What we learned from Gfinity 3

I wrote an article titled Why G3 trumps ESL One Cologne a week or so ago, and in it outlined some of the most interesting storylines that we're looking to figure out at Gfinity 3.

While some answers remain unclear, we've also learned a ton about the current state of the professional CS:GO scene. In turn, that will make ESL One Cologne all the more exciting.

Let's look at the ten most important lessons we took home from two days packed with action at Gfinity 3, which was ultimately won by Virtus.pro over Titan in the grand final.

Sum up of LC, mouz, Epsilon & Heat

Going into Gfinity 3 nearly everyone agreed these four would be the teams battling for the final playoff spots in the stacked groups of the event. They did not disappoint, though group A wound up being so close that a must-win for London Conspiracy actually skyrocketed them all the way to a second place, ahead of both iBUYPOWER and NiP, as well as Epsilon, who were wrongly knocked out of the tournament.

While Epsilon being wronged has garnered reasonable media attention, almost no one has spoken about Heat. Mikail "⁠Maikelele⁠" Bill's team ESG attended G3 under the banners of Heat eSports, and did just what was expected of them - put up some good fights versus the stronger teams, and beat both mousesports and FM - but were still knocked out. G3 messed up here -- you can't make a case against it. They used the wrong rules, and it knocked both Epsilon and Heat out, while saving NiP and mousesports.

London Conspiracy looked the best out of all these four teams. They beat Epsilon, iBUYPOWER and Infused, while also putting up a fight with NiP. They also took a map off of Virtus.pro in the quarter-finals. Though Håvard "⁠rain⁠" Nygaard cooled down on day two, he had a strong event and is going to be one of the players to look out for in Cologne as LC look to get out of the group stage once again, against all odds. Since Preben "⁠prb⁠" Gammelsæter joined the team has only improved, and they are a solid top fifteen squad now.

mousesports are not as good as a team, but they have a much better top player in Aleksi "⁠allu⁠" Jalli. A huge part of their issues are surely the language barriers in the team, and those will get better over time. Before, we expected allu to be flanked by Chris "⁠chrisJ⁠" de Jong on the scoreboard, but the Dutchman was a huge disappointment in London, finishing with team and career-low 0.77 rating for a LAN tournament. Nevertheless, taking fnatic to overtime on the third map means they, much like LC, are also a top fifteen team.

Epsilon are the one who showed the most promise. Joey "⁠fxy0⁠" Schlosser proved he is indeed a great sniper even on LAN - finishing with sixth highest rating of 1.24 and a 0.89 KPR - while helping his team overcome NiP in the opening round. Key mistakes in LC and iBUYPOWER games cost them a spot in the playoffs - though they should have been in regardless, if not for G3 admins' mess-up - but they showed a lot more promise than before. Add in the fact Richard "⁠shox⁠" Papillon was visibly ill, and had one of the worst showings of his career, and it's clear Epsilon will be dangerous in Cologne.

Finally we have Maikelele's Heat. They were somewhat boring in the sense that only the star player Maikelele had a good showing individually, and they did exactly what was expected of them. It doesn't yet seem like they are truly ready to compete with the top eight teams, not even with Marcus "⁠Delpan⁠" Larsson on the team, but they now have three-plus weeks of preparation for SLTV StarSeries X Finals, where things may be different. Good team, but not a contender.

New fnatic is good, but not great

This fnatic roster is already doing better than the previous one ever did in 2014. A 3-4th place finish is solid, though it comes at an event that fnatic seemed better prepared for than most, and their players certainly had to be more motivated for than the average top team. Due to G3 being the roster's debut, there was added pressure. They lived up to it, but the going will only get tougher in the future.

fnatic finished group A with wins over NiP, London Conspiracy, Infused and Epsilon while surrendering a tie against iBUYPOWER. That's a good record. In the playoffs they struggled too much with mousesports - going into overtime on the third map before allu's team ran out of steam - and were then knocked out by Titan, a team that has always been kryptonite for Markus "⁠pronax⁠" Wallsten's troops.

