NovusAnimus, thank you for pointing out a number of (important) things I had completely forgot to mention myself. You succeed in clarifying some of the biggest positive aspects of this ranking system.
<!--quoteo(post=1681627:date=Jun 20 2008, 03:27 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 20 2008, 03:27 PM) <a href="index.php?act=findpost&pid=1681627"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->In the end, awarding a player for winning the match is better than awarding the player for individual actions. It may not be what we want, it may have problems it'll cause with stacking or too competitive of play or whatever, but so far it's proving to be the lesser of the many evils I've seen so far. The only other option is no ranking system, which has many of its own issues shown in NS already.<!--QuoteEnd--></div><!--QuoteEEnd-->
This has been tacitly exposed previously, but I like your explicit remark better: either have a worthwile rank (that is to say none of the classic FPS ranking system seen so far), either have nothing. I prefer having the former, some others the latter. A compromise would consist in a division of ranked and unranked game.
<!--quoteo(post=1681627:date=Jun 20 2008, 03:27 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 20 2008, 03:27 PM) <a href="index.php?act=findpost&pid=1681627"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Try to make this ranking system as complicated as you want, the more complicated the worse it's gonna get and the more problems/exploits/griefers/stackers. I'm for using the ranking system that's been used since the dawn of sports: your team wins, it gets the point of winning, loser gets nothing (in this case, team point goes into player's personal rank).<!--QuoteEnd--></div><!--QuoteEEnd-->
I should have said this long ago. I think I even intended to at some point, but lost track of that particular thought in the midst of this morass of parleys.
Most of the critics have stated that this ranking system is overcomplicated while in fact it is the very opposite. What makes this statement even more abberant is that these same persons have presented -before this topic and even after this topic- several ranking systems that are ridiculously complex, especially in regard to whatever potential they might have (that is to say nigh none, in my opinion). Trying to circumvene the sole factor which should be used (team success) just creates a void which you can, of course, attempt to fill with an infinite number of evermore complex factors, but without ever reaching a satisfactory result.
<!--quoteo(post=1681627:date=Jun 20 2008, 03:27 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 20 2008, 03:27 PM) <a href="index.php?act=findpost&pid=1681627"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->A simple ratio of games won to games played, with perhaps a very small penalty for games uncompleted seems simple enough. Also, some sort of timer to make sure a player doesn't jump from server to server looking for games with 20 seconds left, just so they can join the winning team. I suggest a minimum of 5-10 minutes on a team to get the benefit of the win.<!--QuoteEnd--></div><!--QuoteEEnd-->
Nicely thought out variants.
I would still prefer the initial version, but if these modifications would be necessary for this ranking system to be widely accepted I would naturally be in favor of them. Reasons why I do not overjoy over your propositions: - a small penalty is not coercitive enough and could even prove to cause a perverse effect since it would simply prod the players into quitting matches instead of finishing them (and thus lose less points). And there is no real point in establishing a large penalty in the place of a simple penalty equivalent to a loss. Also, if you're thinking of using some sort of time limit in the end for dividing quitters with a small penalty and those with a loss penalty, it would not be a lot more effective: players would just quit sooner when the risk of losing would be too high. In fact, it could be even more detrimental for the gaming, because dropping players would become a nuisance way before the end of the match and could provoke a domino effect amongst the players of the losing team. - letting people join games in the middle of a match and reap the profit of a win by playing 5-10 minutes is troublesome because rankwh0res would undoubtedly abuse this method. The sheer number of unskilled players who would probably misuse the ranking system could seriously unbalance and screw the whole ladder up, not to mention that it would stir a feeling of unfairness among other players. Another problem, which may not occur as fundamental to you as it does for me, is that you mingle a highly unpredictable element (a totally new player with a variable skill) to a team which has already established a strategy and a cooperative intelligence (i.e. through knowledge of other players' skills and roles) and thus interfere not only with the teamwork but with the overall harmony in the gameplay experience (cf. what I think as a "clean" experience is a set of players from the beginning till the end, without people being afk, quitting, joining, dorking right and left). - you're not sticking to what you said before...you're adding surfeit complexity.
<!--quoteo(post=1681627:date=Jun 20 2008, 03:27 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 20 2008, 03:27 PM) <a href="index.php?act=findpost&pid=1681627"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Edit: Oh, and for the heck of it people can have a profile that contains ratios and percentages of things done, like you're K:D and percentage of games where you gorge'd for 50% of the match's length and such, just to appease people's ego, but have no effect on their rank.<!--QuoteEnd--></div><!--QuoteEEnd-->
Yes. Agreed. I don't know if I have or haven't already mentioned this somewhere before, but there you go. You've put it out in the open.
<!--quoteo(post=1681641:date=Jun 20 2008, 05:20 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 20 2008, 05:20 PM) <a href="index.php?act=findpost&pid=1681641"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Teamplay is more important than individual skill in NS, this isn't CS. A ranking system showing off a person's individual skill is a flawed one in the first place for NS because of this. Teamplay is what counts.<!--QuoteEnd--></div><!--QuoteEEnd-->
Although I don't disagree with your statements here (NS requires more teamplay than CS), I do disagree with what is kind of implied (this ranking system is convenient for NS, not necessarily CS). I think this ranking system could (and should) replace any kind of ranking found in all of the existing team-based FPS games. So far, they have been more than crappy. I have always indulged these systems by thinking that future games would ultimately fix them, but this has never been done...is it that ranking is percieved as such an unimportant issue? I find that hard to believe or in the very least foolish (cf. wake-up call for EA), because ranking has such strong influence on server balance and the overall gaming experience.
I don't think I've made it too complicated, still seems quite simple to me, but I'm a game programmer, complicated is what I work with =P
Perhaps a penalty for dropping from a game that increases the closer to the end of the game you are. Leaving a game with 30 minutes left gives you, say 1% of the penalty, while leaving the game with 1 minute left gives you 100% of the penalty. Make the penalty decently bad too.
