|
Post by HEATHEN HERETIC on Jun 17, 2017 8:02:35 GMT -5
Apparently I'm dyslexic. I read that wrong. ? Will fix. OK, build this hanger in your system: 2 12/12 acilots, and 2 1/1 cossacks, and a 6/6 Orkin rog. Now make a hanger that has all bots and weapons within 1 level of each other that this hanger will face, make them perfectly even. Challenge accepted. Your hanger averages are 4/6 bot size/upgrades & 4/9 weapons size/upgrades. Or Medium size bot/6 upgrade level average & Medium size weapon/ 9 upgrade level. My answer. 5 × Doc/6 w Taran/Orkan/Tulumbas/9. Exact same hanger averages. And all within 1 size or upgrade level as you specified. Of course a computer would be capable of making all these calculations much faster with more variation. This was just quicker for me. 10 weapons slots vs 20? No my formula doesn't take into account weapons numbers but it could. It's just another variable. If I had here's another hanger... 5 × Rogatka/6 w Taran/Orkin/Tulumbas/9. Same number of weapons. I didn't include weapons numbers because then it would be too even as some have complained. By averaging just sizes and upgrades there is still the element of overcoming disadvantages. Another example... Here's a hanger not to your specs but still averages out the same... Fury/12 w/weapons upgrades 12 Doc/6 w/weapons upgrades 6 Rogatka/6 w/weapons upgrades 6 Boa/6 w/weapons upgrades 8/4 Gepard/1 w/weapons upgrades 1 Same hanger averages by size and upgrades.
|
|
|
Post by zer00eyz on Jun 17, 2017 11:21:21 GMT -5
OK, build this hanger in your system: 2 12/12 acilots, and 2 1/1 cossacks, and a 6/6 Orkin rog. Now make a hanger that has all bots and weapons within 1 level of each other that this hanger will face, make them perfectly even. Challenge accepted. Your hanger averages are 4/6 bot size/upgrades & 4/9 weapons size/upgrades. Or Medium size bot/6 upgrade level average & Medium size weapon/ 9 upgrade level. My answer. 5 × Doc/6 w Taran/Orkan/Tulumbas/9. Exact same hanger averages. And all within 1 size or upgrade level as you specified. Of course a computer would be capable of making all these calculations much faster with more variation. This was just quicker for me. 10 weapons slots vs 20? No my formula doesn't take into account weapons numbers but it could. It's just another variable. If I had here's another hanger... 5 × Rogatka/6 w Taran/Orkin/Tulumbas/9. Same number of weapons. I didn't include weapons numbers because then it would be too even as some have complained. By averaging just sizes and upgrades there is still the element of overcoming disadvantages. Another example... Here's a hanger not to your specs but still averages out the same... Fury/12 w/weapons upgrades 12 Doc/6 w/weapons upgrades 6 Rogatka/6 w/weapons upgrades 6 Boa/6 w/weapons upgrades 8/4 Gepard/1 w/weapons upgrades 1 Same hanger averages by size and upgrades. 12/12 ancilot vs 6/6 doc? That isn't clubbing?
|
|
|
Post by bilbobaggins on Jun 17, 2017 12:09:47 GMT -5
Another flaw in this "fix" (lol)... equipping only one or two bots to lower your hangar score even more.
|
|
|
Post by HEATHEN HERETIC on Jun 17, 2017 12:45:26 GMT -5
12/12 ancilot vs 6/6 doc? That isn't clubbing? My answer to your challenge was 5 × Doc/6 w/weapons upgraded to 9s. Or 5 × Rogatka/6 w/weapons upgraded to 9s. If you're referring to my last example then no I don't. We're not talking about 1v1 bot matches as you seem to keep alluding to. In a 5 slot hanger solo matching you know it should(will?) be 30v30 bot matches. Of which sure there might be a few people running lopsided hangers as you described. But how long will someone with a hanger like that last with the variety of bots and upgrades facing it? As you so aptly indicated in an earlier post you can kick @$$ with a 8/9 hanger vs 11/11 - 12/12 hangers. Out of 6 opponents not one or more will have a bot capable of taking out the Ancilot? I'm just interested in leveling the playing field, taking away the ability of highly upgraded tankers and regular paying customers from taking unfair advantage of other players with lesser upgraded materials by simply outgunning then across their entire hanger. And you know as well as I that it occurs quite often with the current match making system. If a player has a full hanger off 12/12 heavy bots then they should be facing the same. If you're running a full 8/8 medium bot hanger then that's what you should be facing. But I'm not taking the ability to mix a hanger up away from the players. I'm just trying to keep the hanger sizes, bot sizes, weapons sizes and all upgrades evenly matched by averaging them out across a hanger.