The Swedes had changed some things around, but in general it was clear we saw another pronax team on the server. That is both good, and bad. It's good because he is one of the better leaders in the game and a group of players led by him, versus not, will almost always be better. At the same time it's bad because that means they haven't changed enough to be surprising in any way -- everyone knows just how they play.

fnatic's win at DreamHack Winter last year was largely accounted to no one being accustomed to their playing style. If you ask almost any team in attendance in Jönköping last November, they will tell you fnatic became easy to counter once you knew they were going to play that aggressive style. Now they have more skill - which will make it work better - but it's still stoppable, as seen in the Titan series.

Olof "⁠olofmeister⁠" Kajbjer struggled at G3 individually, and looking at his stats page, you can tell he hasn't adjusted well to playing in fnatic's system. After averaging close to 0.80 kills per round and 1.12 in rating, he has yet to go above 0.70 for an event in fnatic, and he's had a sub-1.02 rating in every online tournament, and now at G3, since joining the Black and Orange. If pronax can get him more involved, fnatic will become more dangerous. For now, they're a marginally upgraded version of the previous team.

Poor organizing still happens in 2014

Though I stand by my opinion in my article comparing G3 to ESL One for spectators - the arguments still hold true, though execution in running of the event ruined some of it - it's clear the people in charge of G3 made multiple major mistakes of the magnitude that should not be happening in 2014. Luckily the event as a whole wound up being a success, and most of these issues will be forgotten in a week, but we must make sure this doesn't get repeated.

G3 greedily wanted streamers to pay for rights to do so, but not enough of them did so, which led to some games not being broadcasted at all. Luckily they didn't go through with it in the end though. Once contacted by HLTV.org to get at least public GOTVs added, the organizers simply stated we can't do so, as it would take away viewers from their official stream - even though we were talking about different games. I understand it's a business, but this kind of greed serves no one.

We missed games such as Epsilon's upset win over NiP - which no demo, VOD or anything else exists for - and iBUYPOWER's tie against fnatic. Instead we were forcefed lopsided games featuring weak UK teams. It's unacceptable, and can not fly in the future. There were also major issues with the PCs USB ports - leading to tons of delays - that probably should have been taken care of a day or two before the action was set to begin.

Other problems included random lagouts for some seconds that affected rounds' outcomes in multiple games, and the water dripping down on players from the ceiling,IEM Global Challenge New York-style. However, none played a bigger factor in how the event itself played out - the tournament integrity - and were easier to fix than the tiebreaker in the group stage.

Finally, we also can't forget that the organizers kicked out a majority of the paying customers - their spectators - during the grand final. Plenty of people had bought the expensive tickets to watch some of the world's best players play - and all of them probably would have wanted to see the grand final - but G3 decided they couldn't stay, as the venue was past its closing hour. At the very least those poor folks should be refunded, but I doubt it will happen.

It's down right appalling that the organizers of G3 failed to use the correct tiebreaker method in a round robin group stage. In Counter-Strike, when two teams are tied, the tiebreaker has always been head-to-head. That is because maps vary in games, and starting on the easier side will help your round difference. Both groups saw wrong teams advance due to G3's idiocy, and that truly is one of the worst things to happen in professional CS:GO up until 2014.

iBUYPOWER blew their best chance

It's unclear what the players themselves expected exactly, but in our AMA with the team, it seemed Joshua "⁠steel⁠" Nissan and company believed they'd have a solid chance at making the semi-finals. Instead they lost in the quarter-finals, which seems like a fair assessment of their current form. iBUYPOWER are a top team - in the region of the eight best or so - but not exactly a contender - not yet anyway. In addition, they struggled mightily in close games.

iBUYPOWER now get to bootcamp in London for the next ten days to get ready for ESL One Cologne and to better adjust to the playing styles of the European teams. That should increase their level of play significantly, though in comparison to the teams they will be going toe-to-toe with at ESL One, it's unclear how much of an impact this bootcamp will have. Their opponents will also bootcamp, and I'd argue iBP were the better prepared team this time around, aside from Titan.

