After some careful thought and consideration i submit this idea as a potential option for a ranking system in either halo4 or any subsequent halo. Some of these ideas are borrowed and some are of my own invention so if I mention something anyone has already proposed then please don’t take offense.
This system would season based with each year being broken up into four seasons. The seasons would be broken up into fall, winter, spring, and summer. Each season would last three months and you would be required to play at least 100 games to be ranked for the season. 100 games seems like a good metric to work off of and though it may appear to be a lot please keep in mind each season lasts three months, two games a day will get you ranked for that season. Also it would prevent people from only playing a handful of games to get ranked for the season. To ensure balanced matches throughout the season players would be placed into brackets based off of their win/loss ratio. These win/loss percentages could represent the different brackets. 0-20%, 20-40%, 40-60%, 60-80%, and 80-100%. These brackets are only an example and could be modified to ensure a balanced distribution of skill as well as population. Furthermore you can continually move up or down in the brackets throughout the season depending on your win/loss ratio. The bracket you are currently in is not a rank and would only be visible to you in your menu while you are searching a particular playlist.
Your persistent overall rank (h2/h3 style) would appear next to your character name in lobbies and your service record. This rank would be based on your win/loss ratios over multiple seasons and is subject to change for better or worse. Meaning if you got better and improved your win/loss ratio average then your persistent rank would raise to reflect it. The inverse would also be true.
To determine your persistent ranks and obtain them there would be specific playlists you would have to play in. Each playlist giving you a unique rank associated it, similar to h2/h3. Some example playlists could be as follows. Snipers (slayer), swat (slayer), big-team (slayer/objective), 4v4 (slayer/objective), and 2v2 (slayer). No default team slayer ranked list. The reason for this is that slaying isn’t the only skill required to be successful player and objective intelligence needs to be taken into consideration in order to truly assess an individuals skill.
In a system like this i believe boosting and selling accounts could be almost completely eliminated. Some measures could also be implemented to further deter these issues. The first would be to not display your highest rank achieved anywhere because it is always subject to change and have your persistent rank be limitless. Meaning that as more seasons take place the higher an individuals persistent rank can become. No more 1-50 think now of 1-infinity. 50 could still be the high rank as far as the base for someone who, on average, finishes with a 94% win/loss ratio or higher each season. But if one continually places within the parameters to receive a rank of 50 then there rank would go up by one rank level with each subsequent season. For example finish the first four seasons with a 94% win/loss ratio then your overall rank would be 53. Since there is no more rank ceiling there will hopefully be a large enough opportunity cost to deter players who engage in activities such as boosting less skilled players,to better win/loss ratios, or selling accounts. Since doing so will have them being left behind by those who choose not to partake in such endeavors.
I believe everything presented provides an excellent foundation for a system that not only deters questionable behavior but also fosters a balanced competitive community. Please feel free to leave your thoughts and opinions, thanks for reading.
Since each season is three months and there can only be four per year i think the ranks above 50 should be symbols instead of a number rank. For example once your average win/loss ratio is 94% or higher you will receive a visible rank of 50. Fast forward to the end the next season at which time you’ve maintained your 94% win/loss ratio. Now when the season ends instead of getting a visible rank of 51 you’ll receive, for example, a symbol of a halo ring. This is similar to the system that was in h2.
I like it but what If you don’t have good Win/loss ratio and You have great skill then you would be paired with less skilled players that you can defeat easily which wouldn’t make it fair.
You never become locked within a particular bracket. As a season unfolds the bracket you are placed in is directly correlated to your win/loss ratio for that season up to that point. The system would track and calculate your win/loss ratio in real-time. Meaning you would never stagnate and the system would place you in the appropriate bracket linked to you current win/loss ratio. Your bracket placement in a season at any given point in time has no relationship to your career win/loss ratio. So at the beginning of each season everyone will have their win/loss ratio reset. I hope that explains things a little better.
How about something even simpler: In objective type games your rank is your W/L ratio (as a decimal to a maximum of 2 places ie 0.00) and in slayer playlists it’s your KDR (again, to 2 decimal places)? You can only boost to a certain extent, and whenever the boosting is over, your stats normalize again. You’ll also be spotted instantly as a booster and reported by other players (similar to what I and other players did in Halo 3 once boosting became common place) in a really short order, and likely ostracized before you can even play enough games to show off your 2.00 KDR to all the little kiddies.
Thoughts?
