Teamfight Tactics Set Nine YouTube Videos Playlist
Another set has come and gone which prompts the traditional retrospective look back at Teamfight Tactics Set Nine. This was a set that did several things well and had another promising beginning that lasted for the first few weeks. In fact, the first half of the set was pretty decent overall as compared to other TFT sets. Unfortunately one of the keystone set mechanics proved to be a disaster in the long run and the second half of the set following the midset update was an utter train wreck. This was the last set to include a midset update and they will not be missed by the community going forward. Once again, we'll take a quick tour of the good, the bad, and the ugly as we review the six months that comprised Set Nine.
Set Nine was advertised as "Runeterra Reforged", a reworking of the original Set One that simplified some of the unit and trait designs to make the gameplay easier for newcomers to pick up. Due to the geographic theming of the set, most of the units were given traits that corresponded to their lore from League of Legends. Garen had Demacia trait, for example, while Darius had Noxus trait and Teemo had the Yordle trait. Units were given trait pairings that logically fit their character backgrounds, like Jayce having Piltover + Gunner traits or Cho'gath having Void + Bruiser. Every single unit in the set also used their base artwork from League of Legends rather than something bizarre from a skin line. After having so many ridiculous outfits and weird trait pairing from previous sets, especially the crazy stuff like Heart and Hacker and Ox Force from Set Eight, this was a real breath of fresh air.
I want to dedicate a second paragraph to this topic because I think it was that important: using all of the base League artwork and lore-based trait pairings was the single best aspect of Set Nine. It was much, much easier to learn all of the various units and their traits in Set Nine because they were natural matches for the units in question. Lux wasn't a Redeemed unit or an Astral unit or whatever, she was a Demacia unit because OF COURSE she was. Ashe, Lissandra, and Sejuani were all Freljord units since they were the "cold" characters, Renekton and Azir and K'Sante all had Shurima trait because they were the "desert" characters, and so on. The visual clarity for this set was infinitely better than in Set Eight which made learning the set and purchasing units quickly in the shop far easier. And yes, this really does make a difference in terms of attracting players and keeping them around for more than a few games. I loved this aspect of Set Nine and wish that the base skins would be used all the time in Teamfight Tactics, or at least more often than once every 2-3 years. It's not a coincidence that the sets featuring base skins have been some of the most popular ones historically.
One of the big innovations for Set Nine that was also received positively was the introduction of Portals. These were effectively a new version of the Galaxies from Set Three, a new set of rules applied specifically for the game in question, with the big improvement of having multiple different portal options that the lobby would vote on prior to the game starting. Some of the portals were much more popular than others and were chosen over and over again by players while others were virtually guaranteed to see you standing there by yourself if they happened to appear. Most of the portals injected additional loot of some kind into the gameplay, either changing some or all of the augments to Prismatic status or handing out free gold or free shop rerolls or free champion duplicators or free emblems, that sort of thing. My favorite portals were the ones that made modest changes to the gameplay without tossing out a bunch of free stuff, things like God-Willow's Grove that allowed players to gain the traits of a unit on the bench or Immortal Bastion which granted a flag item which units could cluster around to gain additional attack speed and shielding.
Unfortunately most of the portal options were not equally interesting and simply handed out lots of freebies. Even more unfortunately, the portals that gave players the most free stuff also tended to be the most popular options; whenever Jayce's Workshop appeared (all augments will be Prismatic) you just knew that five or six of the other players would run there immediately. I've criticized the TFT development team in the past for putting so many free handouts into the gameplay as it's detrimental to the core TFT mechanics. However, it's clear that this is what a large portion of the playerbase craves so the developers are simply giving people what they want, for better or for worse. Most of the portals were innocuous enough that I didn't mind them too much even though I did think that there were far too many Prismatic augments and trait emblems getting handed out across this set. The portals were a big enough hit that they are apparently going to stay for good, permanently replacing the traditional item carousel at the beginning of each match. Even though the opening carousel had real issues (like five people all clicking for the same starting item), I liked it overall and I'm sorry to see it disappear. Now players have absolutely no control at all over their starting items which is a major loss of agency.