<!--quoteo(post=1681885:date=Jun 24 2008, 02:41 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 24 2008, 02:41 PM) <a href="index.php?act=findpost&pid=1681885"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->I don't think I've made it too complicated, still seems quite simple to me, but I'm a game programmer, complicated is what I work with =P<!--QuoteEnd--></div><!--QuoteEEnd-->
Yeah, I just noticed that I might have exagerrated a bit...it wouldn't overcomplicate the system.
<!--quoteo(post=1681885:date=Jun 24 2008, 02:41 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 24 2008, 02:41 PM) <a href="index.php?act=findpost&pid=1681885"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Perhaps a penalty for dropping from a game that increases the closer to the end of the game you are. Leaving a game with 30 minutes left gives you, say 1% of the penalty, while leaving the game with 1 minute left gives you 100% of the penalty. Make the penalty decently bad too.<!--QuoteEnd--></div><!--QuoteEEnd-->
I like the idea and it could be made to something very functional. However, you would also have to take into consideration the size of the teams in order to measure the impact such an exit would make...so you would get more penalized if you left a small team than a large team.
The only -rather considerable- apprehension I would have would be that players quitting in the middle of a session could alter other players' gaming experience. In addition to what I already stated above (about a "clean" gaming experience), I fear that the fact that a player would leave a game would actually penalize himself less than it would all of his teammates (who obviously should not be penalized at all, even the contrary). Let me explain this in greater detail: say one or a couple of players quit 20 minutes before the end and "only" get a 20% penalty. What about the other players in the team who will lose those teammates and eventually might end up losing 100%? They could be seriously handicapped if the joining player replacing a quitter is really bad or had trouble adapting to the role and strategy imparted to him by his predecessor (i.e. he's bad at fading although that is what the last player did very well and that is what the team needs). Even worse, what if no one joins the server (quickly) to take the quitters' place? Moreoever, who would want to join a server in the middle of a session when you would probably have to join the losing team (since it is highly probable that a quitter does leave his team because it is not in good shape or just not as good as the opposing team)? Also, if you take into account that you lose less points if you quit sooner and with a larger team, the first ones to quit would actually be less penalized although they might be the biggest reason for the team loss and any latent stampede. If 1 or 2 important players quit a team and 2-3 minutes later it becomes evident that because of this the team has set course to a loss, other players might panic and leave the team as well. Ironically, it is the last valiant players standing that will lose the most points!
Hence why I am still more in favor of a fixed starting and ending point for a game, with no possibility of joining or quitting the session. The only viable (but rather complicated) solution to fix your suggestion would be to somehow "reward" players who stayed from the beginning till the end, so as to actually penalize them less than 100% if they did in fact lose the match instead of quitting while other players did quit. Seems like a complex algorithm to put in place, in my opinion, but maybe you can figure it out?
Fixed ending and starting positions would work better, if this were a competitive sport with designated teams. Alas, it's not, it's only like that for clan matches and skrims and such, and I know a lot of people would like to have ranked pubs, and a ranking system that forces players into fixed teams would destroy pubs. No one would wanna play on them. The ranking system needs to work with ######s and twats (most people) who'll leave the game early in mind, make the penalty too harsh and no one will play on ranked servers at all.
As for an algorithm, trial and error. The leaving early penalty should be worse than losing a game, but I figure only at around 50%. You'll get 1 to 100% of that leaving early penalty depending on how long the match is going on, and I guess how many people are in the game. That way it'll be better to stick around if you're losing than to leave right before it's over. Still, allowing a player to leave when he has to when it's viable, ie: 30 minutes left in the match but a player needs to leave for supper, allow him to do this without destroying his rank (so, like I said, 1% of the penalty), should work. It'll dissuade people from leaving as sticking around to lose is better (you won't know you're going to lose when there's 30 minutes left in the match normally), but the penalty for leaving way early won't be huge enough to piss people off entirely. There is no reason a person would leave a match when it's 3 on 3 and there's only 2 minutes left in the game though. I say 100% penalty on those ######s. The idea is to find a spot where losing a game is better than leaving a game when the game is almost over, but leaving the game is better than losing the game when there's 30 minutes left in the game and you need to leave for a wedding or something.
I use the term penalty loosely, it'd take trial and error to find a good amount to hurt a person's rank by using this penalty variable.
I totally understand your point of view and agree with it to some extent. Nonetheless, I have a few observations to make:
<!--quoteo(post=1682070:date=Jun 27 2008, 05:54 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 27 2008, 05:54 PM) <a href="index.php?act=findpost&pid=1682070"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Fixed ending and starting positions would work better, if this were a competitive sport with designated teams. Alas, it's not, it's only like that for clan matches and skrims and such, and I know a lot of people would like to have ranked pubs, and a ranking system that forces players into fixed teams would destroy pubs. No one would wanna play on them.<!--QuoteEnd--></div><!--QuoteEEnd-->
It has been already stated a few times that this ranking system could be described as "semi-competitive". Here are 2 statements related to that: <i>If it seems like a competitive rule-set, then so be it, but I think many will seek such an environment (without the trials and tribulations of joining a clan).</i>
<i>there are players (such as myself) who are neither satisfied with random pubs nor competitive play. Random pubs offer, like their name suggests, too random of a gameplay and chaotic gaming environments. Competitive play requires too much involvement, time and skill. A global ranking system would combine the benefits of both, without their drawbacks; it would permit easy access to somewhat competitive play and controlled environment, without demanding as much involvement, time and skill.</i>
That's my opinion in any case. Like I've also said before, players who want to play semi-competitively could easily muster up the extra requirements (such as waiting for other players and having fixed starting and ending points to a game). And from what I've seen in other genres (e.g. RTS') people are quite happy with these type of gaming conditions. Plus, nothing stops the developers from adding the typical FPS' ranking-per-session/map to the unranked section of the game. So basically this ranking system does not, in any rate, suppress any previous classical element (as K:D rankings), but just adds a much necessary and desired feature.