|
|
|
Post by zer00eyz on Jun 17, 2017 13:02:48 GMT -5
Some things, the quote was messed up... Sorry I asked for a hanger of bots where the bot levels and weapon levels were +/-1 of each other (your +3).... Though you can push weapons ahead of bots, the returns aren't that great on doing so (in both cost and time) when you can upgrade something else for less time and less cost. Were looking for a more typical hanger vs someone trying to manipulate the system.
|
|
|
Post by HEATHEN HERETIC on Jun 17, 2017 14:03:32 GMT -5
Some things, the quote was messed up... Sorry I asked for a hanger of bots where the bot levels and weapon levels were +/-1 of each other (your +3).... Though you can push weapons ahead of bots, the returns aren't that great on doing so (in both cost and time) when you can upgrade something else for less time and less cost. Were looking for a more typical hanger vs someone trying to manipulate the system. ? Your hanger is +3. To get a match the formula has to match yours. Here's a different hanger.... 1 × Doc/8 w/weapons upgraded to 11 1 × Boa/7 w/weapons upgraded to 10 1 × Rogatka/6 w/weapons upgraded to 9 1 × Golem/5 w/weapons upgraded to 8/9 mix 1 × Vityaz/4 w/weapons upgraded to 7/8 mix or... 1 × Vityaz/8 w/weapons upgraded to 7/8 1 × Boa/6 w/weapons upgraded to 10 1 × Golem/4 w/weapons upgraded to 11 1 × Doc/7 w/weapons upgraded to 6 1 × Rogatka/5 w/weapons upgraded to 5 Again the averaged upgrade differences are +3 as you stated if you were referring to the difference between bot upgrade averages and weapons upgrade averages. Because it has to match your original challenge. That's how my formula works. If you have an average bot size of medium bots(4) and the average upgrade of your bots is 6 then you're going to be matched up against hangers with an average size of medium bots with an average upgrade of 6. The same hold true for weapons. Those are medium sized weapons on average with an average upgrade level of all weapons of 9 in your challenge. I'm not using the number of weapons, just the size and upgrade averages. I could do so as I said, but then it would be too even. I can't match a hanger with only a +/- 1 difference between the average upgrades in total with the one that you gave me originally. The numbers have to match as closely as possible(rounding to the nearest whole number) to what you present in your hanger. I can place a medium bot upgraded to 6 w/weapons upgraded to 6 in the hanger. But then either everything is the same or everything is different. I understand that most players will not be upgrading weapons beyond 1-2 upgrade levels from their bots. But haven't you noticed that some people do run bots that are significantly more or less upgraded in comparison to their weapons? My formula takes that into account as well. That's why my first hanger example looks as lopsided as yours does between bot and weapons upgrade averages. They match.
