+- One Man Left Studios Community Forums (http://www.onemanleft.com/forums)
+-- Forum: General (/forumdisplay.php?fid=1)
+--- Forum: Outwitters (/forumdisplay.php?fid=11)
+--- Thread: Is total points really due to player activity alone? (/showthread.php?tid=857)
Pages: 1 2
Is total points really due to player activity alone? - joelduque - 11-08-2012 04:55 AM
My purpose of this thread is to rectify what I perceive as a grossly errorneous idea propagated by others. Pertinent discussion welcome.
I find it amusing that other members imply or even bluntly state that total points is just due to activity alone. So if you play along their line of reasoning, the more you play, the more your total points is. Well, that statement is true. I know where they are coming from. I am aware of the so-called win bias, wins adding more points than losses subtracting less points for the similar skill level. Based on my experience, this is currently at about +18 for a win and -12 for a loss at a similar skill level. Win bias is a related but essentially different topic.
But I beg to disagree that points is due to activity alone. It is also due to skill! Do you think people will keep on playing this wonderful and awesome game if their points is nothing but a sign of their activity? Well, I don't think so! People play this game for fun and also to feel good about themselves. They play to accumulate as much points as they can, to the best of their abilities. Competition. Yes, that is the very nature of league games. People continuously play to become better and because they believe they are much more skilled than the average player. That they are good, better, or even the best!
I don't know for others, but I find the claim that total points is due to player activity alone very much annoying if not downright insulting. I am speaking for myself alone. I do not agree with umbrella statements that belittle the achievements of all players, from the wild card unrankeds, humblest of fluffies to the mightiest of the super-titans.
In addition, such statements basically undermine the choice of the developers of an appropriate ranking system for the game. Yes, the current ranking and points system may not be perfect. Still, these claims essentially downgrade the player rankings to nothing but a ranking of player activity! I do not know about you but for me any ranking system without an element of skill is invalid especially in the context of league games, "Battle players of similar SKILL." It never said, "Battle players of similar ACTIVITY."
For me, these statements are demoralizing even if in my heart I know them to be untrue. So I am only just one player. But what more about other players whom you consciously or unconsciously were able to convince that Outwitters is nothing more than a game based only on activity. They could become less enthusiastic about keeping on playing the game. So now, can you take responsibility for such statements? Taking out fun from the other players playing the game. You could be more responsible. I just believe you could be more responsible and truthful at the same time.
Total points is also due to skill. To prove this, let me show a good example. Assuming there is an uber active player. Bought the 35 games limit. All 35 game slots always fully loaded with 1v1 league games. The player takes his turns at once and sends them all back to the server in a jiffy. But no matter what he does, he just loses all his games. So his points remain at 0 (or even negative if the system allows it). You cannot say he is inactive just because he has a score of 0. Furthermore, the score of 0 can mean a lot of things. It can exist across all leagues; I have seen it at different times in the various leagues.
So to summarize my point, total points is not solely determined by player activity! It is also determined by player skill! Don't just believe in me or other people. Think for yourself. That's the best.
RE: Is total points really due to player activity alone? - GreatGonzales - 11-08-2012 05:11 AM
Sure, it's both. You don't get points for losing! Still, league placement is always going to be a better indicator of skill than points.
RE: Is total points really due to player activity alone? - aaronINdayton - 11-08-2012 06:12 AM
I know I've said something like "don't worry about points, they're just an indicator of activity."
What I meant by this clearly not 100% accurate statement is that it doesn't matter how many points you have in your league, if everyone is winning 50% of their games because they're matched properly to their skill, the players with more points are simply more active than those with less, within the same league. And all players in that league should be considered of similar skill, assuming they're in the proper league.
It's just a lot faster to say "Points just mean activity."
RE: Is total points really due to player activity alone? - wonderpug - 11-08-2012 06:19 AM
Exactly. There are many things that contribute to how many points you gain or lose. When you look at a point total, though, you have no way of knowing which of those many things contributed. Did they win those points in Fluffy league? In Super Titan? By winning a smaller number of high point games, or by winning a large number of low point games?
The only thing you can be sure of is that someone with a high point total has played a lot, which is why people say that's all that points say.
RE: Is total points really due to player activity alone? - garcia1000 - 11-08-2012 11:38 AM
Well, the player matching system adjusts so that you are expected to have a 50% chance of winning your games. That's the whole point of a player matching system!
So if you were really bad, you would be paired up against worse and worse opponents, until you had a 50% chance of winning.
If you were really good, you would be paired up against better and better opponents, until you had a 50% chance of winning.
The only exceptions are if you are the most terrible worst players, or if you are the most amazing best players (like me).
If you have a 50% chance of winning, on average, every 2 games you will gain 6 points.
So I think ultimately it is true that points are mainly an indication of player activity, in the long run.
