minski
Destrier
Posts: 22
Karma: 18
Pilot name: minski
Platform: iOS
|
Post by minski on Oct 15, 2017 17:34:26 GMT -5
|
|
|
|
Post by hurtbot on Oct 19, 2017 9:17:57 GMT -5
Nice that someone took the time to do this, but his analysis is flawed. His first mistake is using league levels as a major indicator. We all know that league levels are meaningless. I'm at Expert 2 at the moment with level 10 bots/weapons on iOS, but I regularly see diamond players with stronger hangers. Second, if you're going to do this, you have to include your win % in the comparisons, since we all know that once you rise above that "magical" 50% win rate the MM goes full on psychotic. If he wanted accurate numbers, he really should have done 2 different sets of comparisons... one above 50% and one below. Then there's the bots... going up against a hanger of maxed out Golems and Vityazs isn't the same as going up against a hanger of maxed Lances and Furies, which isn't the same as going up against hangers full of maxed out Kbots. As it stands his analysis is superficial at best and I seriously doubt it reflects the reality. |
|
minski
Destrier
Posts: 22
Karma: 18
Pilot name: minski
Platform: iOS
|
Post by minski on Oct 19, 2017 9:58:46 GMT -5
Hi Hurtbot, thanks very much for the feedback - much appreciated and some good points. I'd like to respond if I may:
Nikita from Pixonic confirmed at the time of the test that the matchmaker was based on the level of the pilot, and not on bots or hangers. She also confirmed that Pixonic would not be returning to that model in the future. So, based on this feedback, it was understood that league levels were not meaningless, but actually rather relevant.
You make a good observation in that some Diamond players have stronger hangers than you'd expect. This might be for a number of reasons including levelling slowly, purchased upgrades, tanking, or simply the general weapon/bot level increasing over time within the wider game. Either way, we know that at the time of the test, hanger strength was not a factor in the match maker.
The focus of the test was on the match-maker, and its basis for matching: league levels. So, win outcomes wasn't something that were focused upon, as that takes place after match-making has completed and is based on a more complex list of parameters. You did however make an interesting suggestion about testing above and below 50%. Have you any thoughts on how you would you model this given the complexities of multiple squad-mates, and of constantly changing team make-up over a period of days? If so, I'd be interested in working with you on this.
You are quite right that maxed hangers of different bots will pose different strengths. However, as above, at the time of the test, hanger strengths were confirmed to have no bearing on the match-maker, so this wasn't relevant.
You are of course entitled to your opinion and doubts, but bare in mind that one objective behind the test was to put empirical data, from 100 battles, behind a conclusion and avoid subjective statements, urban myths and assumptions. So, for the duration of 100 battles across geographies, league levels, time zone and squad sizes, we know that this does indeed reflect reality. As emphasised in the article, the test was intended to be a rough indication as to what was going on, and the reader was urged to accept the findings in the spirit in which they were intended. Thanks again for taking the time to feedback.