While most European teams - excluding Titan, and teams who won't be at ESL One, such as mouz and Heat - weren't properly prepared, having just come off of vacations, iBUYPOWER had certainly put in whatever effort possible before this event. The stakes were higher for them, and although the preparation they can get in North America doesn't compare to their practice in London in the next ten days, adjusting for their opponents, they were probably still better off here.

Teams such as NiP, Virtus.pro and dignitas attended G3 after breaks from the game and without a bootcamp. Not only will they put in maximum effort for ESL One, but so will teams such as Natus Vincere, HellRaisers and LDLC. The competition will only get tougher from here on out, so I would actually make the case that if iBUYPOWER were going to make the semi-finals during their trip to Europe, it would have happened in London. In Cologne they aren't favored to even advance from their tough group.

With that being said, it's entirely possible Sam "⁠DaZeD⁠" Marine gets his team going in the next weeks, they overcome whatever issues they had in closing out games in London, and have a little bit of luck on their side in the best-of-one battles potentially versus fnatic - who finished 3-4th at G3 - and Virtus.pro, the champions. Odds are currently against iBUYPOWER, who also may have to face dAT in group C, but they are within an arm's reach, and maybe even closer. The window remains open, though pressure is building.

allu broke through as a star

Though mousesports as a whole were underwhelming on an individual level - especially true for chrisJ who was appalling - they had a clear breakout performance from allu, who carried his team to a 5-8th place finish, and damn near over fnatic in the quarter-finals. His 35/5/27 statline in the overtime loss proved he was by far the best player on the server, but reminds everyone that Counter-Strike is a team game. You win as a team, and you lose as a team.

allu had a great run overall, averaging out to 1.29 rating - which would be good for number two in the world over a career - and finishing the tournament with a 0.90 KPR and a kill-death difference of +53. Most importantly, his rating was a ridiculous 0.30 better than his next teammate's. That's the kind of carrying that would have gone down in the history books had mouz won the overtime versus fnatic and clinched a semi-final spot.

For comparison, a 0.30-difference is more than the difference between Patrik "⁠f0rest⁠" Lindberg and Robin "⁠Fifflaren⁠" Johansson. Only one player in mousesports - aside from allu - had a rating even close to or better than their career average, and that was Nikola "⁠LEGIJA⁠" Ninić who topped his career rating of 0.98 by 0.01. In other words, allu is the only reason mousesports beat Virtus.pro, put up a fight versus dignitas, advanced to the playoffs and nearly knocked fnatic out of G3. Not too bad for a debut performance.

Virtus.pro and dignitas still got it

Most people expected these two juggernauts to be good, and even I predicted them both to finish in the top three of group B, but it's still nice to get confirmation that after some time off - in dignitas' case - and some lackluster results - for the Polish side - they can still compete for top placings amidst the world's best teams. Let's look at the runs of both of these teams individually below, starting with the lower placing of them, the Danes of dignitas.

dignitas came into G3 as a huge question mark, not having played an official match since the end of June, not having attended an international event since mid-April, and having only played twenty-or-so matches with Philip "⁠aizy⁠" Aistrup on the roster. To make it worse, they'd been on a lengthy vacation from the game. It was impossible to know whether the team who'd just placed 3-4th at the last two big tournaments, including the previous major, would show up or if this dignitas team would regress slightly.

In reality dignitas looked near impossible to beat until the final two maps - their only losses in the entire event - put them out. They gave single digits of rounds to Titan, Virtus.pro and ESG in their wins, and had a dominant first half versus mousesports in a close game. They wrecked their long-term nightmare match-up NiP, only losing fifteen rounds in two maps, and gave Virtus.pro seven rounds on the harder, defensive side of de_dust2 in the semi-finals.

If this dignitas team truly wasn't playing at 100% of their capabilities, then we're in for a treat in Cologne. All of a sudden they no longer have any match-up they can't overcome, which opens up a huge possibility for dignitas to make a deep playoff run. They're one of the most consistent teams around and have lots of individual skill, and seem stronger than they were at their last two semi-final exits with René "⁠cajunb⁠" Borg on the team. Future looks bright for Denmark's finest.

