Gentleman Jon — Firstly let me apologise in advance for an even lower standard of article than usual but it’s a bit of a rush to fit this around real life so I’m putting it out a bit early as well. I know, if I could cancel rl entirely I would. You’ll just have to do the proof reading for me.
Anyway, it couldn’t have been closer but TLR overcame incredible resistance from AWS to make the ETF2L Season 17 grand final. Reflecting briefly on the game, AWS clearly prepared carefully for TLR and radically changed some tactics. TLR played into their hands to an extent picking one of AWS’s go-to maps in Badlands but the combination of KnOxXx and Hys’s experience was able to eventually overturn their initial disadvantage.
AWS were just seconds away from overturning their Granary jinx in spectacular fashion but ultimately their bogey map proved too much to overcome. Hys was downbeat about his performance in the match and statistically he took a hit, and during the match (particularly on Viaduct) he seemed to be the focus for AWS and Flippy. MiG will have been watching closely.
TLR vs MiG – The Grand Final
For the second season in a row one of the grand final teams is involved in a controversy, blaming admin inflexibility for a scheduling problem despite knowing about it weeks in advance. You can’t fight city hall so it appears TLR will have to play without Kiler and it’s incumbent upon them to find a replacement, I have very little idea who it will be and I’m writing this early so my analysis will focus on the size of the gap Kiler will leave and only briefly ruminate on his potential replacement.
Both teams have had significant roster turnover during the season, TLR’s troubles were documented in the last article, but MiG have swapped medic from Thun to Byte (abandoning their national loyalty in favour of competitive expediency even further, adopting the US, Canada, the UK and Syria as German possessions overseas) and transferred their main caller as well from Ipz to a slightly uneasy medium between the team and natural caller Byte who will run things purely due to his compulsive filling of that vacuum. With the loss of Ipz it’s not just their strategy that’s metamorphosing, they cashed in on the Ayo Gurl drama to acquire the best soldier in the game in MGE Mike so like TLR in many ways they’ve got better as a result.
Their head to head match during the season finished decisively in TLR’s favour but it was in some ways the catalyst for the lineup changes at MiG so it doesn’t really have much relevance to this game. The current MiG team only had a couple of full fixtures, brushing aside Awsomniac over Viaduct and Snakewater and beating Ayo comfortably over Snakewater and Granary. Recent PCWs between the teams on logs.tf have largely gone MiG’s way but TLR were KnOxXx-less and were using other mercs so once again it’s of limited relevance.
Demomen
Logs against the dropped Mais la Moule aren’t counted in the following analysis.
The graph shows damage per minute in blue, K/D ratio in red and the player rating in yellow. The K/D is measured against the top axis while DPM and rating against the bottom axis.
Kaidus is the top rated demo of the season so far under my Droso dis-approved system, Hys does a little more damage but Kaidus has a better KD. With so much aggression from his teammates Kaidus has had the luxury of playing a slightly more orthodox demo role this season and his stats have received a bit of a boost as a result, but he’s still more likely to kill the opposing demo or medic than Hys.
This illustrates the relationship between damage and kills for the individual players. The yellow bar is the average damage they get for every landed shot and is measured against the bottom axis. The blue bar is the amount of damage they had to cause to register a kill or assist and is also measured against the bottom axis. The red bar is the number of kills achieved per minute, which is measured by the top axis.
Kaidus has a stronger death-matching profile, focusing his damage a bit more on getting frags than the general sort of support Hys offers, although unusually it seems when Hys lands damage he gets slightly better quality on it than Kaidus. There’s no particularly strong trend either way on weapons use, with both demos using roughly the same proportion of grenades to sticky bombs, and both getting 5 pain train kills so far.
We’re looking at two closely matched demomen at the top of the game with slightly differing styles so talking about an advantage is questionable, but the numbers seem to indicate that Kaidus has slightly more individual impact.