Kind Regards
Minski
|
|
|
|
Post by NerFB on Oct 19, 2017 10:23:33 GMT -5
Thanks minskiAppreciate what you are looking to do, it’s interesting seeing different results. I looked at doing something similar myself, but looking at the individual pilots score, rather than the league. I thought scores would be a granular metric, as a Champion with 5000 at the end of the month isn’t the same as a Champion with 16,000. I recorded all pilot scores, damage, win loss, map for a few weeks, but the introduction of Kbots completely destroyed the baseline. The results looked similar to yours, in that the total Blue and Red pilots score were usually in a few percent of each other, apart from Squads which were based on the Highest squad player (ouch, if you’re Silver or a Gold don’t squad with a Champion....). My gut feeling was that it needs a stable baseline, run over a calendar month, with no new robots or weapons for the Whale Effect, to have realistic result. Fat chance at the moment. ....always remembering that this is such a tiny percentage of the overall battles as to be insignificant..... |
|
|
|
Post by hurtbot on Oct 19, 2017 10:39:25 GMT -5
Hi Hurtbot, thanks very much for the feedback - much appreciated and some good points. I'd like to respond if I may: Nikita from Pixonic confirmed at the time of the test that the matchmaker was based on the level of the pilot, and not on bots or hangers. She also confirmed that Pixonic would not be returning to that model in the future. So, based on this feedback, it was understood that league levels were not meaningless, but actually rather relevant. You make a good observation in that some Diamond players have stronger hangers than you'd expect. This might be for a number of reasons including levelling slowly, purchased upgrades, tanking, or simply the general weapon/bot level increasing over time within the wider game. Either way, we know that at the time of the test, hanger strength was not a factor in the match maker. The focus of the test was on the match-maker, and its basis for matching: league levels. So, win outcomes wasn't something that were focused upon, as that takes place after match-making has completed and is based on a more complex list of parameters. You did however make an interesting suggestion about testing above and below 50%. Have you any thoughts on how you would you model this given the complexities of multiple squad-mates, and of constantly changing team make-up over a period of days? If so, I'd be interested in working with you on this. You are quite right that maxed hangers of different bots will pose different strengths. However, as above, at the time of the test, hanger strengths were confirmed to have no bearing on the match-maker, so this wasn't relevant. You are of course entitled to your opinion and doubts, but bare in mind that one objective behind the test was to put empirical data, from 100 battles, behind a conclusion and avoid subjective statements, urban myths and assumptions. So, for the duration of 100 battles across geographies, league levels, time zone and squad sizes, we know that this does indeed reflect reality. As emphasised in the article, the test was intended to be a rough indication as to what was going on, and the reader was urged to accept the findings in the spirit in which they were intended. Thanks again for taking the time to feedback. Kind Regards Minski Ok... including hard data in an analysis does not automatically make it objective. In fact, there are plenty of people who make lots of money making data mean things it doesn't. The major pharmaceutical companies come to mind. Confirmations from Pixonic mean exactly NOTHING. They have proven themselves to be deceptive and have 0 credibility at this point. Obviously, a MM based on the level of the pilot is literally meaningless. A level 30 pilot could easily be in any league from Bronze on up, so your numbers themselves show a discrepancy between what Pixonic says and what is actually happening. Leagues were introduced to obfuscate the dirty pool that is happening behind the scenes... it is nothing more than window dressing. Pixonic will NEVER disclose how MM actually works, because it is DESIGNED TO BE AN UNFAIR SYSTEM. Numbers like yours gloss over this fact by showing a superficial "fairness" that doesn't exist because the MM is designed to drive sales, not to be fair. There is NO WAY you can claim an objective analysis unless you include relative hanger strengths. |
|
minski
Destrier
Posts: 22
Karma: 18
Pilot name: minski
Platform: iOS
|
Post by minski on Oct 19, 2017 10:42:44 GMT -5
Hi NerFB, Many thanks for taking the time to feedback - much appreciated. I think you could be right in that individual points score might be an even better indicator than league level. Great idea. I feel your pain with the analysis that Kbots affected - you must have put a lot of time and effort into it thus far. You're right in that 2/3 of results did seem within around 10% of so of blue team, although it is worth highlighting that 11% of battles were more than 10% league level in favour of blue and a quarter in favour of red. So, when it was 'unfairly matched' it was REALLY unfair  You may be right about the longer duration test. Personally that's a little more time than I'd be willing to put towards a game that treats its players (or customers, as most firms call them) so poorly. Again you are right about the % of battles being very small. Thanks for taking the time to respond, and if you want to team up for future tests, send me a direct message. Minski |
|
minski
Destrier
Posts: 22
Karma: 18
Pilot name: minski
Platform: iOS
|
Post by minski on Oct 19, 2017 10:59:56 GMT -5
Hurtbot,
Thanks for your comments. I understand your outrage and anger, although I'm not sure why it is directed at me. I have tried to be proactive and provide an indication of status using a test, rather than guesswork or opinion. I welcome constructive criticism, as this drives discussion, although blind rejection of any of the effort, based on assumption, is a little too blunt to be helpful. I think we'll have to agree to disagree.