<!--quoteo(post=1682070:date=Jun 27 2008, 05:54 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 27 2008, 05:54 PM) <a href="index.php?act=findpost&pid=1682070"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->The ranking system needs to work with ######s and twats (most people) who'll leave the game early in mind, make the penalty too harsh and no one will play on ranked servers at all.<!--QuoteEnd--></div><!--QuoteEEnd-->
The thing I've already explained as well is that that ######s and twats will stay low in ranks which means that they will only be a disturbance at low levels. Plus, the higher you get in rankings, the less players of your level there will be to play with. Eventually, you will be able to spot and remember the players you play with; as I've also said, it would be less about the servers you play in, but more about the players you play with.
<!--quoteo(post=1682070:date=Jun 27 2008, 05:54 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 27 2008, 05:54 PM) <a href="index.php?act=findpost&pid=1682070"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->As for an algorithm, trial and error. The leaving early penalty should be worse than losing a game, but I figure only at around 50%.<!--QuoteEnd--></div><!--QuoteEEnd-->
I'm not sure if I understood you exactly right, but you're saying that leaving early would cause more harm than actually losing the game? So instead of losing 10 points, I would lose 50% more, that is to say 15 points? Is that right? On the rest of your explanation, you got me completely lost...you lose more or less points in the beginning or in the end? In either case, I got the feeling that you've kind of changed your mind by wanting the penalty to be more harsh, even more than losing in certain conditions (e.g. <i>eaving early penalty should be worse than losing a game</i>). So if quitting is penalized more than losing or almost as much (even at half the penalty, I think it's close enough), wouldn't it be just simpler to penalize everything at 100%?
Personally, I think asking for a sitdown of a maximum of half an hour is not excessive. And sure, players will sometimes absolutely need to quit and suffer the penalty, but that happens randomly to everyone and thus remains fair.
First, the basis for rank: Games played, games won, and games left early. The more games won, the higher your rank, the more games left early, the lower your rank, and games played acts as the basis for these ratios. (losing a game also lowers your rank, but that's part of the Games Played condition, so there's no need to track games lost)
Let's say we have a player profile class, it includes 3 variables, GW (for Games Won), GP (for Games Played) and GL (for penalties from Games Left early).
You also have a variable to be used as a basis for calculating penalties, let's call it PNTL, and make it equal 2.
Situation 1: You leave the game early, so there's half an hour left in the game. Teams were 10 on 10 so it's not that big of a deal. We'll assign you 1% of PNTL, so this action will damage your rank by 1% of PNTL, or 0.01 * 2, which comes out to 0.02. 0.02 is a very small amount, as a game won is worth 1 point. So we increase the players' GL by 0.02;
Situation 2: You leave the game with only 1 minute left, and teams were 4 on 4. The fact you left just completely screwed your team over, so we'll give that player 100% of PNTL. We just damaged his rank by by 100% of PNTL, or 1.00 * 2, which comes out to 2. 2 is a very big number, as a game won is worth 1 point. So we increase the player's GL by 2.
Situation 3: You leave the game with 15 minutes left, and teams were 6 on 6. This late in the game, leaving will cause significant damage, but your team may still win. Thus, we damage the player's rank by 50% of PNTL, or 0.50 * 2, which comes out to 1. 1 is a decently large number, as a game won is worth 1 point. So we increase the player's GL by 1.
Now, to calculate rank. Let's calculate Bob's rank: Bob has 300 wins and 200 losses, for a total of 500 games. So Bob->GW = 300, Bob->GP = 500. Bob also has a penalty total of... say 25.3413 (from leaving at various times in various games), so Bob->GL = 25.3413. To calculate Bob's RATIO, which his rank will be based on, we do some simple math: (Bob->GW - Bob->GL)/Bob->GP. Or in this case, (300 - 25.3413)/500, or (274.6587)/500, or 0.5493, or 54.93%.
So Bob's ratio is 54.93%, which is decent, as 0% is as low as you can get and 100% is as high as you can get. Someone with a higher ratio has a higher rank, and vice versa for lower ratios. (you can include a system that forces a player to have played at least X matches before their rank starts getting calculated, doesn't matter to me)
This way, a person can leave way early and it won't affect their rank too much, but if they leave when they're really needed, it'll affect their rank more than if they just stuck around and lost. Asking a person to stick around for a full half hour or suffer severe consequences, even when there's still 30 minutes left in the match is not viable. Players will not stand for that, unless you wanna create two ranking systems, professional and casual. This ranking system I've described would be casual. Professional, which is what you want, could be used for more hardcore servers I suppose, I wouldn't mind playing on a server that forces a person to stick around for the full match, but I know a LOT of people would not play on those servers. The idea to please everyone with a system that works, not just the NS junkies.
<!--quoteo(post=1682128:date=Jun 28 2008, 06:46 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 28 2008, 06:46 PM) <a href="index.php?act=findpost&pid=1682128"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->So Bob's ratio is 54.93%, which is decent, as 0% is as low as you can get and 100% is as high as you can get. Someone with a higher ratio has a higher rank, and vice versa for lower ratios. (you can include a system that forces a player to have played at least X matches before their rank starts getting calculated, doesn't matter to me)<!--QuoteEnd--></div><!--QuoteEEnd-->
I know you might be using ratios just for the purpose of this detailed illustration, but I would just like to clear out one thing: this ranking system could not be based on players' ratios as points are essential for calculating how much a win or a loss weighs on a players' rank. Your ratio, on the other hand, treats every win and loss the same way.
