Benjamin Jakobsen
164 posts

Benjamin Jakobsen
@bentoft
Working as a developer at @HLTVorg
शामिल हुए Haziran 2009
74 फ़ॉलोइंग81 फ़ॉलोवर्स

@cuRs581 @MischiefCS2 As mentioned at the top post, there was a bug.
However .... you cannot compare a vrs result and a forecast like this.
They are based on different timestamps and remember that fazes rank depends on a lot of other stuff than their own results.
English


@MischiefCS2 It was indeed a bug that could happen in rare cases when a team reaches playoff in one tournamen while playing in another tournament.
English

@JesperLarsen222 This was an unrelated problem that happened on the day of a cutoff ranking before valve releases their ranking. The details page shows the live ranking, but should be the cutofff date ranking.
Anyway ... fixed
English

@JesperLarsen222 Found and fixed a problem that could happen when a team gets far in one tournament while also playing another tournament.
English

If anybody is interested you can follow me build out dust2 in LEGO in minifigure scale on my instagram: instagram.com/bentoft_lego/

English

@blameFFFFF @BIGCLANgg Nice .... who would have guessed that blameF belong in a team called BIG!
English

Won Roman Emperium Cup, super happy to win my first tournament with @BIGCLANgg 🥰 i feel super comfortable in this team and I am very lucky that all the guys is trusting me and gave me such a warm welcome to the team 🥰❤️

English

@JesperLarsen222 @Auxzap I think I have fixed it, but it is not yet deployed.
English

Yeah, I think its something that @bentoft could pretty easily improve/fix so that the ranking predictions take these extra points into account for better transparency.
It will probably be live on HLTV monday 😂
It would be nice if Valve employees worked with the same "agility" as HLTV employees though.
Because the real fix should be in the VRS model in my opinion.
I don't think its a deliberate design choice from Valve. Its just a result of the implementation which is probably not considered beforehand.
Just like the event weight change wasn't fully analysed in my opinion.
For instance if you have a main event with 500k prize pool, and a play-in event also with 500k prizemoney paid out, then the play-in event will get 750k VRS impact, while the main event will get 500k VRS impact. So the play-in will give more points than the main event now.
We might see a variation of that with the upcoming Blast Bounty depending on how HLTV assigns the Prize pools.
It just doesn't feel like they are considering all scenarios before implementing🤔
English

The VRS winners of Starladder Major Stage 1 -
Blast Bounty invites secured for many teams -
FlyQuest and Ninjas in Pyjamas the big winners!
Many of the teams in stage 1 have fought for more than just success in this major.
For many this major was also about securing invites for next season. Some succeeded, some failed.
Major Stage 1 counted towards December ranking which will decide Blast Bounty invites.
Top 28 and probably also top 32 will secure an invite.
Btw -> Major stage 2 matches will NOT count towards December ranking as the event completes after the December 1st cutoff, meaning that for the major participating teams this race is now over although results from other LANs this weekend can change the last Blast Bounty invites.
A team like NRG will need the right results from other events this weekend to keep their position for an invite.
Apart from Blast Bounty invites, teams are also fighting for positioning for the January ranking which will be much more important with IEM Krakow (top 24) and PGL Cluj (top 16) invites on the line.
Below table shows the VRS impact of Major stage 1:
Note I: Last column is "hidden" points from event weight that won't be "paid out" until December 14th.
FlyQuest, NiP, M80 were the big VRS winners of Major stage 1 when it comes to Blast Bounty invites, but more importantly they are still alive and fighting for a top 24 spot on January ranking which will be very important for staying in the so-called tier 1 invite bubble.
FlyQuest and NIP in particular showcases just how impactful 3 big LAN wins are when you are low on LAN wins to begin with.
They basically did "a Liquid" on a smaller scale and secured their next season relevance from this stage 1 success. Now they are very much alive and kicking!
All the teams that didn't qualify for stage 2 except Legacy will be in trouble for next season.
Most notably GamerLegion who almost certainly will now miss out on the IEM Krakow invite if they don't play anymore before January and if that happpens they are basically out of the tier 1 invite bubble.
I guess we will see them at DraculaN or something similar? 🤔
Finally, a bit of a complex topic, but good news for the teams from stage 1:
It's worth noting that all the teams have not even gotten their full impact of Major Stage 1 results yet. There are more bonus points to come!
In the VRS, play-in "events" such as Stage 1, gets an added event weight based on the prize pool of the main event.
However bizarrely, Valve has for reasons we don't know decided that in the VRS model the upgrade of the event weight of the play-in event, doesn't happen until the main event has finished!
That means that once the major grand final has completed (and everything is recalculated with new event weights), a bunch of extra points are added for all the stage 1, 2, 3 teams with a delay..
It's over engineering in my opinion, just adding unnecessary complexity to an already very complex model, but that's how it works.
Please @valvesoftware , change this silly mechanic 🙏
In the table I have shown the "bonus-points" from event weight upgrade that the teams have earned but which will not be added before the Grand final of the major completes (meaning they will impact January ranking but not December)
I bet teams like NRG and Imperial would have liked to get these points already now!
With those "delayed points" added they would maybe secured the Blast Bounty invite.
Now they won't be "paid out" until December 14th for no good reason (in my opinion).