The trait design was a bit on the simplistic side for Set Nine, at least in part due to this notion of getting back to basics. Many of the traits were pretty easy to understand, like Bruisers gaining health and Multicasters getting additional ability casts and Gunners gaining stacking attack damage, and I don't think that this was necessarily a bad thing. Not every trait has to be super complicated for the gameplay to be fun and interesting, even if Set Nine did stray a bit too much into retread territory at times. Traits that stood out as being new and interesting included the Strategist trait which granted team-wide shielding to the front lines and ability power to the back lines, Demacia trait giving out a limited number of Radiant items, Zaun similarly providing its own unique Chemtech modifications, Ionia giving out Enlightened bonuses that accentuated what its units were already good at doing, and the Shadow Isles trait having a really oddball shielding + mana generation combination unlike anything we'd seen before. All of these created new aspects of TFT gameplay for players to experiment with and (mostly) worked well during the first half of the set.
Other aspects of the trait design didn't work out quite as well. Juggernauts never really emerged as an additional viable frontline option, leaving the set with just Bruiser and Bastion units which limited the options of how teams could be put together. Rogue trait was supposed to be a replacement for the old Assassin trait, however it was a buggy mess for most of the set and eventually had to be nerfed into irrelevance because when it did work properly there was almost no defense possible on the part of the other players. (How can you protect your carries when these pseudo-Assassins jump into the backline at some random point during the middle of the fight?) The Freljord trait's armor and magic resistance shred was a little bit too good throughout Set Nine while the Yordle trait proved to be too confusing and difficult to use for most players. Only dedicated rerollers could make use of this trait and I don't think anyone ever played Yordle 5 across the entirety of the set.
I also need to single out the Piltover trait as the worst-designed trait of the whole set. This was a really cool concept in theory: Set Nine's version of an economic trait which would stack up a mechanical T-Rex which could be sold for additional loot. In practice the trait was a disaster, with no clear indication for players as far as when to sell T-Hex machine or what they would get for doing so. After a few patches, the developers simply removed the option to sell the dinosaur completely because it was so confusing. Even leaving this aside, the huge problem with Piltover trait was the fact that it was utterly useless if the player didn't go on a massive loss streak. The trait granted a mere 1 gold per round (and no dinosaur stacks) when win streaking, and the T-Hex gained almost no stacks at all for short loss streaks. No, the only way to play Piltover was to get the trait in place by Stage 2-1 and then lose the next 8-10 rounds before cashing out, which would then make the individual in question invincible to the rest of the lobby. Other economic traits have had the same issues in past sets, but Piltover could ONLY be played in this way with no other alternatives possible. Coming on the heels of the excellent Underground trait from Set Eight, this was a gigantic miss that really hurt the set's long term viability.
The other big addition to Set Nine aside from the Portals mechanic was the introduction of Legends, and oh boy, was this ever a gigantic mistake. There's a consensus in the competitive TFT community that Legends were not only a bad addition to the gameplay, but were in fact the worst TFT mechanic since the infamous Shadow Items way back in Set Five. Backing up a bit, Legends were a new mechanic that allowed players to select a League of Legends character from a list of about a dozen choices before queuing up for a TFT game. Each Legend had a predetermined list of augments which would always appear when choosing augments on Stage 2-1, Stage 3-2, and Stage 4-2. These were sorted by the Silver/Gold/Prismatic tiers depending on what type of augment was available, but choosing the Tahm Kench Legend would always give the player the option of Rich Get Richer augment if the first augment happened to be Gold. The Legends only affected one of the three available augments, each of which could be rerolled, therefore theoretically occupying only one of the six possible augment slots. However, the benefits of knowing ahead of time at least one augment which was guaranteed to be offered had an enormous effect on the gameplay in a predictably negative fashion.