<!--quoteo(post=1682128:date=Jun 28 2008, 06:46 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 28 2008, 06:46 PM) <a href="index.php?act=findpost&pid=1682128"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->This way, a person can leave way early and it won't affect their rank too much, but if they leave when they're really needed, it'll affect their rank more than if they just stuck around and lost.<!--QuoteEnd--></div><!--QuoteEEnd-->
Your suggestion is honourable, but I still think it unnecessarily complicates the simple formula of a Win:Loss ratio.
Firstly, there's a serious flaw in your reasoning. You say that this ranking system is not viable because people need to be able to quit a session without being penalized too severely, but then you also say that people leaving at about half-way into the game will actually get penalized more sternly! So your suggestion basically only works for players who need to quit just at the beginning of a game as opposed to players who need to quit later on. This flaw is particularly unconvenient when taking into consideration the fact that players who might need to leave are probably not the ones having started a game 5 minutes beforehand, but rather players who didn't think/plan that they would be stuck in the game for more than say 25-30 minutes.
Secondly, repeating somewhat what I said in my last reply, your suggestion actually ends up punishing quitters more harshly than losers, although your initial argument is that quitters need to be treated less severely. You have to agree with the fact that at least (and probably more than) half the quitters (who leave toward the end of the game) will get penalized more strictly than losers.
Thirdly, and coming back to what I also said before, you have to agree with the fact that the other half of quitters (who leave at the beginning of them game) are the ones who get punished the less although they are the ones who cause the most trouble and might be the reason for a team loss. If you think the ranking system I've suggested is unfair for a quitter, then I have to say that your suggestion is very unfair to all the remaining players.
I respect your good intentions and I fully understand the problem you're pinpointing at. However, although your suggestion manages to solve this particular problem, it ends up creating bigger problems.
<!--quoteo(post=1682128:date=Jun 28 2008, 06:46 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 28 2008, 06:46 PM) <a href="index.php?act=findpost&pid=1682128"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Asking a person to stick around for a full half hour or suffer severe consequences, even when there's still 30 minutes left in the match is not viable. Players will not stand for that, unless you wanna create two ranking systems, professional and casual.<!--QuoteEnd--></div><!--QuoteEEnd-->
How do you explain that players from all the other genres (especially RTS') stand for that kind of treatment then?
In addition, don't you think that casual players might be less interested in the rankings and wouldn't even mind if they lost some points for quitting early on?
<!--quoteo(post=1682128:date=Jun 28 2008, 06:46 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 28 2008, 06:46 PM) <a href="index.php?act=findpost&pid=1682128"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->This ranking system I've described would be casual. Professional, which is what you want, could be used for more hardcore servers I suppose, I wouldn't mind playing on a server that forces a person to stick around for the full match, but I know a LOT of people would not play on those servers. The idea to please everyone with a system that works, not just the NS junkies.<!--QuoteEnd--></div><!--QuoteEEnd-->
Although I'm not agreeing with your distinction between casual and professional players, I will simply reply that a division with ranked and unranked sections would sort this problem out.
Well, I do want quitters to be punished more severely than losers, but only when they have the nerve to leave with only 5 minutes left in the match, when their leaving really affects who wins/loses. I want to punish them less severely when they quit before their presence becomes too important, ie: very early in the match.
<!--quoteo(post=1682251:date=Jun 30 2008, 02:43 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 30 2008, 02:43 PM) <a href="index.php?act=findpost&pid=1682251"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Well, I do want quitters to be punished more severely than losers, but only when they have the nerve to leave with only 5 minutes left in the match, when their leaving really affects who wins/loses. I want to punish them less severely when they quit before their presence becomes too important, ie: very early in the match.<!--QuoteEnd--></div><!--QuoteEEnd-->
I heavily disagree with what you said here and I think you are mixing things up a bit.
Players who quit early on are the ones causing the most trouble and their exit is probably the reason why a team will veer toward a loss. Players who quit later on are the ones causing the less trouble and their exit is probably not the reason why a team will veer toward a loss.
To be as clear as possible, here is my explanation: The players who quit at the end of the game of course deserve to be punished. They've played throughout the whole session and their team has lost despite their contribution. In all fairness, they should be punished at least as much as a player who stayed in the game and lost. However, the fact that they quit is <b>probably</b> not the <b>reason</b> behind the loss. If they leave a couple of minutes before the end, it is highly probable that they did this precisely because the team was already losing and in any case a team must be in a pretty bad shape for it to lose within a couple of minutes just because one of its team-members quit. The players who quit at the beginning of the game do <b>not necessarily</b> deserve to be punished. For one, their team might win the game in spite of them leaving. Also, of course, they haven't been playing for that long so it seems natural to punish them less, right? That's the dilemma here, because these players actually contribute <b>probably</b> more to the loss than the players quitting at the very end. As I've stated before, players quitting in the beginning are even more astute and troublesome than the ones leaving at the end for several reasons: - Some players might of course leave in the beginning for any number of valid reasons, but other players (rankwh0res in particular) might be practicing a more shrewd scheme on a regular basis: they watch how well their team fares, as opposed to the opponents', and based on that information stay or leave the game (knowing that the penalty will be minimal) - The game is at a much more critical stage in the beginning than in the end: having even 1 teammember less affects the team much more in the beginning than in the end. - The team is handicapped at least for the time for a replacement to show up and this replacement will not necessarily have the same skills and not immediately be as operational as his predecessor. Plus, the replacement might never show up or at least for a big period of time. And ultimately who would want to join a team who is missing players and is probably on a less good shape than the opposing team? - The fact that one player quits might finally lead to other players quitting as well. Just imagine one team already having lost some critical territory already in the beginning of the game and due to that one player leaves (which would be even worse, if it was the commander). Don't you think that players worrying about their rank might also be inclined to leave their partners at this moment just to be assured they get minimal loss?