Soldiers
Unsurprisingly Mike tops the soldier chart, but the margin is perhaps a little more than might be expected. Given that he’s only had a couple of relatively simple wins in his officials with MiG they should perhaps be taken with a pinch of salt. KnOxXx is far more likely to get a medic in any of his kills so does attack the most strongly defended area of the enemy team more often. Given that Byte will probably be left to look after himself often he’ll have to be on his toes.
Both roamers play in a more conservative style and both work with the black box although Kalho makes by far the most use of it. As a result he has to play a more careful holding style locking down flank locations rather than going for high value picks which Gear is more likely to do, hence his superior rating.
Mike and Kahlo are locked together for the most efficient use of their damage in terms of getting kills, but Kahlo is by far the least aggressive player while Mike outstrips KnOxXx for kill rate by a significant margin. Gear does the most damage without statistical reward in the shape of a kill or assist, but is significantly more aggressive than his roaming counterpart.
Mike doesn’t make much more use of the shotgun so his lowest damage per landed shot stat might be down to real damage plugins suppressing it slightly as his kill rate is so much higher and those killing shots would be reduced.
Overall it seems that Mike can be expected to pick off fragging classes more often and put out more damage, and Gear’s more aggressive style should reap rewards as well. Statistically it seems that MiG should have a moderate but clear advantage here.
Scouts
So, it’s time to see how much Kiler’s absence is going to cost his team.
Smzi tops the pre-final scout stats, although not the most aggressive scout in this match he has by far the most efficient style staying alive and racking up the kills. His partner Bash is slightly more aggressive, and is more than twice as likely to get medic frags so must take more risks with his positioning attacking the enemy combo’s stronghold.
Kiler is a slightly less prolific version of Smzi, also not attacking the enemy medic particularly often his KD/DPM balance is well in favour of his KD. Even so his overall DPM is so restricted his rating is only a hair above Bash.
Rake has more of a risk taking role for TLR judging by how often he kills the medic although as we see later he works with the combo often, and isn’t far behind Bash statistically.
Interestingly Smzi gets a very high kill rate and tremendous damage to kill/assist efficiency. His apparentl lack of aggression is replaced by a very keen sense of timing to get the clean-up frags, he obviously pays attention to calls.
Kiler and Bash have very similar damage efficiency and kill rates although their roles are somewhat different. Kiler has lower damage per landed shot and certainly gets proportionately more kills with his pistol which probably accounts for most of that difference.
Rake by comparison makes the least efficient use of his damage, having the lower kill rate shows it’s not just down to isolation preventing him getting assists. Even so he’s not very uncompetitive.
So the Kiler sized hole is going to be pretty big for TLR. He’s the better scout and one of the highest rated scouts in the league this season. He’s got a good feel for his role and fulfils an important function for his team. So statistically speaking (personal enmities, loyalties and availability aside) who would be the best replacement?
The closest rated scout would be gf18 who had an excellent season. Although he didn’t play that many games for FNX, one of them was against MiG so his performances were not cherry picked for ease. However he played more of a risk-taking role going after the medic so he might not be a good match in that respect. The closest rated scout who seems to have played a similar role is the other FNX scout sheepy who was similar in both rating and style (yes, FNX had really good scouts this season, when they played).
From PCW logs it looks like fl1p may be the replacement of choice and he’s no slouch, but I’ve expunged the Mais La Moule logs from my records for these articles (there’ll probably be something less optimistic for the TFTV stream from this season if it is fl1p) so there’s no direct comparison for this season, but from last season in TSE he could be similar to sheepy in a strong team, and of course he will have built up some understanding with the team. Obviously all of this depends on MiG’s willingness to accept a reasonably skilled merc, hopefully sportsmanship will win through. Certainly none of these suggestions are outlandishly skilled, although if Flippy came up I can understand denying the inevitable Viaduct headshot spree.
