Search the Community
Showing results for tags 'match maker'.
Found 3 results
The Case Against The 3/5/7 Match Maker (updated with a 400 game sample size in each MM) A few months ago I studied the changes to the match maker (MM) implemented in patch 9.18 and noted that it pushed players into more matches as a low tier vehicle, as would logically happen with a template that is built like a pyramid with more tanks at the bottom than the top. Initially I was neutral about this change and was only curious about how it would affect the game in general and more specifically me as a player. After a few months of playing and studying the effects of this new system I find it to be an overall negative to the game in general and a frustrating bore to me as a player. I wanted to make my viewpoint about this issue clear from the start so that the player reading this who is invested in defending Wargaming's new system can take a moment to pause and consciously listen to my arguments with an open mind instead of becoming defensive as you read. This is difficult for most of us to accomplish, including myself, as our egos are involved, but I am hopeful that some readers will overcome this hurdle. My arguments for removing the 3/5/7 template system are straightforward and backed up with numbers. My main points are that the system failed to achieve its stated goal and that the tradeoffs we surrendered to get this system were not worth giving up for what we received in exchange for them. We lost some obvious and some not so obvious aspects of the game when WG implemented this policy. As a side note, I will use "3/5/7 template" to describe the new system even though it also has other variations (such as 5/10 and all the same tier matches) since that is the name WG has given it and since most matches are in the 3/5/7 format during game play. As I said above, I was open minded about the new template when it arrived in 9.18. It was bundled with a lot of large changes if you remember. In the same patch we also received tier 10 light tanks, reusable consumables and vast artillery changes (stun effect, limited to 3 per side, and removal of artillery from platoons). WG didn't exactly sneak the 3/5/7 template into the game since it was a featured change in their communications to the player base, but since it was part of a larger bundle of changes, I think the player base noticed it less than the other changes and they simply believed the company line that the new template would bring more balance to matches by ensuring an abundance of bottom tier tanks. How would a higher amount of bottom tier tanks in a match create more balanced games? To quote WG in their 9.18 Update video, they said "this will guarantee that you find your target," and this is the only reasoning they gave to us in North America (as far as I know) on how flooding the bottom tiers of a match would balance teams and make more competitive games when this change came out. People have since claimed that they made the change to help weaker players by giving them more targets to shoot at. While this may or may not be true, the original stated purpose of this system to create more competitive matches has not materialized to any significant degree. By their own standards, this template was a failure and that alone should be enough to reverse their decision, but don't hold your breath. In an article titled "Matchmaking in 9.19.1 and the Road Ahead" published on WG's North America website on June 28, 2017, they have doubled down on the 3/5/7 template and have stated that it is successful. How did they come to this conclusion? They arrived at it by reading forum comments about how much happier players are with the new MM. After reading this statement, I had two issues with it. First, only a small minority of players actually post on the forums and the players that consistently post on the forums generally tow the company line on most issues for whatever reason. And second, the weaker players that post seem to be happier because they supposedly have more targets to shoot at in matches. It is the side effect of the 3/5/7 template that players are happy with and not the main objective that was stated by WG. In that same article mentioned above, WG is now pushing the 3/5/7 template more on allowing players to contribute rather than on balancing games as they did back in April. They have shifted their emphasis. They write, "remember the less-than-desirable scenario of playing Tier VIII and getting matched against Tier X? Now think how it feels in the game as it stands today. Having a few top-tiers on the other camp no longer makes you question the point of battling." (Apparently they have missed or ignored the angry complaints on the forums about the new system and being bottom tier all the time in tier 8 matches). So WG has changed their stance from balancing teams to allowing players to contribute more as a bottom tier tank. This shift in emphasis is rather peculiar, but in their defense, they are trying to sell us an idea in this case and a product in general and they will do it however they can. Now if we look at the unintended (or intended?) side effect of this 3/5/7 template change, that of helping weaker players as a bottom tier tank by giving them more targets than they had before 9.18, it could still be a good change and not a total failure. The perception that many have now is that players always have targets to shoot at and players feel that they can always make a difference in this new system. That is what WG told us and that is what they believe. But is it true that these weaker players are now better off or does it even matter if they are better off if the overall gaming experience is being ruined for the greater population? Let's take a closer look at what happened when they implemented the 3/5/7 template. The biggest and most obvious change was the frequencies of being selected top tier, middle tier, or bottom tier in a match. To get information for this I used WoT Replay Analyzer to inspect 400 games from tier 5 to 8 before and after the 9.18 update. I excluded light tanks and tanks what get special tiering. Otherwise, I selected the matches in four separate 100 game periods from before 9.18 to collect that part of the data set. I took an old sample of 100 games that I had under the 9.18 update and added another 300 games from under the 9.19 update to also give some diversity to that sample. Due to the new system having 6 likely breakdowns instead of the old 3, I combined those 6 into 3 to match the old system.