|
|
|
Post by HEATHEN HERETIC on Jun 17, 2017 21:05:58 GMT -5
Another flaw in this "fix" (lol)... equipping only one or two bots to lower your hangar score even more. There is no possible way to lower your hanger averages below that of every other player in the match. My formula takes in every hanger configuration and matches it by averages against other hangers of similar averages. You can make all the changes you want to your hanger and every single time you would be matched against other hangers with the same hanger averages. It won't matter what your league ranking is either. Your hanger will determine your matching no matter what. Why do you keep insisting that such would be possible? I challenge you to prove it using my formula. Try to create a hanger that cannot be averaged and matched to other hangers. Never forget that there are 6 players on each team and every one is going to be averaged to match your hanger too. Yes, I left room for creating lopsided hangers. That's the perogative of every player right now and that doesn't seem to be a problem for you. But as of right now players can have full hangers of 11-12 bots, be ranked as bronze or silver and be pitted against hangers that are much lower in upgrades. There is no balance in that at all and yet that doesnt seem to be a problem to you. Well it certainly is for most players including myself. How about if that cannot happen anymore? League rankings are for teams. I'm not a team. Neither are you. Why are you happy being ranked as one? I'd much rather be ranked as an individual. I want to be able to compare my statistics to those of other players and actually see who the better players are. Can't do that in any way at all with the current system.
|
|
RagingFire
GI. Patton
Not again! (gets put on canyon and shot to death by trebs)
Posts: 127
Karma: 48
Pilot name: RagingFire
Platform: iOS
Clan: WiKalibur
League: Gold
Server Region: North America
Favorite robot: Jesse
|
Post by RagingFire on Jun 17, 2017 21:19:21 GMT -5
This is a great idea my hanger is mostly level 6 and 7 but I'm often matched with level 9 and 10 hangers
|
|
RagingFire
GI. Patton
Not again! (gets put on canyon and shot to death by trebs)
Posts: 127
Karma: 48
Pilot name: RagingFire
Platform: iOS
Clan: WiKalibur
League: Gold
Server Region: North America
Favorite robot: Jesse
|
Post by RagingFire on Jun 17, 2017 21:21:30 GMT -5
This is a great idea my hanger is mostly level 6 and 7 but I'm often matched with level 9 and 10 hangers
|
|
|
Post by shaolinrogue on Jun 19, 2017 21:23:43 GMT -5
The biggest problem with any system will be accounting for: Above average players Below average players Players that artificially "adjust" their performance level Whether we have a hanger/league/other or blend of the 3 type system, only making rewards climb exponentially the higher level your robot/gear/league/tier has a chance of "discouraging" things like "farming" "clubbing" & "sandbagging. Humans are animals and the easiest way to influence an animals behavior is with treats
|
|
Deleted
Deleted Member
Posts: 0
Karma:
|
Post by Deleted on Jun 19, 2017 21:31:22 GMT -5
"Humans are animals, and the easiest way to influence an animal's behavior is with treats."
Hahahaha love that.
|
|
|
Post by mijapi300 on Jun 19, 2017 23:15:49 GMT -5
OK, build this hanger in your system: 2 12/12 acilots, and 2 1/1 cossacks, and a 6/6 Orkin rog. Now make a hanger that has all bots and weapons within 1 level of each other that this hanger will face, make them perfectly even. Challenge accepted. Your hanger averages are 4/6 bot size/upgrades & 4/9 weapons size/upgrades. Or Medium size bot/6 upgrade level average & Medium size weapon/ 9 upgrade level. My answer. 5 × Doc/6 w Taran/Orkan/Tulumbas/9. Exact same hanger averages. And all within 1 size or upgrade level as you specified. Of course a computer would be capable of making all these calculations much faster with more variation. This was just quicker for me. 10 weapons slots vs 20? No my formula doesn't take into account weapons numbers but it could. It's just another variable. If I had here's another hanger... 5 × Rogatka/6 w Taran/Orkin/Tulumbas/9. Same number of weapons. I didn't include weapons numbers because then it would be too even as some have complained. By averaging just sizes and upgrades there is still the element of overcoming disadvantages. Another example... Here's a hanger not to your specs but still averages out the same... Fury/12 w/weapons upgrades 12 Doc/6 w/weapons upgrades 6 Rogatka/6 w/weapons upgrades 6 Boa/6 w/weapons upgrades 8/4 Gepard/1 w/weapons upgrades 1 Same hanger averages by size and upgrades. So... I told myself I'd ignore this thread but just came across this "easy answer". As a community here, we aren't the most adept exploiters. Yet in a matter of minutes after reading your suggested system, we came up with a "problem hangar". Your solution is that the noobs getting clubbed just simply drop $350 (plus an additional $50 if they don't already have five slots) to get five Docs, then equip them with a total of 20 Tulumbas, and somehow get all 20 of those Tulumbas up to level 9. Even with a $400 investment, your solution for noobs to compete with that hangar will take weeks to develop. Face it. Your suggested system would make things worse. Just accept it so we can all move on. We all try to come up with ideas here, and some of them just aren't perfect - case in point. Nobody is just trying to be mean, but when you make suggestions on this forum you get feedback. Take the feedback and move on.
