.

Yes, but it's a fiducial measure. On average, the 12 player game will give you about 11 times as many points as the two player game.

That rarely happens, so there are times when you get more than 20 in a 1v1 game and other times that you get less than 220 in a 12 player game. And yes, if you beat 11 more people in a game, then you should have the potential to earn 11 times as many points from that game.

OK, but why? The problem I have with this perspective is that it delinks score and skill. Score is a meaningful quantity if it represents your skill at playing the game. For this to be true, every point has to be approximately equally difficult to obtain. If there are ways to earn certain points substantially easier than others, then your score no longer accurately reflects your skill, or at least the effort contributed. If the number of points you obtain is not proportional to either of those, then what does it mean? It becomes little more than just a number.

[Night Strike]

That rarely happens, so there are times when you get more than 20 in a 1v1 game and other times that you get less than 220 in a 12 player game. And yes, if you beat 11 more people in a game, then you should have the potential to earn 11 times as many points from that game.

OK, but why? The problem I have with this perspective is that it delinks score and skill. Score is a meaningful quantity if it represents your skill at playing the game. For this to be true, every point has to be approximately equally difficult to obtain. If there are ways to earn certain points substantially easier than others, then your score no longer accurately reflects your skill, or at least the effort contributed. If the number of points you obtain is not proportional to either of those, then what does it mean? It becomes little more than just a number.

By that logic, we have to remove all options from the website that provide any variability in game types and settings. Since some people will gain their points playing ass-doodles while others play only 1v1 tournaments, their skill levels can never be adequately compared. If points are supposed to be the measure of skill, then everyone should be playing essentially the same settings. Simply changing the calculations to punish some settings doesn't make the skill measure any better. Besides, if a player isn't that good but gets a big jump in points by winning a 12 player game, they'll lose those points soon enough.

[greenoaks]

Do you think that a person who wins a 12 player game deserves 11 times as many points as a person who wins a 2 player game?

That only happens if everyone in every game are at the exact same score when the game finishes.

Yes, but it's a fiducial measure. On average, the 12 player game will give you about 11 times as many points as the two player game.

as it should.

defeated players are just as defeated whether they are the sole loser or had 10 others for company. the points they lose should be the same regardless of how many others also lost.

[Fazeem]

current system definitely needs a overhall this sounds like a great suggestion that could be coupled with a few others to level the playing field and make skill as much a factor as luck and overall mastery of a map or settings.

[Metsfanmax]

That rarely happens, so there are times when you get more than 20 in a 1v1 game and other times that you get less than 220 in a 12 player game. And yes, if you beat 11 more people in a game, then you should have the potential to earn 11 times as many points from that game.

OK, but why? The problem I have with this perspective is that it delinks score and skill. Score is a meaningful quantity if it represents your skill at playing the game. For this to be true, every point has to be approximately equally difficult to obtain. If there are ways to earn certain points substantially easier than others, then your score no longer accurately reflects your skill, or at least the effort contributed. If the number of points you obtain is not proportional to either of those, then what does it mean? It becomes little more than just a number.

By that logic, we have to remove all options from the website that provide any variability in game types and settings. Since some people will gain their points playing ass-doodles while others play only 1v1 tournaments, their skill levels can never be adequately compared. If points are supposed to be the measure of skill, then everyone should be playing essentially the same settings. Simply changing the calculations to punish some settings doesn't make the skill measure any better. Besides, if a player isn't that good but gets a big jump in points by winning a 12 player game, they'll lose those points soon enough.

Yes, we can never fully remove that aspect of variability from the main scoreboard. That's no excuse for having a bad aspect of variability. There's a blatant aspect of unfairness here that we should dampen if we can. Again, if you agree that a person shouldn't win 11 times as many points for winning a 12 player game as for winning a 2 player game, then we're on the same page, and all that is left is to debate the formula.

[Night Strike]

Yes, we can never fully remove that aspect of variability from the main scoreboard. That's no excuse for having a bad aspect of variability. There's a blatant aspect of unfairness here that we should dampen if we can. Again, if you agree that a person shouldn't win 11 times as many points for winning a 12 player game as for winning a 2 player game, then we're on the same page, and all that is left is to debate the formula.

I don't agree with you on your premise.