As for Virtus.pro, this win came at a good time. Supposedly they reverted back to Wiktor "⁠TaZ⁠" Wojtas leading the squad, and although Filip "⁠NEO⁠" Kubski still struggled individually, the rest of the team did great. Paweł "⁠byali⁠" Bieliński had a 26/11 performance for a 1.80 rating in the grand final, TaZ returned to his career averages, and Jarosław "⁠pashaBiceps⁠" Jarząbkowski finally played at the consistent all-star level we grew accustomed to earlier in the year.

For Virtus it seems a different player peaks in every different match, but this time a new player emerged as their top dog at this particular event. Janusz "⁠Snax⁠" Pogorzelski had a great showing, finishing with a 1.24 rating and a 0.80 KPR, while having a dominant performance in the semi-final against dignitas. The Poles messed up a lead versus Titan, lost a game versus mouz and dropped the first map to LC, but every time they replied and stood on the highest podium on Sunday night.

That was all on very little preparation - they only started practicing a week before the event. Virtus did not bootcamp before G3, nor was it a huge priority for them. It all shifts to fifth gear now as the Poles set up their bootcamp in hopes of defending their championship from EMS One Katowice on August 14-17. Virtus are likely favorites heading into ESL One - seeing as NiP failed so miserably - and it must feel great. Now let's see if they become the clear number one in the world in Cologne.

Titan return as a contender

They are back. For now it's impossible to know whether it's because they were the only team to bootcamp - to our knowledge - before G3, or if it's long term change, but it doesn't matter. Titan have always been capable of beating everyone, except NiP, and now they've done both. Beating NiP online last fall opened the flood gates to VeryGames becoming number one in the world, and it may well do it again in two weeks.

The French-Belgians had a rough start to G3 and must have been doubting themselves as they were facing a 1-2 start in group A, following a lopsided loss to dignitas and an easy win versus FM, as they were trailing Virtus.pro 11-15 on de_mirage. They forced a tie from the match, and went on to beat both mousesports and ESG to advance to the playoffs. 

In the playoffs Titan overcame a 7-14 deficit against iBUYPOWER on de_dust2 - showing mental strength, again - and went on to win the series with a convincing 16-8 win on de_inferno. Semi-finals saw them beat fnatic 2-1, in an up-and-down series, but pull through in the end nonetheless. Only in the grand final did they find themselves outmatched, but Virtus was on a roll then that wasn't going to be stopped easily.

Most important for Titan isn't this single top two finish, or the $10,000 check they picked up. It's the added confidence after the NiP win online that what they are doing is working, and it will build in the days leading up to ESL One Cologne. Titan are back in the mix for a potential grand final appearance, and it's the redemption their players likely deserved - and Kévin "⁠Ex6TenZ⁠" Droolans is also happy.

Obviously Kenny "⁠kennyS⁠" Schrub was the star of Titan in London, finishing with a 1.31 rating in twelve maps - a huge performance, all things considered probably the best of his career - but both Nathan "⁠NBK-⁠" Schmitt and Adil "⁠ScreaM⁠" Benrlitom also played really well, around the level of their career averages. Edouard "⁠SmithZz⁠" Dubourdeaux still seems like he's a man without a role in Titan, but he will be more content that way when he's playing in the finals again.

G3 format makes for a great viewing experience

This wasn't a surprise. By taking a glance at our viewer's guide for G3 it was painfully obvious that Saturday was going to be one of the most enjoyable days for a Counter-Strike fan in 2014. To me it was the most exciting day to follow Counter-Strike: Global Offensive so far, though the amount of interesting storylines leading up to it certainly played a part. Poor organizing by G3 left us out of a few games that also hurt the experience, but it's impossible not to love this format.

Most of the time we had two very good games on at the same time, while certain rounds had to settle with one clearly interesting game, and two with potential, and at other times three games that saw top teams go head-to-head take place. It's hard to argue against a six team group, as long as it uses a good playoff format with four teams advancing -- IEM already used the format, but they allowed the group winners to directly make the semi-finals, which was stupid at best.