|
|
|
Post by HEATHEN HERETIC on Jun 20, 2017 8:42:49 GMT -5
mijapi300Where is your example of how my formula fails? I answered each challenge hanger submitted with a hanger equal in averages to it. Try to buy your way to an overpowered hanger and you'll just be facing hangers that will be matched to yours by averages. So you would be exploiting who exactly? It wouldn't matter what your league ranking(if that ridiculous system was still in use) your HANGER would determine your matching. Some of you keep claiming that it would still allow high ranking players to club other players. Only with skill. That much I will admit. But not by hanger. Equipment for equipment all hangers would be practically even. There is nothing wrong with more skillful players playing throughout the matches at every level. It would force better gameplay out of those less skilled. Show me exactly how my formula can be exploited continuously. Submit any hanger, any configuration and I will answer it with a hanger or hangers that match its averages for averages. Every. Single. Time. I can't do a thing about player skill. Only the players can do that. Please remember that this formula is for solo matching. Squad matches are a whole different animal. They would have to be matched by squad member averages combining all members' hanger averages and then dividing by the number of players that are supposed to be in the squad instead. But it could be done easily enough. I never claimed it to be perfect if that is what you're suggesting. No way. I admitted to leaving out the number of weapons as a contributing factor. Yes, some players might find a way to use that as a way to overpower their hangers. But then others would just follow suit or find a way to counter it and make such a tactic useless. There are 12 players in the current matchings(or there are supposed to be 12). All will have a different hanger configuration. And they will have to use them effectively for best results. I want players trying out a variety of things to try to get the best out of their hangers. That's just part of the game. And it's up to their opponents to counter it by learning how to build good hangers or not. It seems to me you're assuming that I think everyone will play with even hanger setups. That is most certainly not the case. Every system has flaws and yes mine are quite obvious even to me. But the current system isn't worth a cat fart. It's openly exploitive and noone is submitting any answers for it. I did and I haven't yet seen anyone prove how it can be exploited as you claim you have. I answered your challenges hanger for hanger. Yet you are the ones claiming victory? I also challenged others to come up with some ideas. Incentives would be great. But what kind and how distributed for best effect? Player skill and good builds would dominate as they should. Mediocre and poor players would just have to learn to play better or accept their skill levels as is. Pixonic claims to have been working on a fix now for over 6 months. More minds working on it might get it done quicker.
|
|
|
Post by ⓣⓡⓘⓒⓚⓨ48 on Jun 20, 2017 9:27:25 GMT -5
My suggestion... fwiw...
Pix needs to have an educational epiphany and hire one of these awesome YTs to make a video explaining how it all works and how that, even when everything looks wonky as far as weapons and bot levels, that there is a measure of fairness. How? Well, after playing on my Silver account that has mostly level 6 with a couple of 7s, and keeping up with and playing better than people with level 11 and 12 weapons, I see more clearly what the intended result was. Skill DOES become the overlying factor after people's league score gets settled. Sure, MM will throw some insane ?poo-poo? at you when you least expect it, but overall it seems like the concept is sound... and working.
Do I like it better than the borderline cheating that the old MM afforded us with the ability to maximize our hangars for a particular tier? IDK... my heart says yes, but my head says no.
Things my head liked about the old one: It was easy to sit at 6/8 at the top of the old Silver tier (for example) and just WRECK everyone that wasn't sitting at the top with me. It was easy to adjust my hangar to run with anyone lower than me and drop to their league. That's about it.