Thanks
Minski
|
|
|
|
Post by hurtbot on Oct 19, 2017 12:38:29 GMT -5
First, any outrage and anger you detect is a perceptual misstep on your part. Second, nothing was directed at you. While it is certain there was an effort at collecting data, your motives remain in question. What I have pointed out is that you lack sufficient data to arrive at a conclusion. That being said, your statements to this point appear to be leading; distracting from the crux of the discussion while implying an objectivity that isn't there. Again, applying data to an argument does not make it objective on it's own. The data you have shows only that the league status of the players in a given game is more or less balanced. That's it. That's all. Since league status does not reliably correlate to hanger strength and hanger strength is what determines the fairness of the match, your data is, at best, inconclusive with respect to showing fairness. In the interest of full disclosure, the manipulation of data to manipulate people is among my pet peeves and that is what appears to be happening here. If I had to guess, I'd say you either work for pixonic or are a fanboy.
|
|
|
|
Post by Estoplast on Oct 20, 2017 0:24:19 GMT -5
Well... After 3.3 i have seen truly bizarre things in MM. Especially on my smurf account, which is G3. I see diamond regularly. But. Yesterday i saw a lvl10 account with destrier cossack hangar against my big fat Leo. It's worse than 4-6 tankers that i see every match, because usually they can't play very well. But lvl10 account should never see such a match. Never!
After that i decided i will not play the smurf account anymore.
|
|
minski
Destrier
Posts: 22
Karma: 18
Pilot name: minski
Platform: iOS
|
Post by minski on Oct 20, 2017 12:53:59 GMT -5
Hi estoplast - yes, I saw Adrian running a level 12 Destrier recently The level of bots and weapons in lower leagues is rising fast, which turns the War Robots world upside down. We and a few other clans have started Cossack custom games which has been a fun side interest, to take the edge of the madness in the current game. |
|
|
|
Post by Estoplast on Oct 21, 2017 1:57:30 GMT -5
I think You misunderstood. Player level 10. You know, where the best thing you are allowrd to get is Boa or get some cash for Bulgasari.
|
|
ma'ElKoth
GI. Patton
Posts: 138
Karma: 87
Pilot name: ma'ElKoth
Platform: Android
Clan: [F₂P] Ø Ca$h = Ø Daϟh
League: Expert
Server Region: Europe
|
Post by ma'ElKoth on Oct 21, 2017 2:14:35 GMT -5
Hi minski , thank you for all the effort you put into this. I know there is real work and time involved in these things. The focus of the test was on the match-maker, and its basis for matching: league levels. So, win outcomes wasn't something that were focused upon, as that takes place after match-making has completed and is based on a more complex list of parameters. You did however make an interesting suggestion about testing above and below 50%. Have you any thoughts on how you would you model this given the complexities of multiple squad-mates, and of constantly changing team make-up over a period of days? If so, I'd be interested in working with you on this. ... So, for the duration of 100 battles... Would you be able to visualize the existing data in a way that shows how your stats (league levels, "fair) "evolved over time (ie. over the duration of those 100 battles)? I would assume that win rate must have raised above and dropped below those magical 50% multiple times over those 100 battles, so maybe existing data would already show up/down trends over time. Ideally visualization would include whether the battle was actually won or lost, if that data exists for those battles. A won battle would have made the next battle be based on a equal/higher win rate and vice versa. Actually, for the latter 70(?) battles one should be able to calculate the rate based on the earlier battles, if you know whether you won or lost. |
|
minski
Destrier
Posts: 22
Karma: 18
Pilot name: minski
Platform: iOS
|
Post by minski on Oct 21, 2017 4:50:43 GMT -5
Hi ma'ElKoth,
The scope of the test was objective league levels, squad size and a subjective 'did it seem a fair fight'. I'd be happy to share the Win/loss outcomes with you however this was outside of the original scope so wasn't recorded. If you'd like to re-run another test, I'd be happy to work with you on it.