The TFT developers advertised Legends as the chance to provide some additional customization to players and cater the augments that they would see to their personal playstyle. And perhaps this even would have been the case in some magical dreamland where all of the augments were perfectly balanced. In the real world, however, TFT's augment balance has ranged from mediocre to awful ever since they were introduced back in Set Six. Most competitive players were highly skeptical that the Legends would be sufficiently well balanced and suspected that everyone would wind up running the same one or two choices in every game once the strongest options were determined. This is exactly what happened, of course, which was easily forseen ahead of time and had awful consequences for the gameplay. Remember that Teamfight Tactics is a hotly contested Player vs Player game. This isn't like a Single Player title where you can go off and run variants when things are out of balance. Basically you have to play the strong options yourself or you get crushed by those who do. Thus everyone wound up running the same Legends over and over again depending on whatever was overpowered in each particular patch.
This foolish decision sucked out all of the creativity and variability that made the augment mechanic such a success in the first place. The whole reason why augments worked as a design concept was due to the way that players would see different options from game to game. Players had to think on their feet and adapt from game to game depending on what units, items, and augments they were seeing in each outing. This is the core of TFT's gameplay and it's what keeps things feeling fresh. But Legends had the exact opposite effect: now players could force the identical augments in every single game, and in fact most of the time they *HAD* to do so or they would be defeated by everyone else running the overpowered options. Team compositions had to be optimized by picking the proper Legend ahead of time before the game even started; good luck trying to beat another reroll player if he had the Lee Sin Legend and had all the reroll augments guaranteed to appear while you're trying to play flexibly. Legend choices also dictated which portals each player would pick because some of them had innate synergy with the augments that were guaranteed to show up later. It quickly became clear that Aurelion Sol players always wanted the first augment to be Prismatic while Ezreal players would take Ecliptic Vaults (combining item components grants additional gold) and Vladimir players would take the portal that gave everyone 115 HP instead of the default 100 HP. All of this was terrible for flexible play - it heavily advantaged players engaging in rote memorization and doing the same thing in every single game.
Several different Legends were overpowered at different points in time throughout Set Nine but the overall problem never went away since even as little as a 5% perceived advantage for one Legend would result in 80% of the playerbase picking it in every game. Ezreal was the overpowered Legend early in the set, granting 3 extra gold plus 3 extra item components at the start of the game. There was almost no way for other augment options to keep pace with the tempo advantage that this offered throughout the early game and thus there would be six players running Ezreal in every lobby. There was a hilarious period of 48 hours where the Draven Legend was turbo-broken and had to be hotfixed because Draven's Spoils of War augment was dropping insane amounts of loot and turning the economic side of the game into an utter joke. Ornn was the most popular Legend choice for a long period of time and probably distorted gameplay the least, although the Ornn anvil lottery that followed caused its own problems (pulling Eternal Winter was a massive advantage). There was also a briefer period where Master Yi's Pump It Up augment was used to accelerate Garen-based team compositions. The balancing problems were obvious for the developers in situations like this: Garen was pretty underpowered normally but could turn into a monster when he had 50% more attack speed from the Master Yi Legend. And players could force the problematic augment in every game if they wanted... which would also require them to them hard-force Demacia as a comp to match their Legend choice - ugh.
Everything about this system was bad as its flawed elements cascaded into other TFT mechanics and started breaking them as well. Aurelion Sol's bonus experience forced the TFT developers to increase the XP requirements to hit Levels 6, 7, 8, and 9. This caused a mass rolldown for 4 cost units on Level 7 throughout the set as it was too expensive to reach Level 8 barring a crazy econ opening. Twisted Fate's ability to force perfect items resulted in TFT's aura items being nerfed and then eventually removed from the game entirely since forcing three or four Lockets was so broken for Bastion front lines. I hate that some of my favorite items like Zephyrs and Shrouds had to die for the sins of the Legend system. The popularity of Urf's emblems out of Tome of Traits caused massive problems at the end of the set as it became way too easy to hit chase traits like Demacia 9 or Noxus 9. During the Set Nine World Championships, Urf was the strong option and the whole contest turned into an emblem-pulling lottery. This led to incredible data like the following stats from the first day of worlds:
At the absolute highest tier of play possible, roughly 85% of the games were played using the same Legend - and the few individuals who didn't take Urf as their Legend actually performed worse in terms of average finish! There was no benefit here from eschewing the most popular option and heading in a different route, it genuinely was the best choice to follow the crowd and take exactly the same augments as everyone else. Needless to say, this didn't make for games that were much fun to play or games that were much fun to watch.