When I think about the games I've had on the new system... I hear my Heart chime in with this argument... "no matter who you were facing, they have all been outstanding quality games as long as no tankers are involved." Blood spattering and Bone-Shaking knock out, drag-out, tactical sessions that had me punching above my weight more often than not and still winning or putting up a damn good fight if on the losing side. On both of my accounts... the 6/6 one and the 9/9 one. I complain about only 1 thing, and this is the one thing that I feel could help it feel better to the most people that matter... Make Silver have a weapons cap (and I used to think Gold too but not now), possibly, so that tankers can't drop down to the seals and club like a raver with a new Molly hook-up. And also, so people don't hang out in Silver and just slowly accumulate weapon levels so that they are 12/12 in freaking Silver. That's it. Everything else is as good as it should get in order to keep from stagnating the tiers/leagues.
The problem peeps seem to have with the new one, is that they like winning, and knowing that they will win. They like winning easily. They shy at the challenge of playing a game that only gives them a chance to win... UNTIL!!! Until they win that first game with a 9/9 hangar that had most Reds sitting on 12/12 bots and weapons... and not only did they win, they had more damage than their higher leveled opponents. Then, it all clicks. That feeling you get from doing more with less, is why I joined the branch of the Military that I did... and it is what changed my mind on the MM.
So, my only suggestion is a little bit more Seal Clubber repellent... but otherwise, so far, so good.
JMO, YMMV.
EDIT: and scaled rewards. THAT is a good damn idea.
|
|
|
Post by mijapi300 on Jun 20, 2017 11:01:30 GMT -5
ⓣⓡⓘⓒⓚⓨ48I like your ideas, and have always been a proponent of scaled rewards. The one thing I'd argue with (just slightly) is the new points distribution doesn't accurately measure skill. When you went up with a win and down with a loss, it more accurately measured a player's skill in the long run. If your teams consistently are winning, and you're the only common variable in all those wins (you weren't being teamed with the same people over and over), that means you were the strong link on all those teams. Or at least one of the strong links. Now, with the focus back on inflicting damage win or lose, it doesn't accurately measure skill. It only measures damage output. And you can place top two in damage while being the one and only reason your teams keep losing, because you keep ignoring proper area control and capturing beacons. I felt like the previous points distribution led to the most balanced matches and made the most sense. But I do agree that the league system has worked - whether people want to admit it or not. And you could always tell when people are complaining about 50% win rates. Complaining that the matchmaker isn't working while simultaneously providing evidence that it is. Always seemed ironic to me.
|
|
|
Post by ⓣⓡⓘⓒⓚⓨ48 on Jun 20, 2017 11:24:34 GMT -5
ⓣⓡⓘⓒⓚⓨ48 I like your ideas, and have always been a proponent of scaled rewards. The one thing I'd argue with (just slightly) is the new points distribution doesn't accurately measure skill. When you went up with a win and down with a loss, it more accurately measured a player's skill in the long run. If your teams consistently are winning, and you're the only common variable in all those wins (you weren't being teamed with the same people over and over), that means you were the strong link on all those teams. Or at least one of the strong links. Now, with the focus back on inflicting damage win or lose, it doesn't accurately measure skill. It only measures damage output. And you can place top two in damage while being the one and only reason your teams keep losing, because you keep ignoring proper area control and capturing beacons. I felt like the previous points distribution led to the most balanced matches and made the most sense. But I do agree that the league system has worked - whether people want to admit it or not. And you could always tell when people are complaining about 50% win rates. Complaining that the matchmaker isn't working while simultaneously providing evidence that it is. Always seemed ironic to me. Fair enough... I don't know all the variables that Pix's stats show, so I can only go by what I am seeing lately compared to previous months... and I am not as dismayed by all of it as I used to be. Whichever point distribution is best, I am happy that things are moving the way they are and hope that the (IMO) most salient points are going to be more evident to people as time goes by. Those points are that: 1.) Gear matters a helluva lot less than it seems to. 2.) Skills can be measured in many ways and not just in league points. 3.) There will always be better teams/players than yours/you, but the opposite is true, too. Wins and losses are the measurement we go by but it isn't always the best measuring stick to show how well we are doing or how bad we are doing. Each game is different and will have different amounts of skill:gear level ratio per player. A "good" MM system will have to be able to account for the disparity while still giving us good matches. Once we get our skills at a level that is considered up to par... that disparity becomes a LOT harder for the average player to see. All they see is the weapon/bot disparity, and they despair. I am going to make a thread that shows a recent match that proves my point to a degree... Even though i know it won't change a lot of minds. I was in that "This Sucks" camp for a while... but I'm pretty sure I am past it now. UPDATED: The thread mentioned is war-robots-forum.freeforums.net/post/129443/threadIMO, YMMV.