RE: Is total points really due to player activity alone? - terenceshiu2005 - 11-08-2012 12:00 PM
I would say yes. the total number of points of all players in Outwitters will be increased ~ 5 to 10 points when a game is over because there will have net gain points(adding more to winner, subtracting less from a loser). Who will get those points for a long run? Should be the active players.
RE: Is total points really due to player activity alone? - daHsu - 11-08-2012 12:27 PM
What's more important than total points is the league you are in. Even though you can get into gifted right away after 5 preliminary matches, from there on the league you are in is purely based on skill, not activity.
i. e. I see players stuck in gifted with a hundred, sometimes three hundred wins. I've managed to get my main account into master, and I have yet to reach 100 wins. (but I'm getting there
)So just know that it shouldn't piss you off about what other people say, just keep winning your matches and hold your head (and your rankings) high.
RE: Is total points really due to player activity alone? - awpertunity - 11-08-2012 02:39 PM
Assuming your skill never changes, you should eventually (theoretically..) be matched up so that you win half your games at all times. At this point, with the +6 point bias, you are gaining an expected 3 points for every game that you play.
The only way to not gain points by playing as many games as possible is if you are losing more than 67% of them, but that shouldn't happen. As your number of games goes to infinity, your points will also go to infinity. Q.E.D.
RE: Is total points really due to player activity alone? - Eijolend - 11-09-2012 12:25 AM
You're right that it's technically not correct to say that activity is the only thing that determines points. But it's still the single most important factor once the system figures out what your skill rating is and from the number of points a player has it's NOT possible to draw a connection to how good the are.
As multiple people have already pointed out, it's possible to get a huge number of points by winning against a lot of fluffy players before getting paired up with better players and ranked properly. (Which is even worse than the point inflation imo)
The thing, what people mean when they "bluntantly state" that points are only made up by activity, is that once the underlying system has placed you correctly your ranking (and therefore your points) shouldn't change anymore because you win as much as you lose - but that's not the case, because there's this +6 or sth point injection thingy in place, so while your hidden skill ranking stays the same, the displayed points will increase over time with games won/played.
If you get better with time, you'll get a bit more points for a while, but you'll end up with the same amount of points as the other guy that is probably far worse but just played a lot. Although it might take you less games to reach that amount of points, your ranking is still worthless and no indication of your skill.
There are other factors as well that arise from this system. If you start 4 games and the matchmaking system is only able to find 3 opponents much lower than you which you beat, and one opponent a bit higher than you that you lose to. Assuming the skill ratings of the opponents are in a way that your rating stays the same, you'll get more points (3 wins) than if you played against 4 people with an equal skill as you and won half the games (2 wins). So even though your performance had the same worth (effect on skill rating), one case will give you more points than the other. So for this case you get more points because of how you were matched (i.e. who else was online at that time) and not because you did better.
I don't think that's a big issue though, since this should even out over time, but activity is still what we're left with.
The only relevant exceptions are people so far above or below the general skill curve, that the system just cannot figure out where to rank them and cannot find appropriate opponents, so the artificial boundaries of the system kick in. (afaik there is a minimum and maximum amount of points you can win/lose from one game) This lets highly ranked players increase their points at a much higher rate, because they get more points and lose less points than they should. (or vice versa for very low skill ratings)
But this applies only to very very few people, that constantly experience running into these boundaries. (I think p1noyboypj mentioned somewhere he only gets the minimum amount of points for most of his wins)
TLDR: Before the system properly determines your skill, points can vary by a lot of factors, which isn't really good in the first place.
After the system has found out your skill rating, your ranking points continue to increase because there are new points generated into the system for every game played. How much you get out of this is mostly based on activity, but also on the opponents you're matched with. If you get better over time, you'll get more points per game than someone who doesn't improve, but the effect on your points is small and indistinguishable from just playing more.
Your experience may vary if you're either very very very good or very very very bad, which is only an edge case.
If "It's based on activity!" is a reasonable way to summarize these things, is open for discussion - but I certainly think it's enough for someone just wanting to get a quick answer to their points mean. Especially since explaining the details doesn't really make the system sound any better.
And still the main problem stays: Points aren't meaningful at all.
RE: Is total points really due to player activity alone? - Thrutchy - 11-09-2012 02:00 AM
I agree that points by themselves aren't that meaningful. But, when you look at both points and wins you can get a rough measure of skill within a given league. Of course league is a more important first-order measure of skill. I look at the ratio to see points/win. I don't think you can access the number of wins when looking at the players in your division (tell me how if you can), but you can see this for your opponent (chat->player's icon). Having the total wins include 1v1 and 2v2's muddies things further, but you can still use the ratio of the sum of points vs. wins. There are plenty of flaws in using this ratio as a measure of skill, but I still think that points with wins can be meaningful.