We got to see top teams play many more matches than we're usually spoiled with at tournaments, and for once you can say that the group stage was as interesting as the playoff portion of the tournament, and actually mean it. During G3 no one uttered the words said by fans during DreamHack Summer and Copenhagen Games -- something along the lines of why can't the group stage already be over?

Who under or overperformed?

Given how many results were up in the air coming into this tournament and how multiple teams surprised us, we've decided to take a deeper look at all the players on an individual level. We've gone through all sixty players and compared their career rating to their rating at Gfinity 3 to see who did better than expected of them, who disappointed, and how that may have affected their team's overall performance. What makes this practice more interesting this time, is the reasonable sample size for everyone - fifty-five players played at least five maps, and fourty at least seven.

All of this took place in two days, so obviously current form was king in London. That will continue to be the case in the future as well, but with over ten days to prepare for Cologne, some will improve while others will be worse off. Let's take a look at the players who had the biggest change in their rating over their CS:GO career versus at G3 -- for better or worse. We've listed below the players whose rating was at least 0.10 better or worse than their career average in our statistical database.

Player	G3 rating	Career	Difference
Mikail "⁠Maikelele⁠" Bill	1.37	1.08	+0.29
Håvard "⁠rain⁠" Nygaard	1.09	0.87	+0.22
Peter "⁠dupreeh⁠" Rasmussen	1.31	1.14	+0.17
Janusz "⁠Snax⁠" Pogorzelski	1.24	1.10	+0.14
Kenny "⁠kennyS⁠" Schrub	1.31	1.19	+0.12
Aleksi "⁠allu⁠" Jalli	1.29	1.19	+0.10
Andreas "⁠Xyp9x⁠" Højsleth	1.18	1.08	+0.10
Tyler "⁠Skadoodle⁠" Latham	1.02	1.13	-0.11
Alexander "⁠SKYTTEN⁠" Carlsson	0.85	0.97	-0.12
Braxton "⁠swag⁠" Pierce	1.00	1.13	-0.13
Kévin "⁠Ex6TenZ⁠" Droolans	0.80	0.94	-0.14
Marcus "⁠Delpan⁠" Larsson	0.91	1.05	-0.14
Bjoern "⁠Skurk⁠" Maaren	0.77	0.91	-0.14
Olof "⁠olofmeister⁠" Kajbjer	0.96	1.10	-0.14
Filip "⁠NEO⁠" Kubski	0.92	1.07	-0.15
Fabien "⁠kioShiMa⁠" Fiey	0.90	1.06	-0.16
Richard "⁠Xizt⁠" Landström	0.94	1.13	-0.19
Philip "⁠aizy⁠" Aistrup	0.78	0.99	-0.21
Joshua "⁠steel⁠" Nissan	0.87	1.09	-0.22
Robin "⁠Fifflaren⁠" Johansson	0.72	0.97	-0.25
Christopher "⁠GeT_RiGhT⁠" Alesund	1.05	1.31	-0.26
Richard "⁠shox⁠" Papillon	0.93	1.21	-0.28
Patrik "⁠f0rest⁠" Lindberg	0.97	1.26	-0.29
Adam "⁠friberg⁠" Friberg	0.77	1.10	-0.33
Chris "⁠chrisJ⁠" de Jong	0.77	1.15	-0.38

First thing you'll notice is there are obviously tons more players who underperformed, than those who played better than through their career. That's because the level of teams at G3 was high, and everyone's stats tend to get inflated to their natural level in online matches versus weaker competition. You should also note that the career ratings listed above are after G3 -- which means the difference in reality would've been even larger. That being said, let's go over some of the more interesting entries in the table above.

NiP's combined rating was 1.32 lower than their career averages. To put that to perspective, it's almost as if they were playing four-on-five with GeT_RiGhT - whose career average is 1.31 - sitting out. When a team such as NiP underperforms this much, you should always look at their star players. They carry the heavy loads when they win, so surely they shouldn't be excused when they play like average human beings -- or below.

mousesports' chrisJ had a super poor showing as well, and it makes you wonder how much of an effect that had on the team's performance -- after all, they were a few rounds away from making the semi-finals at a very competitive event, that saw eight of the world's top twelve, and ten of fifteen best teams attend. That gives allu some hope of maybe not having all the weight on his shoulders in the future, but at the same time they face a major issue in making the two star AWPers fit together in the team.

shox had a bad event, but he was ill, which makes you want to cut him some slack. That, and he was the best player in Epsilon's win over NiP. Another surprisingly bad performance came from olofmeister, who clearly didn't show his a-form in his debut under the fnatic banner. The sole dignitas player on the underperformer list is aizy, whose LAN debut with the top team saw him possibly crumble under pressure.