Thanks
Minski
|
|
minski
Destrier
Posts: 22
Karma: 18
Pilot name: minski
Platform: iOS
|
Post by minski on Oct 21, 2017 4:52:08 GMT -5
Hi estoplast - ah - I see what you mean. Yes, this was indeed a strange match up for a level 10 player.
|
|
|
|
Post by NerFB on Oct 21, 2017 5:10:12 GMT -5
I think You misunderstood. Player level 10. You know, where the best thing you are allowrd to get is Boa or get some cash for Bulgasari. That’s awful. Many pilots would just give up after a few of these. |
|
|
|
Post by NerFB on Oct 21, 2017 5:29:52 GMT -5
Hi ma'ElKoth, The scope of the test was objective league levels, squad size and a subjective 'did it seem a fair fight'. I'd be happy to share the Win/loss outcomes with you however this was outside of the original scope so wasn't recorded. If you'd like to re-run another test, I'd be happy to work with you on it. Thanks Minski and minski, Attached is is my template which includes win/loss, map, gold and silver, damage and pilots scores for 70 odd battles in August. The Data is junk, totally irrelevant now ?, due to BR, Kbots etc, Attachment Deletedbut the template might be a starter for 10. I’m not sure if it’s better to start at the beginning of the month, when everyone starts from scratch in their league, or now when it reflects whales spending and rising, Tankers dropping, and F2P pilots getting smashed up in the middle.... I’ll start populating it again from Monday. |
|
|
|
Post by HEATHEN HERETIC on Oct 21, 2017 15:50:41 GMT -5
minskiI was a trouble shooter in manufacturing and warehousing for nearly two decades. I use not just statictical analysis but observational and user feedback as well in my analyses. So I'd be interested to know if you kept all of the match information from every match or if you kicked those that tankers or wildly divergent participants were a part of? The problem with the mm is the fact that it is so easy to tank. Because unless someone at Pix was/is incredibly stupid it was purposely designed to be so. They have made not one move to eliminate that aspect of their mm since it was first initiated. So one can only come to the conclusion that Pixonic actually did purposely build the mm that way. Besides the fact that Pixonic has made no such moves to fix it, I spent six months in their player purgatory, and I can tell you from personal experience and observation that the only tankers found in the Leavers Queue are those too inept to use Pixonic's mm properly to do so. The tankers that are in the LQ got there only because they were leaving/skipping maps too often. I very rarely ran into any player in the LQ below Diamond with anything in their hanger better than 10. And never on a full hanger basis. Others who've been there know this if they were paying attention as I was. It's not the Tankers & Leavers Queue, the name is also by design. Tankers are rewarded with easier opposition until they rank up again while players with lesser upgrades are whipped by their masters into opening up their wallets to compete on "equal" footing with the tankers. In my own experiments and mathmatical observations there are 4-7 to 1 odds, based on the 50/50 win/loss rate, of tanking occurring with your hanger just by being the lowest scorer in any match. The 4-7 is actually based on where you came in in the bottom four points losers in any losing match in the 50/50 scenario. You won't be placed in the Leavers Queue if you do just enough damage to not be considered as leaving a match by Pixonic's standards. Thus very few tankers in the LQ.
|
|
|
|
Post by somepilot on Oct 21, 2017 16:48:48 GMT -5
My match making analysis:
80% games are fixed, MM decides which team will win before hand!
At least 80% of the matches in my case are fixed, at least If no one disconnects.
Only 1 in 5 matches, I see both teams are equal. And we have a real tough match.