* For top tier, I combined 3/5/7 and 5/10 matches as a top tier. For middle tier, I combined middle tier (3/5/7) and all the same tier matches since they were closest to each other in dynamics. For bottom tier, I combined bottom tier games in the 3/5/7 and 5/10 formats. Here is the breakdown of what I found: top, middle, and low tier in the old system were 36.0%, 35.5%, and 28.5% respectively and in the 3/5/7 template system they were 7.0%, 36.0%, and 57.0% respectively. Before I gathered the data I guessed that the old system would break down into about a 1/3 for each tier. This perception was wrong on my part. The middle tier games were slightly greater at about 36% and my memory of being top and bottom tier were also off by a little. I was shocked that bottom tier only occurred 28.5% of the time. I thought it would have been a little more than that. This shows how faulty our memories can sometimes be. As regards the 3/5/7 template, I already had a good idea of what those numbers would be since I had studied them already and posted them here a few months ago. My greatest curiosity was in gathering the second set of 3/5/7 numbers (another 300 games) and whether or not they would still mirror the old numbers. They generally did, except that there were more overall bottom tier matches, but mainly in the 5/10 set. Those numbers essentially speak for themselves in what we had to give up to get this 3/5/7 system. The general player lost almost 30 points of top tier matches and gained about 30 points in bottom tier matches. If the general player base was better off with these changes, then maybe it would be a good tradeoff to have this 3/5/7 template, which leads us to the question: is the general player base better off being a bottom tier tank 57% of the time now as compensation for more bottom tier targets in those matches, and are the weaker players, the real supposed beneficiaries of this policy, even better off now (keeping in mind that being bottom tier was their struggle in the first place)? The general understanding of this issue by weak players can be summed up in a recent quote I have from a response I received on the official game forums. It was written by Pipinghot and he said, "players have been complaining for years that they don't like being bottom tier when there are only 2 or 3 tanks on each team that are bottom tier. Lots of people have been saying for a long time that bottom tier would be more fun if there were more tanks on both teams that are bottom tier, and this would allow all of the bottom tier players to be more relevant to their battles, and be able to contribute more to their teams." I think he did a good job of summarizing the concerns of weak players. While he is correct in understanding why a player would struggle as a group of 2 or 3 bottom tier tanks, he is misguided in thinking that the old system had a great frequency of matches like that and also that the current system is better for weak players. It is at this point that you may say, "now hold on Liberty! Bottom tier players are guaranteed at least 7 tanks and that is much better than being in a pathetic group of 2 or 3. Of course the new system is better for these players!" If those are your thoughts, Pipinghot would support you as he added to his post saying, " Many (many, many) people hated the old system that limited your ability to help your team when you were bottom tier, that problem has been significantly improved now that there are more middle and bottom tier tanks." Let's see if this perception that weak players or players in general are better off now or has WG pulled the wool over our eyes. When I analyzed those 400 pre-9.18 games for information, I also recorded the team breakdowns and my positional tier. This took some time, but I think it was worth it. I wasn't sure what I was going to find either. Perhaps players were getting crushed as bottom tiers before this 3/5/7 template and I was all wrong about this. What did I find in those 400 games taken from tier 5-8 tanks I was in? Well, first I should define what I was looking for. I sought out battles that had 5 or more top tier tanks and 3 or less bottom tier tanks, as defined by Pipinghot. I chose 5 or more top tier tanks since 3 is apparently the current magic number and I figured that 1 more wouldn't be much of a change from 3. With those parameters, out of 400 matches, I found 102 instances of these horrible match ups. 102 out of 400 is an alarming number too, but before we faint, we need to remember our rough frequency breakdown of 35% top, 35% middle, and 30% bottom. Out of that 102 we would probably see only 30% of those matches as a bottom tier. After taking that 30% into account, we get 13 matches. So out of 400 matches, players would get horrible matches about 31 times, which equates to 8% of our matches. Although, after getting these numbers, I realized that the more high tier tanks in a match, the less likely it would be to get stuck as a bottom tier tank. That 30% of bottom tier matches is the wrong percent to use to find out the frequency of horrible match-ups. So now I instead individually counted the actual matches that I was bottom tier in that set of 102. To my own shock, I came up with 6 matches that I was bottom tier out of that 102. So out of 400 games, I was put in a horrible match 6 times (that is 1.5%). Even if I were lucky, the likelihood of someone getting a horrible match would probably still be very low, probably no more than 5% with the worst luck. If you think the criteria I used is too extreme and I should have broaden the numbers, I did. I looked at matches that had 5 or more top tier tanks and now 5 or less bottom tier tanks. This criteria gave me 185 matches out of the 400 and I was bottom tier a total of 29 times in these matches, only slightly over 7% of the time. In general, if I look at the frequency of matches that had 3 or less top tier tanks, it came to 18.5 %. Matches with 4 or 5 top tier tanks was 37.0%. Added together, 55.5% of the old matches had 5 or less top tier tanks, a decent majority. Matches with 6 or 7 top tier tanks had a frequency of 30.75% and matches with 8 or 9 top tier tanks was 10.25%. Matches with 10 or more top tier tanks was at 3.5% (and of that 3.5%, 4 matches, more than a 1/4 of them didn't contain any bottom tier tanks). Now I ask again, are we helping players by making them bottom tier tanks significantly more of the time because they had bad experiences about 2% of the time in the past (or even 7% of the time!)? We must remember when addressing this question that most of the bottom tier matches in the old MM had a good mix of tanks already without a 3/5/7 template. It was only a tiny amount of matches, as defined by Pipinghot, that players struggled with. And if weak players struggled with 4 or 5 top tier tanks in the old system then what difference would the new system make for them? They will actually be struggling more because they see bottom tier matches 57% of the time now instead of about 30%. These weak players and the player base in general would be better off having 35% top tier matches instead of 7% as they can control their fate more in those matches. In top tier matches players have more armor, hit points, firepower, etc. over their opponents. When we strip them of these top tier matches, they are not better off, but worse off. While being bottom tier 57% of the time is bad enough, we also sacrificed some other more subtle characteristics of the game when we moved to this new MM. These tradeoffs may concern players that otherwise don't care about this issue on a top or bottom tier perspective. These other issues are more about aesthetics, but still affect our interaction with the game greatly. One new trend is that the current system becomes tedious after a while. We are constantly thrown into the same match-ups game after game. We get either 3/5/7, 5/10, or all the same tier. I have seen other variations than these only 3 times in over 1000 games, so these 3 are basically all that is expected. In a game that is repetitive already, we don't need mechanics that make it worse. This system encourages a boring environment. The old MM, with its flaws, was able to create over 130 different team lineups in a 3 tier pool of tanks (tier 8, 7, and 6 for example). When I did my research on pre-9.18 games, I identified 114 unique team lineups out of that 200 games. There was a certain beauty in seeing that mosaic of different teams listed across my Excel spreadsheet. There were matches that were 1/8/6, 2/10/2, 5/3/7, 8/1/6, 12/3/0, and everything in-between. After 9.18, in those 400 games, I saw 3 different team lineups over and over. Maybe it is just me, but I find this to be on the lame side. They took a dynamic way of making teams and made it strictly uniform now with no chances of an exciting or odd lineup of tanks. Flat out, this is boring. On top of being dull, the new 3/5/7 template takes an element of chance out of the game too. I find that as much as I dislike random number generation (RNG) consciously, humans have a subconscious affinity for games of chance and keep coming back to them. There is an element of excitement when you get what you want, whether it be a top tier match or you high roll a Skorpion G with an HE round. It keeps us coming back for more despite the disappointments in-between our triumphs as long as the odds aren't that bad. That aspect of the MM is now unfortunately gone. Another issue that we have is also connected with the new monotonous format that we went over above. WG is "dumbing down" the game. In the article on the MM mentioned earlier, there is a line that caught my eye on this topic. They wrote, "It’s much easier now to analyze the power balance and the situation before the countdown is over, meaning there’s time to think over the enemies’ moves and plan your own." I'm not sure about every player out there, but this was not an issue for players that cared about doing better. I never had a big problem with this skill in the past once I consciously tried to master it. I recall watching the old DeathsArrow video series called "WoT I'm Thinking" and listening to him go over the team lineup and that in turn prompted me to do the same thing and improve as a player. When a player decides that they want to get better, they will figure it out without the help of WG. Also, WG should have higher expectations for its customers than what they wrote there. They make it seem like players weren't capable of doing this in the past. And the players that didn't analyze the teams in the past will doubtfully all of a sudden start to do it now. The most helpful tool they could give, and did give to the player base was the list of tank classes at the top of the screen which is not dependent on a 3/5/7 template. These issues in the new system, being mundane, taking away an element of chance from the game, and dumbing it down, are secondary to my main concern about tiering and how that affects the player base, but they are still serious concerns. Many of us have enjoyed playing this game for years and I would hate for us to be pushed away from it because it became boring due to a failed MM change. The new 3/5/7 template isn't helping the vast majority of players in the game and it