The worst of the Legends was Twisted Fate, not even because it was overpowered for most of the set but because of the way that it warped the gameplay. Twisted Fate as a Legend allowed players to guarantee that Pandora's Items would be offered as the first augment, which would shuffle any item components on the bench at the start of each round. I was never a huge fan of Pandora's Items as an augment as it encourages players to greed for "perfect" items instead of making due with what they've been getting from their random drops; I've tended to take this augment when getting particularly ill-suited items for my team and otherwise usually ignore it. However, when players could know ahead of time that they would be able to take Pandora's Items in every single game, and have it operating the whole game instead of getting it for the second or third augment, it distorted the way that they played. There were a handful of champions in this set (Nilah and Mordekaiser being the worst culprits) who were ordinary champions with normal itemization and suddenly became overpowered with the perfect best-in-slot items. Twisted Fate allowed players to force teams that otherwise wouldn't be viable because they could reroll items until having four or five bows, something which would otherwise never happen. This wrecked game balance further as unintended item interactions which would otherwise happen once in a handful of games were now being forced by multiple players in every lobby.
Backing up and looking at this from a big picture perspective, Legends were toxic to TFT gameplay because they undermined everything that gives the game its longevity. TFT is always at its best when players are trying new things in each game and playing flexibly around what they find in the shop; it's at its worst when players are hard-forcing the same thing over and over again by rote. Legends essentially forced players into the same repetitive patterns which became stale very fast. Twisted Fate was especially bad in this regard because players could literally sit there with a guide and run the exact same items on the exact same champs with the same augments in every game - no thinking required. Supposedly the Legends were very popular in the more casual portions of the TFT playerbase for exactly this reason, however I think this is one of those situations where it's important to watch what players do and not what they say. Not having to think much and using Twisted Fate to run the same gameplay back over and over again, well, it might sound good to more casual TFT players... but my impression is that this became very boring very fast for pretty much everyone. Very few people can be antisocialmunky and run Kayle reroll in 200 consecutive ranked games, they generally get tired after a dozen or so games and go off to do something else. Don't watch people telling you that Legends are a fun mechanic, watch what they're actually doing, and the player retention rate for Set Nine was quite poor indeed.
As if the Legends themselves weren't bad enough, the TFT developers decided to embarass themselves further by initiating a new policy of hiding all augment data from third party websites. It was blatantly obvious that they didn't think that they could properly balance the various augments and thought to put a Band-Aid on the problem by simply hiding which ones were performing poorly from the playerbase. This would be like Major League Baseball responding to the discovery that on base percentage (OBP) is a much more effective analytic tool than the old runs batted in (RBIs) metric by conspiring to hide OBP data for the whole league. Much like this made-up example would have been, the attempt to hide augment data was an abject failure as players were well aware of which Legends were overpowered and third party websites found ways to get the hidden augment data anyway. The TFT developers made fools of themselves by saying things like "just play 2-3 games with each augment rather than relying on data", leaving aside the fact that there were 200+ augments and a third of them had their numbers changed in every biweekly patch. Sure, I'll get right on that, I only need to play 500 TFT games to get a feel for each augment, and I'd better do it before the next patch comes out with its own hundreds of changes. This was not a serious position and fortunately the TFT developers relented from this policy after receiving overwhelming negative feedback. The fact that they attempted it at all was a sign of how poorly the Legends were designed as a mechanic. It's honestly a real shame: if Set Nine had left things alone with Portals as the only set mechanic, it would have been a much better set!