I am in Android Expert 2.
|
|
|
|
Post by Russel on Oct 22, 2017 3:34:10 GMT -5
minski I was a trouble shooter in manufacturing and warehousing for nearly two decades. I use not just statictical analysis but observational and user feedback as well in my analyses. So I'd be interested to know if you kept all of the match information from every match or if you kicked those that tankers or wildly divergent participants were a part of? The problem with the mm is the fact that it is so easy to tank. Because unless someone at Pix was/is incredibly stupid it was purposely designed to be so. They have made not one move to eliminate that aspect of their mm since it was first initiated. So one can only come to the conclusion that Pixonic actually did purposely build the mm that way. Besides the fact that Pixonic has made no such moves to fix it, I spent six months in their player purgatory, and I can tell you from personal experience and observation that the only tankers found in the Leavers Queue are those too inept to use Pixonic's mm properly to do so. The tankers that are in the LQ got there only because they were leaving/skipping maps too often. I very rarely ran into any player in the LQ below Diamond with anything in their hanger better than 10. And never on a full hanger basis. Others who've been there know this if they were paying attention as I was. It's not the Tankers & Leavers Queue, the name is also by design. Tankers are rewarded with easier opposition until they rank up again while players with lesser upgrades are whipped by their masters into opening up their wallets to compete on "equal" footing with the tankers. In my own experiments and mathmatical observations there are 4-7 to 1 odds, based on the 50/50 win/loss rate, of tanking occurring with your hanger just by being the lowest scorer in any match. The 4-7 is actually based on where you came in in the bottom four points losers in any losing match in the 50/50 scenario. You won't be placed in the Leavers Queue if you do just enough damage to not be considered as leaving a match by Pixonic's standards. Thus very few tankers in the LQ. : Some things to add: I back up everything that Heathen Cat says (well, except for the fact he was a trouble shooter, obviously). Recently I got really disappointed in a game and decided to analyze some things I was eager to know but was thinking to be inappropriate to do, like LQ, repair costs, low winrate and such. What can I add to your analysis: 1) MatchMaking is ALWAYS fair in LQ. 99 out of 100 you get balanced teams. Each team got an equal number of appropriate league standings. Not so in the Normal queue. I was really really surprised by that. In normal Queue, it's completely normal (no pun intended) to get 1-2 champs on one team+master+experts, and other team gets 2 masters, 2 experts, and 2 diamonds. This, plus the fact that in LQ there should be _LESS_ active players is making me wonder. 2) There are almost NO Experts and Masters in iOS. I walked all the way Expert3 -> Champions and checked every league roster (the one that supposed to have top100 players). All Experts and Masters are usually 86~99 people each tier. While in Champs there were 68 tiers, 100 players each. That makes Expert+Master leagues less than 600 players, while Champions are 6800 players pool, and it's only getting bigger. 3) Battles in Expert was harder for me than battles in Champions. In Champions you are either put against more or less balanced teams or put against some weak team from Diamond-Expert, where you steamroll the opposition really fast. Having said all that, I can't agree with your terminology. Saying that 2\3 of the battles are fair sounds like saying that a bank robber robbed ONLY 10 banks out of 10,000 so he's mostly innocent. My impression is that MM is a highly manipulative algorithm main function of which is to keep you hooked (thus some fair matches) while constantly putting your face in the dirt (hence unbalanced teams and obviously unfair matches). P.s. and another observation I made - I still do NOT think that MM is trying to push your winrate to 50% If you still got the data from the battles - could you check the following theory: Matchmaking is trying to balance TEAM winrate so the mean ("mean" is like the "average", but better, I bet you know the term) would be ~equal for both teams, and close to 50% |
|
minski
Destrier
Posts: 22
Karma: 18
Pilot name: minski
Platform: iOS
|
Post by minski on Oct 22, 2017 7:20:48 GMT -5
Heathen Cat - thanks for the feedback - I completely agree about the observation. This is a very important tool in analysis. The reason I decided to try and run a test is to put some sort of numbers against what I was observing. I wasn't clear which people's view were correct and which were just conjecture. And I totally agree that tanking can be manipulated. A common part of freemium games is the need to frustrate players to drive sales, so I can see why Pixonic aren't keen to solve this issue.