So to sum it up. All in all the players who quit at a late stage of the game are much less to blame for the loss than the ones quitting in the beginning. The latter ones might have contributed for the loss for a much smaller period of time, but their impact is probably far greater: a player who quits at the end might have been playing for 30 minutes and might have done a few small mistakes during that time that would have in the end cost the game. But think of the player who quits at the very beginning, having played only for 5 minutes and already been able to create mass disarray and dismay within his team! The ones quitting in the end are naturally to be punished at least as severly as losers. Then again, the ones quitting in the beginning should be punished at least as much as the other quitters since they are actually more to blame. So, you end up with punishing all quitters at least as much as losers, which means you can as well punish all quitters with a simple loss.
Oh, and i forgot about another thing: what to do with players joining in the middle of the game?
If a player joins the losing team at either the middle or the end of the game and quits, how should he be punished? He might have only played for 5 minutes, so should he be just slightly punished? Also, what to do with players who play for a minimal amount of time and end up winning? Do they get 100% profit for that?
Comments
You succeed in clarifying some of the biggest positive aspects of this ranking system.
<!--quoteo(post=1681627:date=Jun 20 2008, 03:27 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 20 2008, 03:27 PM) <a href="index.php?act=findpost&pid=1681627"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->In the end, awarding a player for winning the match is better than awarding the player for individual actions. It may not be what we want, it may have problems it'll cause with stacking or too competitive of play or whatever, but so far it's proving to be the lesser of the many evils I've seen so far. The only other option is no ranking system, which has many of its own issues shown in NS already.<!--QuoteEnd--></div><!--QuoteEEnd-->
This has been tacitly exposed previously, but I like your explicit remark better: either have a worthwile rank (that is to say none of the classic FPS ranking system seen so far), either have nothing.
I prefer having the former, some others the latter. A compromise would consist in a division of ranked and unranked game.
<!--quoteo(post=1681627:date=Jun 20 2008, 03:27 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 20 2008, 03:27 PM) <a href="index.php?act=findpost&pid=1681627"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Try to make this ranking system as complicated as you want, the more complicated the worse it's gonna get and the more problems/exploits/griefers/stackers. I'm for using the ranking system that's been used since the dawn of sports: your team wins, it gets the point of winning, loser gets nothing (in this case, team point goes into player's personal rank).<!--QuoteEnd--></div><!--QuoteEEnd-->
I should have said this long ago. I think I even intended to at some point, but lost track of that particular thought in the midst of this morass of parleys.
Most of the critics have stated that this ranking system is overcomplicated while in fact it is the very opposite. What makes this statement even more abberant is that these same persons have presented -before this topic and even after this topic- several ranking systems that are ridiculously complex, especially in regard to whatever potential they might have (that is to say nigh none, in my opinion).
Trying to circumvene the sole factor which should be used (team success) just creates a void which you can, of course, attempt to fill with an infinite number of evermore complex factors, but without ever reaching a satisfactory result.
<!--quoteo(post=1681627:date=Jun 20 2008, 03:27 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 20 2008, 03:27 PM) <a href="index.php?act=findpost&pid=1681627"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->A simple ratio of games won to games played, with perhaps a very small penalty for games uncompleted seems simple enough. Also, some sort of timer to make sure a player doesn't jump from server to server looking for games with 20 seconds left, just so they can join the winning team. I suggest a minimum of 5-10 minutes on a team to get the benefit of the win.<!--QuoteEnd--></div><!--QuoteEEnd-->
Nicely thought out variants.
I would still prefer the initial version, but if these modifications would be necessary for this ranking system to be widely accepted I would naturally be in favor of them.
Reasons why I do not overjoy over your propositions:
- a small penalty is not coercitive enough and could even prove to cause a perverse effect since it would simply prod the players into quitting matches instead of finishing them (and thus lose less points). And there is no real point in establishing a large penalty in the place of a simple penalty equivalent to a loss. Also, if you're thinking of using some sort of time limit in the end for dividing quitters with a small penalty and those with a loss penalty, it would not be a lot more effective: players would just quit sooner when the risk of losing would be too high. In fact, it could be even more detrimental for the gaming, because dropping players would become a nuisance way before the end of the match and could provoke a domino effect amongst the players of the losing team.
- letting people join games in the middle of a match and reap the profit of a win by playing 5-10 minutes is troublesome because rankwh0res would undoubtedly abuse this method. The sheer number of unskilled players who would probably misuse the ranking system could seriously unbalance and screw the whole ladder up, not to mention that it would stir a feeling of unfairness among other players. Another problem, which may not occur as fundamental to you as it does for me, is that you mingle a highly unpredictable element (a totally new player with a variable skill) to a team which has already established a strategy and a cooperative intelligence (i.e. through knowledge of other players' skills and roles) and thus interfere not only with the teamwork but with the overall harmony in the gameplay experience (cf. what I think as a "clean" experience is a set of players from the beginning till the end, without people being afk, quitting, joining, dorking right and left).
- you're not sticking to what you said before...you're adding surfeit complexity.
<!--quoteo(post=1681627:date=Jun 20 2008, 03:27 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 20 2008, 03:27 PM) <a href="index.php?act=findpost&pid=1681627"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Edit: Oh, and for the heck of it people can have a profile that contains ratios and percentages of things done, like you're K:D and percentage of games where you gorge'd for 50% of the match's length and such, just to appease people's ego, but have no effect on their rank.<!--QuoteEnd--></div><!--QuoteEEnd-->
Yes. Agreed. I don't know if I have or haven't already mentioned this somewhere before, but there you go. You've put it out in the open.
<!--quoteo(post=1681641:date=Jun 20 2008, 05:20 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 20 2008, 05:20 PM) <a href="index.php?act=findpost&pid=1681641"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Teamplay is more important than individual skill in NS, this isn't CS. A ranking system showing off a person's individual skill is a flawed one in the first place for NS because of this. Teamplay is what counts.<!--QuoteEnd--></div><!--QuoteEEnd-->
Although I don't disagree with your statements here (NS requires more teamplay than CS), I do disagree with what is kind of implied (this ranking system is convenient for NS, not necessarily CS).