For iBUYPOWER three players make the list. Biggest dropper was steel whose return to his old home in London clearly couldn't have gone much worse, but swag and Skadoodle also underperformed clearly. That's about it for the interesting players who underperformed, so let's take a look at players who surpassed their career averages.

Number one is obviously Maikelele, who had another stellar event. His teammates offered little help - he had a similar task as allu in carrying his team - and Delpan underperforming didn't help. allu also shows up, though his career average is already so high that greatly overperforming it is a much harder task.

Biggest surprise of Gfinity 3 was undoubtedly rain, whose events page was all red prior to G3, but now boasts a solid 1.09 rating from an event where his side played eight maps against top level competition. Two dignitas player unsurprisingly show up, as does Snax from Virtus.pro, and kennyS, who had arguably his best event to date.

NiP are in serious trouble

It's official. I suggested following their losses online that NiP may not be in very good shape, but it was impossible to know until they attended a LAN tournament and put forth their best effort where it counts. They did just that in London last weekend, and came back empty handed. If G3 organizers knew how round robin group stages work, NiP would've been knocked out in groups, but due to the organizer's stupid rules - not understanding the difference in Counter-Strike's rounds  with different maps and the World Cup - NiP instead got to exit the groups and get run over by dignitas in the quarter-finals.

Looking at NiP's results one match at a time in London only makes their case worse; They lost a 13-4 lead versus Epsilon, couldn't close out a 14-14 game versus fnatic as terrorists on de_dust2, barely beat Infused, lost against iBUYPOWER and then barely snuck by London Conspiracy. The way dignitas beat them in the quarter-finals, on another day no less, proves this wasn't all for nothing. Add in the previous week's telling online results, and it's possible something is brewing within NiP.

The Ninjas are still coming off of a vacation and have roughly ten days - which they will probably spend bootcamping - to get in shape for ESL One Cologne. However, this is the second weak tournament result in a row for NiP, having placed 5-6th at ESEA Invite Season 16 Global Finals at the end of June, and marks the second time in the team's history they've placed outside of the top four. That's a bad sign, and could suggest a bigger problem in the team.

The last tournaments where the team's star player and the world's number one fragger GeT_RiGhT played to his career average of 1.31 rating were Copenhagen Games and EMS One Katowice. The former ended nearly four months ago, and since then GeT_RiGhT's resume includes ratings ranging from 0.96 to 1.11 at SLTV StarSeries IX Finals, DreamHack Summer, ESPORTSM Finals, ESEA Invite Finals and Gfinity 3. He is not the player currently that we've grown to consider the world's best.

The team's other star f0rest hasn't struggled as consistently - though he also had two catastrophic sub-1.00 rating events since May - and in-game leader Xizt has been playing around as well as he has been throughout his CS:GO career. On the other hand, friberg has had a string of performances in the red now, and Fifflaren has dipped far enough that it surprises you to scroll all the way down and see that in the early days of CS:GO he was legitimately a good player.

Clearly any team lives and breathes their best player, and just getting GeT_RiGhT back on track should put NiP back into title contention. However, this raises a whole new issue heading into Cologne. What if NiP get knocked out in groups? What if they're sent home in a lopsided quarter-final once more? Will the team finally make a roster change? Will results prevail over friendship? It's a great storyline we could find an answer to in two week's time. Either way, Xizt promises us a different team by the time ESL One starts.

ESL One Cologne will take place on August 14-17 and we're starting our pre-coverage for the event in a day or two. We will also be on-site in Germany for full coverage of the event.

Follow HLTV.org's @lurppis_ on Twitter.