Ever since Set Three, TFT has split its sets in half with a midset update that runs for an additional three months after the initial three month period expires. These midset updates exist for exactly one reason, to sell another Battle Pass, and most of them have been significant downgrades over the first half of their respective sets. Only Sets 5.5 and 7.5 were improvements, the first case because Shadow Items were removed and the second case because Set 7.5 did away with the stupid restriction of only having one Dragon unit allowed on each team. Set 9.5 absolutely did not break this pattern as it proved to be a massive drop in quality as compared to the first half of the set, bad enough that I barely played it after the initial week or two. Set 9.5 had all of the same problems common to midset updates, swapping out units and traits to make changes for the sake of making changes, and replacing them with hastily designed new stuff that didn't fit together properly. To run through a few quick examples:
* Noxus was designed as a tempo trait, intended to win streak early and put pressure on the rest of the lobby but then fall off in the lategame (due to lacking a 4 cost carry unit) unless Darius and Katarina could be 3 starred. This identity was completely destroyed in Set 9.5 as Darius and Katarina were nerfed into becoming trait bots while Mordekaiser was added at the 4 cost tier. After the midset update, Noxus was played as either a reroll team based around Samira/Cassiopeia or a Mordekaiser-based lategame comp which felt nothing like its previous incarnation and didn't fit the thematic nature of Noxus.
* Challenger as a trait desperately needed Kalista to cut through tanky frontlines and put early pressure on the lobby before the 4 cost carries of Yasuo and Kai'sa arrived. After Kalista and Yasuo were removed in the midset update, the trait didn't have any units to pressure the lobby in the early game except Samira who was better played in a Noxus comp anyway. The trait became centered on Fiora who was a poorly designed unit with invulnerability on her ult and spent the whole set either unkillably overpowered or laughably weak. Challengers simply felt "wrong" for the whole second half of the set as they clearly missed the hole left behind by Kalista.
* Invoker was a popular trait throughout Set Nine and an additional option for AP builds if an early Shen or Ryze showed up. The trait became unplayable in Set 9.5 once Lissandra was taken out and it never recovered, falling out of the meta without ever being replaced.
* Deadeye trait was removed entirely and replaced with Vanquishers which was a much less interesting and less flexible design; Vanquishers basically had to be played with vertical Ionia and were useless in any other context. Aphelios being shifted from Deadeye to Gunner was a change that also failed to work as he was never played outside of Piltover cashout boards.
* Zaun was a trait based around its 4 cost carries, Urgot and Zeri, and therefore its Chemtech mod items were also designed around these units. All of them were either intended for tanky melee DPS (Urgot) or were attack speed based (Zeri). However, both of these units were removed in the midset update and replaced with Silco, a slow-acting AP caster who had no use for any of these stats. None of the Chemtech items were ever updated which left the Zaun trait in a permanently broken state, with reroll Jinx and Ekko doing a poor job of filling in for their more expensive departed predecessors.
I could continue but that should be more than enough examples. Again, this was typical from past midset updates as the changes from the second half of the set were awkwardly grafted onto systems which had been designed with other purposes in mind. As bad as this was, Set 9.5 was made far worse by having a truly heinous amount of bugs and poor balancing decisions, with one patch after another forced to release immediate hotfixes because so many things were outright broken. The Bilgewater trait was ludicrously overpowered when the midset update first released (despite everyone telling the developers on the PBE that this was the case), then there was a patch where Mordekaiser was severely overtuned, followed by a patch where untargetable Fiora was crushing every lobby, and so on. There was almost always a single unit or trait which crowded out every other option and which was made more forceable than ever due to the Legends operating in the background. Towards the end of the set, Multicasters were inexplicably reworked for no reason and buffed into absurdly overpowered state, with the trait averaging something like a 3.60 average placement. Then it was hotfixed nerfed and it *STILL* had something like a 4.10 average placement afterwards. This was the patch on which qualifying for Set Nine Worlds took place and it was a total crapshoot as far as who could hit the overpowered Multicasters team first when everyone rolled down at Level 6 on Stage 3-2. The balancing was disgracefully bad throughout this patch and the midsetup update more generally.