I think this ranking system could (and should) replace any kind of ranking found in all of the existing team-based FPS games. So far, they have been more than crappy. I have always indulged these systems by thinking that future games would ultimately fix them, but this has never been done...is it that ranking is percieved as such an unimportant issue? I find that hard to believe or in the very least foolish (cf. wake-up call for EA), because ranking has such strong influence on server balance and the overall gaming experience.
Perhaps a penalty for dropping from a game that increases the closer to the end of the game you are. Leaving a game with 30 minutes left gives you, say 1% of the penalty, while leaving the game with 1 minute left gives you 100% of the penalty. Make the penalty decently bad too.
Yeah, I just noticed that I might have exagerrated a bit...it wouldn't overcomplicate the system.
<!--quoteo(post=1681885:date=Jun 24 2008, 02:41 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 24 2008, 02:41 PM) <a href="index.php?act=findpost&pid=1681885"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Perhaps a penalty for dropping from a game that increases the closer to the end of the game you are. Leaving a game with 30 minutes left gives you, say 1% of the penalty, while leaving the game with 1 minute left gives you 100% of the penalty. Make the penalty decently bad too.<!--QuoteEnd--></div><!--QuoteEEnd-->
I like the idea and it could be made to something very functional. However, you would also have to take into consideration the size of the teams in order to measure the impact such an exit would make...so you would get more penalized if you left a small team than a large team.
The only -rather considerable- apprehension I would have would be that players quitting in the middle of a session could alter other players' gaming experience. In addition to what I already stated above (about a "clean" gaming experience), I fear that the fact that a player would leave a game would actually penalize himself less than it would all of his teammates (who obviously should not be penalized at all, even the contrary).
Let me explain this in greater detail: say one or a couple of players quit 20 minutes before the end and "only" get a 20% penalty. What about the other players in the team who will lose those teammates and eventually might end up losing 100%? They could be seriously handicapped if the joining player replacing a quitter is really bad or had trouble adapting to the role and strategy imparted to him by his predecessor (i.e. he's bad at fading although that is what the last player did very well and that is what the team needs). Even worse, what if no one joins the server (quickly) to take the quitters' place? Moreoever, who would want to join a server in the middle of a session when you would probably have to join the losing team (since it is highly probable that a quitter does leave his team because it is not in good shape or just not as good as the opposing team)?
Also, if you take into account that you lose less points if you quit sooner and with a larger team, the first ones to quit would actually be less penalized although they might be the biggest reason for the team loss and any latent stampede. If 1 or 2 important players quit a team and 2-3 minutes later it becomes evident that because of this the team has set course to a loss, other players might panic and leave the team as well. Ironically, it is the last valiant players standing that will lose the most points!
Hence why I am still more in favor of a fixed starting and ending point for a game, with no possibility of joining or quitting the session. The only viable (but rather complicated) solution to fix your suggestion would be to somehow "reward" players who stayed from the beginning till the end, so as to actually penalize them less than 100% if they did in fact lose the match instead of quitting while other players did quit. Seems like a complex algorithm to put in place, in my opinion, but maybe you can figure it out?
As for an algorithm, trial and error. The leaving early penalty should be worse than losing a game, but I figure only at around 50%. You'll get 1 to 100% of that leaving early penalty depending on how long the match is going on, and I guess how many people are in the game. That way it'll be better to stick around if you're losing than to leave right before it's over. Still, allowing a player to leave when he has to when it's viable, ie: 30 minutes left in the match but a player needs to leave for supper, allow him to do this without destroying his rank (so, like I said, 1% of the penalty), should work. It'll dissuade people from leaving as sticking around to lose is better (you won't know you're going to lose when there's 30 minutes left in the match normally), but the penalty for leaving way early won't be huge enough to piss people off entirely. There is no reason a person would leave a match when it's 3 on 3 and there's only 2 minutes left in the game though. I say 100% penalty on those ######s. The idea is to find a spot where losing a game is better than leaving a game when the game is almost over, but leaving the game is better than losing the game when there's 30 minutes left in the game and you need to leave for a wedding or something.
I use the term penalty loosely, it'd take trial and error to find a good amount to hurt a person's rank by using this penalty variable.
<!--quoteo(post=1682070:date=Jun 27 2008, 05:54 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 27 2008, 05:54 PM) <a href="index.php?act=findpost&pid=1682070"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Fixed ending and starting positions would work better, if this were a competitive sport with designated teams. Alas, it's not, it's only like that for clan matches and skrims and such, and I know a lot of people would like to have ranked pubs, and a ranking system that forces players into fixed teams would destroy pubs. No one would wanna play on them.<!--QuoteEnd--></div><!--QuoteEEnd-->
It has been already stated a few times that this ranking system could be described as "semi-competitive". Here are 2 statements related to that:
<i>If it seems like a competitive rule-set, then so be it, but I think many will seek such an environment (without the trials and tribulations of joining a clan).</i>
<i>there are players (such as myself) who are neither satisfied with random pubs nor competitive play.
Random pubs offer, like their name suggests, too random of a gameplay and chaotic gaming environments.
Competitive play requires too much involvement, time and skill.
A global ranking system would combine the benefits of both, without their drawbacks; it would permit easy access to somewhat competitive play and controlled environment, without demanding as much involvement, time and skill.</i>
That's my opinion in any case. Like I've also said before, players who want to play semi-competitively could easily muster up the extra requirements (such as waiting for other players and having fixed starting and ending points to a game). And from what I've seen in other genres (e.g. RTS') people are quite happy with these type of gaming conditions.
Plus, nothing stops the developers from adding the typical FPS' ranking-per-session/map to the unranked section of the game. So basically this ranking system does not, in any rate, suppress any previous classical element (as K:D rankings), but just adds a much necessary and desired feature.