is leading to frustration and creating a dull playing environment. Many players seem to have been given a false impression by WG and the player base is falling prey to the myths created that support the company line. WG wants the players to think that this new system is a runaway success. Perceptually to some players it is, but in reality It isn't. This system hasn't fixed the problem of team balance that it was stated to have been created for and it also doesn't help bottom tier players. The old system usually had a good spread of tanks in each battle and the few battles that didn't should have been bearable compared with the current 57% of bottom tier matches we now have. The old MM certainly needed to be altered. The old MM needed to be tweaked to make sure that there were the same amount of top, middle, and bottom tier tanks on each side, and also to make sure that each side had a similar amount of tank types (TD, Arty, etc.) per tier, with a +1/-1 differential for variety and to cut down on MM queue time. It also needed a cap on top tier tanks in tier 1 and 2 battles to help protect the new players a little, but beyond that it was pretty good and dynamic compared to what we have now. These changes are simple and would please most of the player base, if not all of it. To conclude, I will leave you with two quotes that are related to this topic. One is from a great player, LemmingRush. He recently stated in one of his videos after he was destroyed at the end of a difficult match, "so when you're bottom tier like this, it's very difficult to actually come out on top, but you just have to understand that, you have to go in into battle with the expectation that you are not going to win and still understand that and be okay with it. Because when you're bottom tier, you know, you're bottom tier." The other quote is from, well, a lesser great player, Claus Kellerman (who essentially represents the average Joe World of Tanker). He said recently when ranting about the old preferential matchmaking premium tanks (IS-6, KV-5, etc) that are getting into too many bottom tier or same tier matches and only seeing tier 8 and 9 tanks now, "shouldn't you be able to make some credits with your IS-6 or KV-5 without being frustrated as f**k, just getting owned by every f**king tank in the game now and having guns that can't do s**t and facing tier 9s and 8s. Like what happened to the preferential matchmaking? What happened to playing against tier 8s and 7s and 6s? I don't know." Constantly being bottom tier is becoming frustrating for many old players and especially our new players. Let's get it back down to 30% bottom tier matches and bring top tier matches back up to 35% so more players can enjoy the game again. ----------------------------------------------------------------------------------------------------------------------------------- * The complete numbers for the 3/5/7 template broke down as follows: Top Tier 3/5/7: 4.25%; Top Tier 5/10: 2.75; Middle Tier: 18.5%; Same Tier: 17.5%; Bottom Tier 5/10: 12.25%; and Bottom Tier 3/5/7: 44.75%. Further Reading: http://forum.wotlabs.net/index.php?/topic/26886-is-something-wrong-with-the-new-match-maker/ http://forum.wotlabs.net/index.php?/topic/27372-hopefully-strict-template-mm-will-die-soon/ http://forum.worldoftanks.com/index.php?/topic/551557-the-great-mm-debate-357-trash-or-triumph/ ----------------------------------------------------------------------------------------------------------------------------------- Appendix: Other charts I later created with the data A more specific chart of player position frequency. Frequency of top tier tanks in a match, but with a players percent of games in those matches from their total matches. These next two are the amount of games in the 400 game sample that the players were bottom tier in a select number group of bottom tier tanks in the match. One chart is the exact number, and the second chart is that percent out of all of a players games. Finis
I am usually better than average in tier 6 tanks so when I struggled to get wins I began to notice that most of my matches were tier 8 and I was constantly becoming powerless to influence games enough to pull out the odd win here or there to pull my win rate above 50%. It was so bad actually that my win rate after 50 games is 42%, 21 wins and 29 losses. I played these 50 games over three days with no other battles in any other tanks. The new tiering possibilities seems to be broken down like this, using the VK3000D as the sample tank and in order of difficulty: Top tier (ex. Tier 6 to 4) Partial Top Tier (Tier 6 to 5) Middle Tier (Tier 7-5) Only Tier (Only tier 6) Partial Bottom Tier (Tier 7-6) Bottom Tier (Tier 8-6) After 50 games, a decent sample size, here are the results (WN8 2435 for these 50 games): Top = 5 (4-1 80%) Partial Top = 0 (NA) Middle = 5 (3-2 60%) Only Tier 6 = 13 (5-8 38%) Partial Bottom = 1 (1-0 100%) Bottom = 26 (8-18 31%) Bottom tier games dominated my play time more than any other bracket combined. If I add together all the other brackets to get a win rate, it is 54%. Which is lower than I hoped it would be, but compared to being bottom tier with a WR of 31%, it is amazing. Is this an isolated incident or have others been experiencing the same thing? I'd personally like to know so that I can save whatever is left of my WR and not play tier 6 anymore until this is addressed by WG.
The_Illusi0nist posted a topic in Casual PurpleWe have all had more than our fair share of suicide teams... you know, matches where the end score is something like 1-15. But what do you think about when you have the team that kills everything before you even get the chance to fire a shot. I find this situation even more frustrating because at least I could do some damage if my team flops on their back like a turtle. So I am going to leave you with two questions: What's you opinion on the matter? How would you make matches more "even"?