[Metsfanmax]

Yes, we can never fully remove that aspect of variability from the main scoreboard. That's no excuse for having a bad aspect of variability. There's a blatant aspect of unfairness here that we should dampen if we can. Again, if you agree that a person shouldn't win 11 times as many points for winning a 12 player game as for winning a 2 player game, then we're on the same page, and all that is left is to debate the formula.

I don't agree with you on your premise.

So do you explicitly agree that a 12 player game is worth 11 times as many points as a 2 player game, on average?

[koontz1973]

So if a player earns less points for this, players would lose less points as well? That might make a few play more games but I doubt it as players play what they like.

Playing a (X) sized game may not be (Y) times more differcult, but you forgot to add the (U) quantity. All games including 1v1 have an element of luck but with the larger game I would surmise that the luck factor is spread between 8 different players and is less of a factor in games.

I could go further and say that because the luck factor in 1v1 games is far higher, the points won should be less. When you consider that in a 1v1 game, anything from a bad drop, early bonus, bonus drop, a round or two of bad dice, who goes first could all swing a game towards a player that is not as good as the opponent, means the game can be won or lost very early on and no matter how skilled you are, if the luck is not with you, you lose. Hence, players should not lose points for having bad luck.

Either way you look at it, the scoring system is not bad when you consider all things.

[betiko]

i would definitely never play multi singles again. this would remove all the fun.
I still didn't get an answer, who gives a f*ck about people having fluctuating scores? i can tell you that you premise that people playing lots of games get back t a sort of eqilibrium is completely false. In my experience, when I play tons of games is when i feel like playing tons of speeders, I never have too many active games. Well I can tell you that my score fluctuates over 1000 during the week i play tons of speeders. Winning much above average a given multi single setting is about skill. the score is supposed to reflect the skill. you have very little chances to win a 12 player game and when you do well jackpot, that's jst the way it is. It's a great feeling to win 200+ points and it doesn't happen very often to anyone on this site. Even if the conqueror was a 12 player game player, why would we question him beig the conqueror if he only plays those sort of games? it would be a remarkable way to become conqueror.

also you are not talking about battle royales, with 52 players; we re talking about different proportions of point bonanza there!

[Metsfanmax]

So if a player earns less points for this, players would lose less points as well? That might make a few play more games but I doubt it as players play what they like.

I don't expect that this will really change the games people play. It's actually because of this that I want to fix it.

Playing a (X) sized game may not be (Y) times more differcult, but you forgot to add the (U) quantity. All games including 1v1 have an element of luck but with the larger game I would surmise that the luck factor is spread between 8 different players and is less of a factor in games.

I don't think it's a large variation. For example if you play an 8 player game that's not escalating, at least half the players will get screwed by the drop most likely.

I could go further and say that because the luck factor in 1v1 games is far higher, the points won should be less. When you consider that in a 1v1 game, anything from a bad drop, early bonus, bonus drop, a round or two of bad dice, who goes first could all swing a game towards a player that is not as good as the opponent, means the game can be won or lost very early on and no matter how skilled you are, if the luck is not with you, you lose. Hence, players should not lose points for having bad luck.

Under the alternate system I've proposed (with the square root), you do get less points for a 1v1. You just don't get 11 times less.

[BigBallinStalin]

Let's only consider the issue of determining one's skill by using the score.

With the current system, the score is an insufficient proxy for skill since an aggregate value is not enough. With recruitment, I always go through maprank to view win rates in standard, dubs, trips, quads, and 1v1s. Points are irrelevant because complements matter more (e.g. being on a good team) and having a high 1v1 or standard win rate doesn't mean one is a good team player (i.e. they may have potential, or they may be too much of a lone wolf).

Let's assume this suggestion is implemented. Will it accurately enough reflect a player's skill?

No. I'd still have to go through maprank to disaggregate one's overall score.


From this POV, I don't see why CC should bother with this.

[Metsfanmax]

Let's only consider the issue of determining one's skill by using the score.

With the current system, the score is an insufficient proxy for skill since an aggregate value is not enough. With recruitment, I always go through maprank to view win rates in standard, dubs, trips, quads, and 1v1s. Points are irrelevant because complements matter more (e.g. being on a good team) and having a high 1v1 or standard win rate doesn't mean one is a good team player (i.e. they may have potential, or they may be too much of a lone wolf).

Let's assume this suggestion is implemented. Will it accurately enough reflect a player's skill?