It was hard to explain how the gameplay balancing could continue to exist in such a poor state after four years and eight previous sets of Teamfight Tactics. The developers seemed to make the same mistakes over and over again, engaging in the classic balance thrashing that the community has come to hate, overbuffing and overnerfing units in a merry-go-round that created a new overpowered team composition with each new patch. I've written in previous retrospectives how the designers aren't stupid and that they're clearly engaging in deliberate balance thrashing because they think it drives player engagement. But then I saw this post on Reddit TFT in response to a completely predictable triple set of buffs to Taric that made the unit stupidly overpowered and led to an immediate hotfix nerf. What if there's a darker explanation at work here? Here's the post in question:
Like I said, I had already come to the conclusion that the TFT developers are not actually trying to balance the gameplay, that was obvious after several years of playing this game. Keeping the gameplay in a state of constant flux clearly had higher priority than maintaining proper balance; that's why every patch has 100 unnecessary changes in it to keep shuffling things around endlessly. But what I hadn't considered was that the TFT developers might be *DELIBERATELY* breaking their own game by releasing patches that they know will require immediate hotfixes. There's an evil logic at work here: everyone discusses how dumb the proposed changes are ahead of time, everyone has to see the ridiculous nonsense for themselves after the patch releases, everyone has to discuss potential fixes and complain on social media afterwards. Engagement, engagement, engagement. Then the TFT developers can swoop in afterwards with the hotfix patch that they already knew would be needed ahead of time and look like the responsible good guys to get praise from their most devoted players. I'd like to think that's an insane conspiracy theory but... it honestly tracks really well with how TFT's patching history has proceeded. I certainly hope it's not true because man, that would be absolutely ghoulish and messed up.
Anyway, I could write and write and write on this topic but it all boils down to one thing: Set 9.5 wasn't fun to play. It grew stale incredibly fast between its poor balancing, buggy patches, illogical trait design, and the ongoing plague caused by the Legends. I don't think I even made it to two dozen games before I was ready to abandon the set and wait until the arrival of Set Ten. I was hardly the only one either, as the first half of Set Nine performed similar to Set Six (significantly better than Sets Seven and Eight) only to collapse back down to their level after the midset update. The aggregate data matched the experience of our local TFT group as pretty much everyone dropped the set after a couple of weeks due to lack of interest. It was obvious that Set 9.5 received barely any developer attention whatsoever while most of the Riot team was working on Set Ten. From that perspective, Set Nine looks a bit less like a "return to basics" set and more of a case of "reuse lots of old assets so that the set can be done on the cheap".
If there's good news here, it's that Set Nine will be the final set to have a midset update at all. The TFT developers are switching over to three sets of four months per year instead of two sets with two midset updates each lasting three months. This should be a major improvement and it would have happened much sooner if Riot hadn't been obsessed with selling additional Battle Passes. (I thought they would have to give up some of that income stream but nope! They're going to keep selling two Battle Passes per set even though the full sets will now be four months instead of six months, plus release more and more exclusive time-limited chibis at outrageous prices. I genuinely can't discuss the monetization of TFT without wanting to vomit.) The midset updates are going the way of the dinosaurs and that decision has been met with much rejoicing. Their removal going forward is the single best thing that can be said about Set 9.5 - good riddance.
Overall then, the first half of Set Nine was pretty decent before the second half turned into absolute garbage. As always, the TFT developers were their own worst enemies as they tried to do too many ambitious things at once without having the proper resources in place. Legends were a misconceived idea from the start that undercut the core mechanics of TFT and in the end they wrecked what could have been a promising set. Set Ten is bringing back the Chosen mechanic from Set Four in a revamped form, and while that's an exciting idea that should have been done much sooner, I have little doubt that the developers will try to do 25 things at once again and find some way to screw things up. At least we have the winter holidays to look forward to: about a month when the TFT team is locked out of the building and physically can't patch the game no matter how much they want to do so. It's my favorite period of the year for TFT and speaks volumes about how the game plays best when the developers aren't allowed to make changes. Until then, thanks for reading - this set will not be missed.