|
|
|
Post by mijapi300 on Jun 20, 2017 11:30:33 GMT -5
ⓣⓡⓘⓒⓚⓨ48I was fortunate (or unfortunate?) enough to not really experience the hangar based MM. I started playing two weeks before they scrapped it, so I wasn't playing long enough to experience anything but the wrong side of seal clubbing. That probably had a lot to do with my quick acceptance of the new matchmaker. It also reminded me of my Halo 2 days. Their matchmaker was very simple, if you won games, you leveled up and faced better players. Obviously there weren't as many variable with gear and equipment, but it made sense then, and it makes sense now.
|
|
|
Post by lindenwood on Jul 27, 2017 2:11:23 GMT -5
Give me just two-three Golems leveled to 3 with level 5 weapons and your 12/12 Ancilot is badly damaged or destroyed if I play smart. This is my main problem with your system. You are assuming that players will have the knowledge to manage the game against said Ancilot. That hangar average matches up with probably mid-Bronze, with players who have only been around perhaps a week or so and have barely started developing tactics. I am in high Gold and play exclusively solo and I still often find myself in situations where my teammates dont know how to pool their energies into an Amcilot to take it down. So, in reality, what you have is a guy who will jump around in 4 Cossacks grabbing 6-10 beacons in a 5 minute period. Then he'll spawn his Ancilot and mosey around til the end of the game smashing everyone in sight. There is nooooo way a team of players with a week or two of experience would suddenly know how to deal with this. By far, the best solution has been to simply limit or eliminate awards if you drop more than two tiers below your current one. Something like, for example, if you drop from Diamond 1 to Diamond 3, you start gaining half rewards, and in Gold 1, you gain 1/4 rewards, and in Gold 2 and below, you gain no rewards. This would allow players a little leeway with bad losing streaks (I have lost 200 league points in a single day!), to include hangar experimentation. But, it would almost completely eliminate the incentive to artificially drop your hangar acore by significant amounts.
|
|
|
Post by HEATHEN HERETIC on Jul 27, 2017 6:28:24 GMT -5
Give me just two-three Golems leveled to 3 with level 5 weapons and your 12/12 Ancilot is badly damaged or destroyed if I play smart. This is my main problem with your system. You are assuming that players will have the knowledge to manage the game against said Ancilot. That hangar average matches up with probably mid-Bronze, with players who have only been around perhaps a week or so and have barely started developing tactics. I am in high Gold and play exclusively solo and I still often find myself in situations where my teammates dont know how to pool their energies into an Amcilot to take it down. So, in reality, what you have is a guy who will jump around in 4 Cossacks grabbing 6-10 beacons in a 5 minute period. Then he'll spawn his Ancilot and mosey around til the end of the game smashing everyone in sight. There is nooooo way a team of players with a week or two of experience would suddenly know how to deal with this. By far, the best solution has been to simply limit or eliminate awards if you drop more than two tiers below your current one. Something like, for example, if you drop from Diamond 1 to Diamond 3, you start gaining half rewards, and in Gold 1, you gain 1/4 rewards, and in Gold 2 and below, you gain no rewards. This would allow players a little leeway with bad losing streaks (I have lost 200 league points in a single day!), to include hangar experimentation. But, it would almost completely eliminate the incentive to artificially drop your hangar acore by significant amounts. That's part of the problem with those trying to find fault with this formula. Ya'll seem to be assuming that league ranks would somehow be involved in it. The league ranks are not a consideration in this formula at all. Hanger sizes, equipment size averages and equipment upgrade averages. That's it and nothing else. Come to think of it this formula would make the current ranking system