Sniping
MiG doesn’t really have a prodigious sniper, and Rake suffered against Flippy in the last game so there may well be a lack of sniping until one team gets into desperate straights. Even so despite the semi-final Rake’s overall stats remain solid and he represents a superior sniping threat statistically speaking.
Medics
In this graph the blue bar shows the kills and assists per death and is measured against the bottom axis, the red shows healing per minute and is measured against the top axis and the yellow shows assists per minute also measured against the bottom axis.
Byte has a marginally higher assist rate but a slightly lower KAD than Mirelin which reflects the more aggressive overall style of MiG. The margins are pretty small but it shows that TLR are slightly more methodical overall, whether they can consistently make use of this to pick off over extenders depends on MiG’s discipline. You might assume that Byte would seek to control this with some drill sergeant discipline but the inner dynamic of MiG probably doesn’t really allow or require this as there are senior players in the team who know their jobs, rather than raw recruits aching for the browbeating.
In terms of ubers it’s all good news in MiG’s favour with a lower drop rate and faster uber rate, although only by very small margins. Mirelin’s drop stats are of course slightly skewed by an alleged semi-comatose performance against AWS so assuming he gets his beauty sleep these disparities may disappear for the match itself.
Team Assists
This table shows the assisted kills between players within the team for games in which all the lineup played.
The scout pair of Kiler and Rake are the most fruitful fragging partnership in TLR, and Kiler also takes part in their 3rd and 4th most important partnerships with KnOxXx and Kalho, illustrating his importance in the core of the team. The traditional pocket/medic pair of Mirelin and KnOxXx are the 2nd most prolific partnership.
Hys’s kill assists are spread amongst the team, slightly favouring Rake over Kiler and both scouts over his soldiers. He’s also Mirelin’s favoured second favourite healing target in combat, also followed by Rake with Kiler and Kalho (with his magic healing rockets) a little isolated from their medic. Interestingly Kiler has more of a defensive scout’s kill profile but with more of a flank scout’s assist profile.
MiG’s current lineup only have a couple of matches behind them so these numbers are comparatively low but they give some idea of the internal workings.
Smzi and Bash are clearly the top fragging partnership, interestingly Byte doesn’t really feature as a top assister anywhere. As we’ve seen MiG frequently have a higher kill rate but it seems Byte isn’t always in with them. Ipz criticised Thun for not being able to survive by himself without dedicated protection, it seems Byte is frequently left to fend for himself with confidence justified by the medic stats above.
Gear & Bash and Mike & Smzi tie for the next most important partnerships, in contrast to TLR the scout/soldier pairs forming important fragging units.
Similarly to Hys, Kaidus doesn’t really share assists with anyone in particular although he rarely combines with Gear.
A (hedged) Prediction
MiG were hot favourites at the start of the season but KnOxXx’s arrival in TLR forced their hand making fundamental changes to the team. Like building up a chess position it appears that MiG has a series of small advantages that add up to a clear edge. But the gap between the teams isn’t big, something that could conceivably be bridged easily by one or two players being at anything other than the top of their game.
Even so TLR seem to be talking themselves into a losing position with the loss of Kiler, whether they’re trying to pile some psychological pressure onto MiG or whether they genuinely expect that they won’t be able to challenge is unclear. If it’s a genuine sentiment then the risk of being psychologically beaten when behind but before the result becomes decisive is much greater.
On the flip side the new calling structure in MiG hasn’t really been put to the test in a really tough game with something on the line, so if TLR can put up enough resistance as nerves fray it may prove difficult to solve problems on the fly in game.
It’s a stats article so I have to go with the numeric conclusions, MiG are favourites but with TLR holding a pretty good underdog’s chance at pulling off the upset.
Originally published on VanillaTV (vanillatf2.org) on 2014-04-05 by diemasher.
Recovered from the Internet Archive (archived copy) as part of tf2esports.com's preservation of VanillaTV's TF2 coverage. The original site is no longer online. All rights remain with the original author; reposted for historical reference.