English

As Stage 2 gets closer a small warning:
HLTV Simulator does NOT predict Stage 2 opening matchups for your favorite team if you think so😉
Below a peak into what we already know about stage 2 opening matchups if you can't wait 👇
Maybe you play around with the very nice HLTV major simulator like I do. It's a great tool but in the current version it has its limits and it could potentially misguide you when it comes to the next stages.
In the tool, you can finalize Stage 1 and go to Stage 2, but be aware that when you do so it will NOT give you the correct opening matchups yet as the correct seedings are not coded into the tool.
This is just a warning for those who might think so.
I've found this tool be be taking Valves seeding rules into account for correctly predicting opening matches:
majors.im/2025/budapest
But back to HLTV simulator and how you may be misguided.
Below I used HLTV's simulator to finalize stage 1:
And when I then jump to Stage 2 the Simulator will show me the "correct" 16 teams, but the seeding and therefore the opening matchups are not correct:
So don't use HLTV simulator for stage 2 before it has been updated with the correct matchups.
Here is instead how it the seeding works:
For stage 2 the 8 initially qualified teams will be seeded 1-8 according to most recent official VRS ranking (November ranking):
1 Aurora
2 Natus Vincere
3 Liquid
4 3DMAX
5 Astralis
6 TYLOO
7 MIBR
8 Passion UA
These 8 teams will then be matched up against seed 9-16 this way for the opening round:
1vs9, 2vs10, 3vs11... etc until -> 8vs16
The 8 teams advancing from stage 1 will be seeded according to first their W-L score and second according to their Buchholz score, which is actually still depending on todays results, even for 3-0 and 3-1 teams. So "work in progress" for most teams.. 😉
But we already know this:
- M80/FlyQuest will be seed 9/10
- B8 will be seed 11 (todays results cant change it)
- fnatic/NiP will be seed 12/13
- The 3 teams advancing today will be seed 14-16
With this knowledge we have the following most likely opening round matchups for Stage 2 and even one of them already set in stone (Liquid-B8) :
If your favorite team is among the 3 that can advance today, then you know that they will face either TYLOO, MIBR or Passion UA and vice versa if you're a fan of those 3 teams.
You can play around with all the possible combinations in the tool I linked above.



English

🆕🛠️ VRS rankings and impact forecast for matches have been introduced in the latest update to the HLTV Android App
And new map insights, which can be inspected after veto process is complete 👀
Get it on Google Play Store:
➡️ play.google.com/store/apps/det…

English

@JesperLarsen222 Fixed on next deploy. Also: what a stupid case!! 😂
English

HLTV live ranking is slightly out of sync with reality due to HOTU's "creative" gaming of the VRS model.
HOTU has not played a game since these shenanigans described below took place and why would they..
They might deny another team a Blast Bounty invite from this very creative gaming of the VRS system!
We can't say they have broken the rules, but they have stretched them to the limit..
A summary of what happened:
HOTU wanted to forfeit a game in ESL Challenger League back in October, because they wanted to focus on ESL Pro League.
However, if they forfeited it would cost VRS points, so they were "clever" and fielded a roster with only 2 players from their current roster but with 1 benched player so that they still did not break the invite rules.. They played first map and forfeited the next map and went on to forfeit the Lower bracket game completely.
hltv.org/matches/238646…
Very "clever", but also "too clever" in my opinion.. as in bad sportsmanship. They were obviously gaming the system to avoid getting the minuspoints from forfeiting the games.
The only way to defend their actions is if they claim they did it to put a spotlight on an issue in the VRS model.
It is objectively an issue that it would cost HOTU points to accept the last minute replacement invite for EPL, but then again, the schedules weren't conflicting and 2 players did play, so I assume all 5 players could have played. They just didn't want to do it "blameF style" and play multiple events while at a LAN! So they decided to be "creative".
It gets even worse though.. This "fake" lineup not only absorbed the minus-points from the 2 forfeits from the main roster, it also shared players for some of HOTU's old matches with overall negative ranking impact. 26 old matches swapped from main team to "fake roster" each time they "played" a ranked match respectively.
Since the last forfeited match in the ESL Challenger was scheduled some hours after the last match in EPL that the real lineup played, the "fake" lineup now "holds" quite a few minuspoints that belong to the real HOTU lineup. Without even playing!!
But as soon as HOTU's real lineup plays their next ranked match they will rightfully get those old matched back again from the fake lineup.
I'm guessing that won't happen until after Blast Bounty invites have been handed out though as its very convenient that the fake lineup holds those minuspoints!
Be aware (@bentoft) that HLTV live ranking can not handle these shenanigans (HOTU's tricks broke HLTV live ranking! 😉).
HLTV live ranking shows HOTU as rank 31 on the December 1st ranking that will decide Blast Bounty invites:
However, reality is that HOTU is placed 25th with 1308 points due to their "fake" lineup currently "holding" all their 26 matches from before 10.09.2025.
Here from the official Valve gitHub from the November ranking:
You can see how this "fake" lineup that only played one map and forfeited the rest (the 2 matches from ESL Challenger in the yellow box), also holds 26 old matches. Matches that HOTU's real roster will claim back when they play their next match.
Probably only after the Blast Bounty invite has been secured.
I guess a fix is needed on HLTV live ranking as it actually affects the ranking displayed for top teams in a minor way too since HOTU did play in EPL.
Example: Vitality lost to HOTU and should have a few more points than HLTV live ranking shows because HOTU should have more points than they have... and so on.. minor differences, but yeah better get it corrected.


Udknud@JesperLarsen222
A very interesting way to "abuse" the VRS to get a free forfeit has surfaced as pointed out below👇 Just clever (ab)use of the system or bad sportsmanship? Personally I think it's over the line, but yet totally within the rules as far as I know, so again.. it's the rules that can't handle being bent to the max.. 🤔
English

A little surprise for players competing in Chengdu 👀
HLTV and @1xBet_Esports have teamed up to design another special MVP medal
Who is leaving China with this memorable award? 🇨🇳🐼



English

@pimpwinneche Has to be m0NESY. Just fun to watch when he delivers.
English


@StrikerHLTVorg Thats a green light to bring a beer on stage - taking a sip every time you have completed a sentence. Would be something!
English