completely irrelevant. Yay!! Rankings on a league level in a game like this is simply ridiculous. League are for teams and I certainly do not see any teams listed in the rankings. So what is its purpose then? It sure isn't indicating what kind of shooter you are. How much damage on average you manage each match. How well you pilot your different machines. Assists on kills. Things that actually can be used to compare your skills to those of others. The current system just shows how lucky you are in who you're matched up with in most cases. As you said you've lost 200 points in a day. That means you finished 3rd or less more times than not over a period of time. And that's all those silly league rankings are going to reflect. Big whoop. But I do understand some of ya'll's concerns and I have been thinking it over. I think the best solution would be an in game tutorial phase, say from level 1-10 that guides players through certain aspects of the game. After that player experience and skill dictates the outcomes of hanger builds and matches. You'll build a hanger the mm can match or you'll not find any matches to play. No tanking, no clubbing, ever. You mentioned rewards and incentives. This formula would make the current ranking system irrelevant as I said. Replace it with a statistical ranking system instead. That would give Pixonic so many more ways to reward players. There could be several categories of rewards. Best shooters by bots. Most damage on average by size of bot. Any number of ways to reward players and keep interest in the game. You might decide to go for a best shooters trophy. Next time maybe a beacon hunters trophy. Best Griffin pilot by averages. All at the same time. Etc, etc, etc.
|
|
|
Post by moses on Sept 1, 2017 23:41:06 GMT -5
Now I see.
This system would combine the problems of both original tier system and the current league based system when put in practice.
1) All bots / weapons are not created equally
The system assumes that all bots / weapons of a given category are of equal value on the battlefield - this is clearly not true and so would require some adjustment factor to be applied subjectively. This is one more subjective metric that would constantly need tweaking behind the scenes. This was an issue with the original,system. Opaqueness
2) players will optimise hangars to be matched with lower players
As illustrated above in the previous responses people will design hangars specifically to optimise their metrics for our system. It may not be as bad as mag gep clubbing, but it is blazingly obvious that guys with balanced 8/9 hangars of plain vanilla bots will get crushed by guys running 2-3x maxed lances / dash bots / butches plus 1-2 low level junk bots that they keep in the hangar to average down. Especially as you will need to lump tiers together (see point 3)
3) you will have a trade accuracy of matchmaking for time to match so you will end up with tiers
As the current system shows, one of the biggest issues faced even with a single numerical metric matching system (as we have with leagues today) is quickly finding people in a narrow spread all wanting to enter a match at equally the same time window.
The current league system is designed to sweep up and down from your league points score expanding the search until the match is full. What we have got instead is consistently games mixed 2-3 leagues in either direction precisely because this matching is very hard to do with the size of the player base. No one wants to wait 5 mins to be matched in a solo game.
Apply this to your system and all of a sudden you need to put in place some kind of tiers. But instead of champions playing with diamond league players you have 12/12 ancilot hangars playing with 8/8 golem hangars. Then 8/8 golem hangars playing 4/4 destroyer hangars.
If you decide not to use an expanding search approach then you will have to fix tiers. E.g. Hangar ratings A-B are first tier, C-D are second and so on. Then what you create is a system where people will be optimising hangar mix specifically with the goal of having the best possible hangar for a given tier - exactly the mag gep problem.