No. I'd still have to go through maprank to disaggregate one's overall score.


From this POV, I don't see why CC should bother with this.

That's not the only perspective by one which should analyze the meaning of a score, because there are benefits to having a high score no matter how you get it (like being Conqueror). If you don't value score as an independent quantity that's fine, but lots of people do, so it should be calculated fairly across settings.

However, I grant you that there's a big difference between skill when playing alone and skill when playing in teams. I don't think these two should be on the same scoreboard.

[patrickaa317]

I still don't understand why it's so bad for someone to go up or down 200 points. Why is 200 such an issue but 150 isn't? Or 100?

If you are trying to curb that, pretty soon the next issue will be going up or down 50 points in a game.

I went from 2500+ to 1800 in about 10 days because i started playing a lot more games that I typically don't. I knew that was a risk but it is what it is. I don't want some safety net coded into the site for my score not to drop so drastically.

Also, I may have missed it but how do you see this working with terminator games?

[Metsfanmax]

I still don't understand why it's so bad for someone to go up or down 200 points. Why is 200 such an issue but 150 isn't? Or 100?

It's not "so bad," as if there's some cutoff between what is acceptable and what is not. The greater the degree of fluctuation is, the less a player's current score can be trusted to reveal their skill.

Also, I may have missed it but how do you see this working with terminator games?

It would work the same in Standard and Terminator, as proposed.

[OliverFA]

I am afraid this time I have to disagree. Winning 12 player games is difficult enough to make them pay even less.

The real problem is that players only get points if they finish 1st, and the only alternative is an unfair terminator mode or a random assassin mode.

[OliverFA]

I still don't understand why it's so bad for someone to go up or down 200 points. Why is 200 such an issue but 150 isn't? Or 100?

If you are trying to curb that, pretty soon the next issue will be going up or down 50 points in a game.

I went from 2500+ to 1800 in about 10 days because i started playing a lot more games that I typically don't. I knew that was a risk but it is what it is. I don't want some safety net coded into the site for my score not to drop so drastically.

Also, I may have missed it but how do you see this working with terminator games?

Extreme variation is a problem for the score system. In average, the system is supposed to represent the actual player level, but because a single game can create such a huge variation, it introduces errors in the representation.

However, I think this is not the solution.

[patrickaa317]

I still don't understand why it's so bad for someone to go up or down 200 points. Why is 200 such an issue but 150 isn't? Or 100?

If you are trying to curb that, pretty soon the next issue will be going up or down 50 points in a game.

I went from 2500+ to 1800 in about 10 days because i started playing a lot more games that I typically don't. I knew that was a risk but it is what it is. I don't want some safety net coded into the site for my score not to drop so drastically.

Also, I may have missed it but how do you see this working with terminator games?

Extreme variation is a problem for the score system. In average, the system is supposed to represent the actual player level, but because a single game can create such a huge variation, it introduces errors in the representation.

However, I think this is not the solution.

Seems like rather than a dynamic score where your 10,000th game is calculated the same way as your first (only dependent on your score and the opponent(s) score), it should take into consideration how many games have been finished by both you and the other person.

[Shannon Apple]

I disagree with this completely based on the fact that you may already have lost a hell of a lot of points in playing multiplayer (8-12 player) games before winning any. Sure, you can have a win streak and shoot up by a few hundred points, which may be followed by a losing streak where you lose a hell of a lot more. I don't think it's at all fair to disadvantage the winner in these situations.

[OliverFA]

I disagree with this completely based on the fact that you may already have lost a hell of a lot of points in playing multiplayer (8-12 player) games before winning any. Sure, you can have a win streak and shoot up by a few hundred points, which may be followed by a losing streak where you lose a hell of a lot more. I don't think it's at all fair to disadvantage the winner in these situations.

Of course. If the winner gets less points the loser should also get less points, as points won minus points lost should still be 0.

[Metsfanmax]

I disagree with this completely based on the fact that you may already have lost a hell of a lot of points in playing multiplayer (8-12 player) games before winning any. Sure, you can have a win streak and shoot up by a few hundred points, which may be followed by a losing streak where you lose a hell of a lot more. I don't think it's at all fair to disadvantage the winner in these situations.

Of course. If the winner gets less points the loser should also get less points, as points won minus points lost should still be 0.

Indeed, that would still happen in this suggestion, of course.