Somepilot - Again, thanks for the stats. Are you able to share your data? I'd be really interested to look through. Also how many battles was this based on? Cheers.
Russel - I found your observation around fairness in LQ really interesting. This isn't something that I've looked at but it does beg the question how fairness compares in LQ and outside LQ. I also find your analysis around numbers of players in different league levels intriguing. How did you get hold of the totals? I agree that I rarely see or meet Master players, although I know and squad with many Expert players. I can we believe the Champ tiers are cavernous, and always wondered whether there was limited space, or it could fill up as long as you made the grade.
Thanks to all three of your for the constructive feedback - I really enjoy the conversation, sharing and learning!
|
|
ma'ElKoth
GI. Patton
Posts: 138
Karma: 87
Pilot name: ma'ElKoth
Platform: Android
Clan: [F₂P] Ø Ca$h = Ø Daϟh
League: Expert
Server Region: Europe
|
Post by ma'ElKoth on Oct 23, 2017 3:02:10 GMT -5
ma'ElKoth and minski , Attached is is my template which includes win/loss, map, gold and silver, damage and pilots scores for 70 odd battles in August. Thanks NerFB for sharing! I have taken a quick look and realized, you are already fighting in Champion, which removes MM's ability to add higher league players. Given that I'd like to see that for a non-squading player whit sufficient room for higher league players, I have now started to capture match results myself (*aaaaargh* stupid me!) to see what it turns up. Just will take some time to gather sufficient match data. |
|
|
|
Post by Russel on Oct 23, 2017 4:48:21 GMT -5
Heathen Cat - thanks for the feedback - I completely agree about the observation. This is a very important tool in analysis. The reason I decided to try and run a test is to put some sort of numbers against what I was observing. I wasn't clear which people's view were correct and which were just conjecture. And I totally agree that tanking can be manipulated. A common part of freemium games is the need to frustrate players to drive sales, so I can see why Pixonic aren't keen to solve this issue. Somepilot - Again, thanks for the stats. Are you able to share your data? I'd be really interested to look through. Also how many battles was this based on? Cheers. Russel - I found your observation around fairness in LQ really interesting. This isn't something that I've looked at but it does beg the question how fairness compares in LQ and outside LQ. I also find your analysis around numbers of players in different league levels intriguing. How did you get hold of the totals? I agree that I rarely see or meet Master players, although I know and squad with many Expert players. I can we believe the Champ tiers are cavernous, and always wondered whether there was limited space, or it could fill up as long as you made the grade. Thanks to all three of your for the constructive feedback - I really enjoy the conversation, sharing and learning! Well it’s kinda easy and hard at the same time :-D You get into a tier and look for top-100 pilots roster - often you see that there are less than 100 players. Then you proceed to next tier, check again. Roster (am I using the word correctly?) shows only first 100 players, so you do not know if there are more. However, you know if there are less than 100. As an example:  As for champs - you are considered a champ if you get more than 5000 trophies, the place is unlimited. Though they are no differently rewarded/badge separated tiers, each tier holds only 100 pilots. Dunno why is that. And only top-50 pilots considered “Legends”, but you need to play a LOT to get there. Last time I checked it was 16,750 league points for iOS Picture is two weeks old, there were a lot more tiers and max league points added since then  When you move up it changes only a bit:  |
|
|
|
Post by Russel on Oct 23, 2017 4:52:22 GMT -5
Oh, and those rosters are the same across countries (checked that by using VPN) and for LQ/NQ (I played in both).