<!--quoteo(post=1682070:date=Jun 27 2008, 05:54 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 27 2008, 05:54 PM) <a href="index.php?act=findpost&pid=1682070"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->The ranking system needs to work with ######s and twats (most people) who'll leave the game early in mind, make the penalty too harsh and no one will play on ranked servers at all.<!--QuoteEnd--></div><!--QuoteEEnd-->
The thing I've already explained as well is that that ######s and twats will stay low in ranks which means that they will only be a disturbance at low levels.
Plus, the higher you get in rankings, the less players of your level there will be to play with. Eventually, you will be able to spot and remember the players you play with; as I've also said, it would be less about the servers you play in, but more about the players you play with.
<!--quoteo(post=1682070:date=Jun 27 2008, 05:54 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 27 2008, 05:54 PM) <a href="index.php?act=findpost&pid=1682070"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->As for an algorithm, trial and error. The leaving early penalty should be worse than losing a game, but I figure only at around 50%.<!--QuoteEnd--></div><!--QuoteEEnd-->
I'm not sure if I understood you exactly right, but you're saying that leaving early would cause more harm than actually losing the game? So instead of losing 10 points, I would lose 50% more, that is to say 15 points? Is that right?
On the rest of your explanation, you got me completely lost...you lose more or less points in the beginning or in the end? In either case, I got the feeling that you've kind of changed your mind by wanting the penalty to be more harsh, even more than losing in certain conditions (e.g. <i>eaving early penalty should be worse than losing a game</i>).
So if quitting is penalized more than losing or almost as much (even at half the penalty, I think it's close enough), wouldn't it be just simpler to penalize everything at 100%?
Personally, I think asking for a sitdown of a maximum of half an hour is not excessive. And sure, players will sometimes absolutely need to quit and suffer the penalty, but that happens randomly to everyone and thus remains fair.
First, the basis for rank: Games played, games won, and games left early. The more games won, the higher your rank, the more games left early, the lower your rank, and games played acts as the basis for these ratios. (losing a game also lowers your rank, but that's part of the Games Played condition, so there's no need to track games lost)
Let's say we have a player profile class, it includes 3 variables, GW (for Games Won), GP (for Games Played) and GL (for penalties from Games Left early).
You also have a variable to be used as a basis for calculating penalties, let's call it PNTL, and make it equal 2.
Situation 1: You leave the game early, so there's half an hour left in the game. Teams were 10 on 10 so it's not that big of a deal. We'll assign you 1% of PNTL, so this action will damage your rank by 1% of PNTL, or 0.01 * 2, which comes out to 0.02. 0.02 is a very small amount, as a game won is worth 1 point. So we increase the players' GL by 0.02;
Situation 2: You leave the game with only 1 minute left, and teams were 4 on 4. The fact you left just completely screwed your team over, so we'll give that player 100% of PNTL. We just damaged his rank by by 100% of PNTL, or 1.00 * 2, which comes out to 2. 2 is a very big number, as a game won is worth 1 point. So we increase the player's GL by 2.
Situation 3: You leave the game with 15 minutes left, and teams were 6 on 6. This late in the game, leaving will cause significant damage, but your team may still win. Thus, we damage the player's rank by 50% of PNTL, or 0.50 * 2, which comes out to 1. 1 is a decently large number, as a game won is worth 1 point. So we increase the player's GL by 1.
Now, to calculate rank. Let's calculate Bob's rank:
Bob has 300 wins and 200 losses, for a total of 500 games. So Bob->GW = 300, Bob->GP = 500.
Bob also has a penalty total of... say 25.3413 (from leaving at various times in various games), so Bob->GL = 25.3413.
To calculate Bob's RATIO, which his rank will be based on, we do some simple math:
(Bob->GW - Bob->GL)/Bob->GP. Or in this case, (300 - 25.3413)/500, or (274.6587)/500, or 0.5493, or 54.93%.
So Bob's ratio is 54.93%, which is decent, as 0% is as low as you can get and 100% is as high as you can get. Someone with a higher ratio has a higher rank, and vice versa for lower ratios. (you can include a system that forces a player to have played at least X matches before their rank starts getting calculated, doesn't matter to me)
This way, a person can leave way early and it won't affect their rank too much, but if they leave when they're really needed, it'll affect their rank more than if they just stuck around and lost. Asking a person to stick around for a full half hour or suffer severe consequences, even when there's still 30 minutes left in the match is not viable. Players will not stand for that, unless you wanna create two ranking systems, professional and casual. This ranking system I've described would be casual. Professional, which is what you want, could be used for more hardcore servers I suppose, I wouldn't mind playing on a server that forces a person to stick around for the full match, but I know a LOT of people would not play on those servers. The idea to please everyone with a system that works, not just the NS junkies.
Thanks, I got it this time.
<!--quoteo(post=1682128:date=Jun 28 2008, 06:46 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 28 2008, 06:46 PM) <a href="index.php?act=findpost&pid=1682128"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->So Bob's ratio is 54.93%, which is decent, as 0% is as low as you can get and 100% is as high as you can get. Someone with a higher ratio has a higher rank, and vice versa for lower ratios. (you can include a system that forces a player to have played at least X matches before their rank starts getting calculated, doesn't matter to me)<!--QuoteEnd--></div><!--QuoteEEnd-->
I know you might be using ratios just for the purpose of this detailed illustration, but I would just like to clear out one thing: this ranking system could not be based on players' ratios as points are essential for calculating how much a win or a loss weighs on a players' rank. Your ratio, on the other hand, treats every win and loss the same way.
<!--quoteo(post=1682128:date=Jun 28 2008, 06:46 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 28 2008, 06:46 PM) <a href="index.php?act=findpost&pid=1682128"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->This way, a person can leave way early and it won't affect their rank too much, but if they leave when they're really needed, it'll affect their rank more than if they just stuck around and lost.<!--QuoteEnd--></div><!--QuoteEEnd-->
Your suggestion is honourable, but I still think it unnecessarily complicates the simple formula of a Win:Loss ratio.