PS, OT, that system takes way too many games. TrueSkill only takes about 40 games (depending on variables and give or take) to lock you into a skill slot and then the back half of your ranking time would be spent matched up against similar players over and over again, with very little fluctuation. It appears that you’ve taken a good chunk of your idea from here and while there are ideas here that I both agree and disagree with, I have to agree with his insistence on displaying the skill number from the beginning, but that could just be my own level of sick curiosity…
A win/loss ratio bracket has problems as it is not always an accuate measure of skill. The normal skill system (and it’s various tweaks over the years) has the biggest probability of creating a balanced match. Unbalanced matches could lead to good players failing based on a teammate’s poor performance. That teammate may have got there from a few matches of a good team outskilling another. The constant violent flipping of a strict win/loss bracket system just isn’t effective.
Another thing that encourages this unbalance is in a three month span not everyone will be starting at the same time. And throughout those months people will be popping up to get their rating. Think of this analogy: You have a jar full of different varieties of sand. You shake and the sand seperates into seperate layers. You add more sand of a random mix and keep shaking. Before the new sand settles you add yet more sand. By now the fine layers are riddled with pockets of different colored sand. Then you empty and repeat. See the problem?
I don’t fully understand your reference to decimals if you could elaborate more it would be appreciated. Also i don’t agree with ever relating your rank to anything other then your win/loss ratio since the purpose of each match is to win. Halo is also a team game that requires communication so if someone is not willing to seek out like minded individuals to play with then that is on them. Ranked playlists can be played solo but randoms are abundant and should be expected. Those who will thrive will have a team that communicates and works together. As for boosting i am not referring to it in the broken sense that it was abused in h3. That was due to a fundamental flaw in that ranking systems infrastructure. I am simply referring to boosting in the sense of helping someone move into a higher bracket by improving their win/loss ratio for that particular season. Which i don’t see it being possible for someone to be boosted to a 94% win/loss ratio if they aren’t in some way deserving of it. Helping someone boost at something happens in our everyday lives and as such it is inherently impossible for us expect to be able to expel it fully from any potential rank system. Though since this system is based solely on your win/loss ratio and having no relationship with those you play with, i see it being less of an issue then it was in h3.
Now to address your second part,and by the way, thanks for that link i hadn’t seen it before. According to the author him and i are thinking along the same lines in relationship to establishing a rank for a season of play. The author says anywhere between 60-120 games is fair where as i believe it should be set at 100 because that would be enough games to measure someones skill and deriving a percentage from 100 is simple.
As far as the time it will take for you to match with those of similar skill level I see my system needs some revision. I never intended for it to be construed that it would take you 100 games to be placed into competitive matches. But i do see now that there should be a requirement for advancing past a certain bracket before a set number of games have been played. This would need to be implemented so that people who join into a season late and, for example, win ten games in a row, with losing any, aren’t immediately thrown into the highest bracket where they may or may not deserve to be. Since there are five brackets i feel a player should have to play at least ten games in that bracket. Once ten matches have been played if that player has the requisite win/loss ratio to advance to the next bracket then they will be promoted. This solution also aligns more closely with your approximation of how many games it takes to receive a trueskill rating. Now regarding your desire to see a rank from the outset, i understand. I think most people would agree with you. Though the only solution i see would be to shorten the first season but i don’t know what kind of strain, if any, that would put on the integrity of the system as a whole.
To start with every ranking system is going to have its problems. Implement mine and you will inevitably get matched with randoms occasionally, who for whatever reason, can’t compete against the opposing team. Implement a system that is based on personal merits and you are promoting selfish gameplay. Also merit based systems tend to overlook the individuals who are grinding, objective and supports players, in favor of main slayers. The truth is that the purpose of each match is to win and if you lost, regardless of the reason, you should not benefit.
I must stipulate that my proposal merely provides examples for season length, brackets, and ranks. As such any or all of these could be modified. But i will point out that with a three month season you will give everyone, from the casual to the hardcore, the ability to play enough matches to allow the system to accurately assess their skill level.
As for the problem with people joining mid-season and those who would stop once they played their 100 matches would be no different if the system was simply an enhanced clone of h3. In h3 for example many players played till they got their 50 and then moved on to another playlist for fear of losing their rank or simply because they had accomplished a 50. While many others, who also reached 50, continued populating that playlist. Furthermore getting a 50 in h3 could easily be accomplished in less then 100 games. I also don’t see there being a population issue either. Based on past precedents we can expect a portion of those who play 100 games with a 94% win/loss ratio to stop playing and a portion to continue and then there would many more, not in the highest bracket, who continue to play to further improve their ratio in hopes of increasing their rank by seasons end.