|
|
minski
Destrier
Posts: 22
Karma: 18
Pilot name: minski
Platform: iOS
|
Post by minski on Oct 23, 2017 6:11:19 GMT -5
Russell thanks so much for this - very helpful.
|
|
|
|
Post by HEATHEN HERETIC on Oct 23, 2017 12:30:08 GMT -5
Oh, and those roosters are the same across countries (checked that by using VPN) and for LQ/NQ (I played in both). Yeah, they are roosters alright.   Those are roosters. I think you meant rosters? |
|
Deleted
Deleted Member
Posts: 0
Karma:
|
Post by Deleted on Oct 23, 2017 15:31:27 GMT -5
Russel - I found your observation around fairness in LQ really interesting. This isn't something that I've looked at but it does beg the question how fairness compares in LQ and outside LQ. I also find your analysis around numbers of players in different league levels intriguing. How did you get hold of the totals? I agree that I rarely see or meet Master players, although I know and squad with many Expert players. I can we believe the Champ tiers are cavernous, and always wondered whether there was limited space, or it could fill up as long as you made the grade. I second Russel's statement. I currently play in normal iOS Champion's League, and have been put in a few matches in LQ for skipping / crashing Moon and skipping Springfield. There is usually an equal of amount of players of the same league in the same team which usually results in a match where the results are pretty close. In normal Champion's League (at least right now), matches are very unfair. In almost all matches, there is one team with 2 more Champion / Masters than another team and that other team getting 2 more Experts / Diamonds which results in a lopsided match. |
|
|
|
Post by HEATHEN HERETIC on Oct 23, 2017 17:54:19 GMT -5
Russel - I found your observation around fairness in LQ really interesting. This isn't something that I've looked at but it does beg the question how fairness compares in LQ and outside LQ. I also find your analysis around numbers of players in different league levels intriguing. How did you get hold of the totals? I agree that I rarely see or meet Master players, although I know and squad with many Expert players. I can we believe the Champ tiers are cavernous, and always wondered whether there was limited space, or it could fill up as long as you made the grade. I second Russel's statement. I currently play in normal iOS Champion's League, and have been put in a few matches in LQ for skipping / crashing Moon and skipping Springfield. There is usually an equal of amount of players of the same league in the same team which usually results in a match where the results are pretty close. In normal Champion's League (at least right now), matches are very unfair. In almost all matches, there is one team with 2 more Champion / Masters than another team and that other team getting 2 more Experts / Diamonds which results in a lopsided match. I'm curious. How do their hangers match up in upgrades? If they are 11-12 all throughout isn't it still a fair matching regardless of those ridiculous rankings? Not trying to be confrontational. I'm seriously curious as to what is considered fair matching at those tiers in both iOS and Android. |
|
|
|
Post by Russel on Oct 24, 2017 5:00:29 GMT -5
I second Russel's statement. I currently play in normal iOS Champion's League, and have been put in a few matches in LQ for skipping / crashing Moon and skipping Springfield. There is usually an equal of amount of players of the same league in the same team which usually results in a match where the results are pretty close. In normal Champion's League (at least right now), matches are very unfair. In almost all matches, there is one team with 2 more Champion / Masters than another team and that other team getting 2 more Experts / Diamonds which results in a lopsided match. I'm curious. How do their hangers match up in upgrades? If they are 11-12 all throughout isn't it still a fair matching regardless of those ridiculous rankings? Not trying to be confrontational. I'm seriously curious as to what is considered fair matching at those tiers in both iOS and Android. Pretty equal. I met tankers in LQ only occasional, and every battle (save for Yama\Springfield) was fun to play; Namely: - You can't predict match outcome in first 2-3 minutes
- A LOT of different stuff on the field, which brings diversity and fun
- Most often match ends in meching out, but not in a stupid massacre. Meaning that people were pushing for beacons when in disadvantage, not being afraid of their League score drain
- Players usually either don't show or play whole battle. I can't recall a match where players are ejecting mid-battle or ditching bots in a row.