So while it is a nice idea, you run into some important conceptual problems that will require adding significant extra complexity and subjectivity to the hangar rating and then significant practical problems on the implementation.
|
|
|
Post by T34 on Sept 2, 2017 0:55:48 GMT -5
The biggest problem with any system will be accounting for: Above average players Below average players Players that artificially "adjust" their performance level Whether we have a hanger/league/other or blend of the 3 type system, only making rewards climb exponentially the higher level your robot/gear/league/tier has a chance of "discouraging" things like "farming" "clubbing" & "sandbagging. Humans are animals and the easiest way to influence an animals behavior is with treats Yes but than you introduce an in game pricing issue and a revenue issue for pix.
|
|
|
Post by moses on Sept 2, 2017 1:09:22 GMT -5
The biggest problem with any system will be accounting for: Above average players Below average players Players that artificially "adjust" their performance level Whether we have a hanger/league/other or blend of the 3 type system, only making rewards climb exponentially the higher level your robot/gear/league/tier has a chance of "discouraging" things like "farming" "clubbing" & "sandbagging. Humans are animals and the easiest way to influence an animals behavior is with treats Yes but than you introduce an in game pricing issue and a revenue issue for pix. And it somehow seems unfair to be weighting the rewards to those who presumably have already pretty full hangars with top equipment, etc. People in higher leagues are already doing well enough, the tankers are just trying to cheat more out of the system. Maybe better to have a maximum league drop - for example, say you could only drop max 2 full leagues below your highest ever league - so if you make it to champions league, no matter what you do then you will never fall below expert. Since experts through champions are in the same pool anyway it won't make a difference to those players and will keep the sharks out of silver, bronze, etc
|
|
|
Post by carnage on Sept 2, 2017 2:12:34 GMT -5
Serious necro posting here, even though I agree with moses on his arguments. To me a good MM should be based on the actual VALUE of the best robot. All it would take here is to set a calculation value for WSP and gold to be translated in silver and we would simply have a value associated with each robot. You have some games using that system, and it works great. It's clear, it's transparent and it's fair. But anyway, this ship has sailed a long time ago. I can live with the current MM, my only problem is the fact that more often than not, I play players from higher leagues. Until Diamond II, I had a good MM. Suddenly when you arrive in Diamond I, it's a total mess and that really kills the point of having leagues if anyway the MM doesn't respect that. I would have no problem waiting a few seconds more for a proper matchup than being thrown versus a bunch of champions with my level 9 hangar. That's really my only real problem with the MM currently. Pix, you created leagues, just please respect them in your MM, seems kind of logical to me.
|
|
|
Post by moses on Sept 2, 2017 2:29:35 GMT -5
Serious necro posting here, even though I agree with moses on his arguments. To me a good MM should be based on the actual VALUE of the best robot. All it would take here is to set a calculation value for WSP and gold to be translated in silver and we would simply have a value associated with each robot. You have some games using that system, and it works great. It's clear, it's transparent and it's fair. But anyway, this ship has sailed a long time ago. I can live with the current MM, my only problem is the fact that more often than not, I play players from higher leagues. Until Diamond II, I had a good MM. Suddenly when you arrive in Diamond I, it's a total mess and that really kills the point of having leagues if anyway the MM doesn't respect that. I would have no problem waiting a few seconds more for a proper matchup than being thrown versus a bunch of champions with my level 9 hangar. That's really my only real problem with the MM currently. Pix, you created leagues, just please respect them in your MM, seems kind of logical to me. Brought it back up as Heathen Cat and I had discussed on another thread where it was off topic. Agree fully that the main issue with current MM is the failure to deliver on the promise to actually match in a tight group of league points. On the rare occasion when that happens the match is usually pretty good.
|
|
|
Post by ᕲΣΣᕲƧ on Sept 2, 2017 14:52:00 GMT -5
League ranking is a poor metric to match from. We all know it's easy to go down. But going up above where you should be is also not that difficult. I just changed clans and went from expert 2 to master 2 in about a week. It's amazing how much easier it is to win when you have full squads and aren't depending on randoms. So my league rating shows how the whole squads I play with perform instead of truly representing my skill level. Also getting minus points for a loss doesn't make sense to me. You could finish third on the losing team, have the 10 kills, 7 beacons, and 800k damage and still go backwards. Yeah I know the point is to win not rack up personal stats, but stats do show your skill.
|
|