Firstly, there's a serious flaw in your reasoning. You say that this ranking system is not viable because people need to be able to quit a session without being penalized too severely, but then you also say that people leaving at about half-way into the game will actually get penalized more sternly!
So your suggestion basically only works for players who need to quit just at the beginning of a game as opposed to players who need to quit later on. This flaw is particularly unconvenient when taking into consideration the fact that players who might need to leave are probably not the ones having started a game 5 minutes beforehand, but rather players who didn't think/plan that they would be stuck in the game for more than say 25-30 minutes.
Secondly, repeating somewhat what I said in my last reply, your suggestion actually ends up punishing quitters more harshly than losers, although your initial argument is that quitters need to be treated less severely. You have to agree with the fact that at least (and probably more than) half the quitters (who leave toward the end of the game) will get penalized more strictly than losers.
Thirdly, and coming back to what I also said before, you have to agree with the fact that the other half of quitters (who leave at the beginning of them game) are the ones who get punished the less although they are the ones who cause the most trouble and might be the reason for a team loss. If you think the ranking system I've suggested is unfair for a quitter, then I have to say that your suggestion is very unfair to all the remaining players.
I respect your good intentions and I fully understand the problem you're pinpointing at. However, although your suggestion manages to solve this particular problem, it ends up creating bigger problems.
<!--quoteo(post=1682128:date=Jun 28 2008, 06:46 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 28 2008, 06:46 PM) <a href="index.php?act=findpost&pid=1682128"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Asking a person to stick around for a full half hour or suffer severe consequences, even when there's still 30 minutes left in the match is not viable. Players will not stand for that, unless you wanna create two ranking systems, professional and casual.<!--QuoteEnd--></div><!--QuoteEEnd-->
How do you explain that players from all the other genres (especially RTS') stand for that kind of treatment then?
In addition, don't you think that casual players might be less interested in the rankings and wouldn't even mind if they lost some points for quitting early on?
<!--quoteo(post=1682128:date=Jun 28 2008, 06:46 PM:name=NovusAnimus)--><div class='quotetop'>QUOTE(NovusAnimus @ Jun 28 2008, 06:46 PM) <a href="index.php?act=findpost&pid=1682128"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->This ranking system I've described would be casual. Professional, which is what you want, could be used for more hardcore servers I suppose, I wouldn't mind playing on a server that forces a person to stick around for the full match, but I know a LOT of people would not play on those servers. The idea to please everyone with a system that works, not just the NS junkies.<!--QuoteEnd--></div><!--QuoteEEnd-->
Although I'm not agreeing with your distinction between casual and professional players, I will simply reply that a division with ranked and unranked sections would sort this problem out.
I heavily disagree with what you said here and I think you are mixing things up a bit.
Players who quit early on are the ones causing the most trouble and their exit is probably the reason why a team will veer toward a loss.
Players who quit later on are the ones causing the less trouble and their exit is probably not the reason why a team will veer toward a loss.
To be as clear as possible, here is my explanation:
The players who quit at the end of the game of course deserve to be punished. They've played throughout the whole session and their team has lost despite their contribution. In all fairness, they should be punished at least as much as a player who stayed in the game and lost. However, the fact that they quit is <b>probably</b> not the <b>reason</b> behind the loss. If they leave a couple of minutes before the end, it is highly probable that they did this precisely because the team was already losing and in any case a team must be in a pretty bad shape for it to lose within a couple of minutes just because one of its team-members quit.
The players who quit at the beginning of the game do <b>not necessarily</b> deserve to be punished. For one, their team might win the game in spite of them leaving. Also, of course, they haven't been playing for that long so it seems natural to punish them less, right? That's the dilemma here, because these players actually contribute <b>probably</b> more to the loss than the players quitting at the very end. As I've stated before, players quitting in the beginning are even more astute and troublesome than the ones leaving at the end for several reasons:
- Some players might of course leave in the beginning for any number of valid reasons, but other players (rankwh0res in particular) might be practicing a more shrewd scheme on a regular basis: they watch how well their team fares, as opposed to the opponents', and based on that information stay or leave the game (knowing that the penalty will be minimal)
- The game is at a much more critical stage in the beginning than in the end: having even 1 teammember less affects the team much more in the beginning than in the end.
- The team is handicapped at least for the time for a replacement to show up and this replacement will not necessarily have the same skills and not immediately be as operational as his predecessor. Plus, the replacement might never show up or at least for a big period of time. And ultimately who would want to join a team who is missing players and is probably on a less good shape than the opposing team?
- The fact that one player quits might finally lead to other players quitting as well. Just imagine one team already having lost some critical territory already in the beginning of the game and due to that one player leaves (which would be even worse, if it was the commander). Don't you think that players worrying about their rank might also be inclined to leave their partners at this moment just to be assured they get minimal loss?
So to sum it up. All in all the players who quit at a late stage of the game are much less to blame for the loss than the ones quitting in the beginning. The latter ones might have contributed for the loss for a much smaller period of time, but their impact is probably far greater: a player who quits at the end might have been playing for 30 minutes and might have done a few small mistakes during that time that would have in the end cost the game. But think of the player who quits at the very beginning, having played only for 5 minutes and already been able to create mass disarray and dismay within his team!
The ones quitting in the end are naturally to be punished at least as severly as losers. Then again, the ones quitting in the beginning should be punished at least as much as the other quitters since they are actually more to blame. So, you end up with punishing all quitters at least as much as losers, which means you can as well punish all quitters with a simple loss.
If a player joins the losing team at either the middle or the end of the game and quits, how should he be punished? He might have only played for 5 minutes, so should he be just slightly punished?
Also, what to do with players who play for a minimal amount of time and end up winning? Do they get 100% profit for that?