- It is wild experience in Yamantau or Springfield. Really random. You either get 1 vs 3 game, or play vs snipers or it's just a throwdown brawler match.
- You are almost guaranteed NOT to get an Ancile parade in a battle.
However, downsides are: - Lower silver payout (often each team got 1-2 Bronze\Gold players with lower level bots, so less damage)
- No chance to play against a coordinated team (I met squad only once)
- You MUST have 8+ equipment, otherwise, you will be crushed every game
- Matchmaking for Domination (well, dedicated Domination in the past but anyway) is taking long time (my record is 30min). Beacon rush is ~1min usually though.
|
|
|
|
Post by NerFB on Oct 24, 2017 6:30:27 GMT -5
ma'ElKoth and minski , Attached is is my template which includes win/loss, map, gold and silver, damage and pilots scores for 70 odd battles in August. Thanks NerFB for sharing! I have taken a quick look and realized, you are already fighting in Champion, which removes MM's ability to add higher league players. Given that I'd like to see that for a non-squading player whit sufficient room for higher league players, I have now started to capture match results myself (*aaaaargh* stupid me!) to see what it turns up. Just will take some time to gather sufficient match data. Yes, ma'ElKoth you're absolutely right, but theres a huge difference between the top and bottom of Champion. Its over 21k for the top of the legends V 5K at the bottom. You really feel it when you run into the legends, a top clan squad against randoms (me) last....oh....about 3 minutes. I'm tracking stats, about 30 games to date, and have added Win / Loss ratios. Its a tiny data set, but W/L, map and damage numbers don't follow any trends in iOS, Europe, 5K+ level. |
|
|
|
Post by HEATHEN HERETIC on Oct 24, 2017 7:55:52 GMT -5
I'm curious. How do their hangers match up in upgrades? If they are 11-12 all throughout isn't it still a fair matching regardless of those ridiculous rankings? Not trying to be confrontational. I'm seriously curious as to what is considered fair matching at those tiers in both iOS and Android. Pretty equal. I met tankers in LQ only occasional, and every battle (save for Yama\Springfield) was fun to play; Namely: - You can't predict match outcome in first 2-3 minutes
- A LOT of different stuff on the field, which brings diversity and fun
- Most often match ends in meching out, but not in a stupid massacre. Meaning that people were pushing for beacons when in disadvantage, not being afraid of their League score drain
- Players usually either don't show or play whole battle. I can't recall a match where players are ejecting mid-battle or ditching bots in a row.
- It is wild experience in Yamantau or Springfield. Really random. You either get 1 vs 3 game, or play vs snipers or it's just a throwdown brawler match.
- You are almost guaranteed NOT to get an Ancile parade in a battle.
However, downsides are: - Lower silver payout (often each team got 1-2 Bronze\Gold players with lower level bots, so less damage)
- No chance to play against a coordinated team (I met squad only once)
- You MUST have 8+ equipment, otherwise, you will be crushed every game
- Matchmaking for Domination (well, dedicated Domination in the past but anyway) is taking long time (my record is 30min). Beacon rush is ~1min usually though.
The opposite is true in the LQ for Android for a lot of what you stated. But fun is relative I suppose. I hated every second of it despite having 8-10 equipment. Every single match is wildly divergent in player levels, ranks and upgrades. As a Recruit in the Android LQ you will see Champions, even Legends in your matches. Level 2 players(I posted pics of those in some threads here) with Legends. Imagine the shock of a brand new player in their second match driving a barely/if upgraded Destrier, facing off for a mere half a second against a full 12/12 Tarancilot or 12/12 Zeus's Fury. Full squads of Champion clans stomping their way through opposition, or squad skipping certain maps. That's just some of that "fair gameplay" as Pixonic likes to expect from us players, but not from themselves. |
|
