Jump to content
CyReN

Halo World Championship 2018 Teams, Seeds, and Discussion‏

Recommended Posts

My dream for esports is for it to grow to the point where they have scout teams and practice squads that they utilize.  This would require more financial incentive, longer contract timelines etc but would be cool to see.

 

One thing i have brought up before is that i see opportunity for growth with the way that teams practice.  Instead of scrimmaging 5 days a week, do it 2 or 3.  The other few days get with a quality AM team or hell even another pro team that is willing to experiment, and setup specific circumstances and run them over and over.  For example, instead of playing Truth flag once in a scrim, set it up as a 2 minute, 1 cap game and play it 5 or 6 times in a row like that.  videogames don't have the real life limitations that traditional sports do, so there is real opportunity for creative practice.

Over in COD Eunited picked up a team of talented U18's (inelgible for CWL) as their reserve team for the main team to scrim and work on strategies with. 

 

Definitely a step in the right direction, and also gives them first access to talented players already in their system once they turn 18. 

  • Upvote (+1) 2

Share this post


Link to post

Over in COD Eunited picked up a team of talented U18's (inelgible for CWL) as their reserve team for the main team to scrim and work on strategies with.

 

Definitely a step in the right direction, and also gives them first access to talented players already in their system once they turn 18.

I’ve been wanting major orgs to partner with smaller ones for a while now. Someone like NV could “partner” with a relatively unknown org for practice and talent pool. The smaller orgs can’t keep the best players anyway so it seems like a win win if they could work out the details.

  • Upvote (+1) 1

Share this post


Link to post

tl;dr?

They've adjusted it to include factors outside of just win/loss. Match result prediction rate has rise from 52% to 68% with the new system.

  • Upvote (+1) 1

Share this post


Link to post

They've adjusted it to include factors outside of just win/loss. Match result prediction rate has rise from 52% to 68% with the new system.

I guess I should really just read it all but which factors and is it for a specific game or globally to any system they've applied it to? Second question being how does this system change and how easy is it to abuse if players are conscious of how it works? Its easy to use k/d as an example when nobody knows about it but once players understand that if they die less their rank is better what happens? Everyone just baits their team? Do the entry fragging type of players come out with a lower rank despite being essential to winning? It would be a fun conversation to have with the people developing it as people who actually play FPS games

  • Upvote (+1) 1

Share this post


Link to post

 

 

Wow, that footage of Bravo and Onset casting the last match was awesome. You rarely get to peek behind the curtains like that. So much energy/tension just in the casters. Sick stuff.

 

 

tldr: GoW4's system.

 

Kills and Deaths affect individual ranking in addition to win/loss.

 

It's fucking stupid.

 

 

I guess I should really just read it all but which factors and is it for a specific game or globally to any system they've applied it to? Second question being how does this system change and how easy is it to abuse if players are conscious of how it works? Its easy to use k/d as an example when nobody knows about it but once players understand that if they die less their rank is better what happens? Everyone just baits their team? Do the entry fragging type of players come out with a lower rank despite being essential to winning? It would be a fun conversation to have with the people developing it as people who actually play FPS games

 

Give it a chance. Just because kills/deaths are accounted for doesn't mean it's broken. When they say the predictive power went from 52% to 68%, that's a huge increase. The system clearly works better on predicting outcome (they used 23 million games to train the algorithm and tested on 3 million games to get the 68% number). It's important to note that they don't disclose the actual weights the system tuned for the parameters associated with kills/deaths. Relative to other factors, the impact of kills/deaths might actually be small. 

 

In regards to gaming the system to just go for KD to try and boost your rank:

 

 

Another key aspect of these papers is that they use latent variables whose sole purpose is to model the scoring ability of a team. In TrueSkill2, the goal is to correlate kill/death counts with the existing player skill variable. In game modes where the objective to score the most kills, then we expect this correlation to be high. In game modes where the objective is to capture territory or simply stay alive as long as possible, we expect this correlation to be low. Even in modes where the objective is to score kills, there may be teamwork effects where players can help their team win without scoring kills themselves. We ultimately want player skill to reflect a player's ability to win, not their ability to score kills.

 

Our solution to this problem is to predict individual statistics from the performance variables. As explained in section 2, the performance variables determine the winner of the match. Therefore the model cannot choose skill ratings that only predict individual statistics. They must predict the match winner first, and secondarily predict the individual statistics. We expect that the most benefit from including individual statistics will come in team games. In such games, the players on a team can have different performance values, as long as their total performance is consistent with the match result. TrueSkill2 uses this flexibility to give more credit for a win to players with the best individual scores, something that classic TrueSkill cannot do.

 

Here's a table showing a player's kill rate (kills/minute) and win %, along with the system's ability to predict their win %:

vTLy9n9.png

 

It's also important to note that the parameters for kill/death (and all other parameters) are tunable per game mode, so the overall skill calculation would be different between HCS and Super Fiesta (I think in the paper by game modes they meant playlists, but they might've meant game modes as in slayer vs. CTF, but I'm not sure).

  • Upvote (+1) 7

Share this post


Link to post

snip

I agree that's a big jump my concern is that their data changes once player habits change due to understanding how the system works. Obviously the weightings can be adjusted but this was actually already tried in Reach and you just had people statting and refusing to push because they knew losing and going super positive was actually an avenue of ranking up. They'll have to plan for that and come up with a way of mitigating it for it to work out when players know its working that way rather than taking data from people playing under the impression that it was just about the W

Share this post


Link to post

-snip-

Great photos and an awesome video man! Look forward to seeing your work at World's.

Share this post


Link to post

We’re bootcamping now from the 5th I’m happy to say! I was recently diagnosed with Crohn’s disease and given medication, which is why we wasn’t going to go because I wasn’t well. Hopefully the meds work out well, as the last few events like Gfinity I was in a lot of pain with it and it’s stressful in general when you have to stay on stage and carry on through it.

Really glad to hear you guys will be bootcamping. You have a solid chance if making it out of groups which would be awesome to see.

 

Sorry to hear about thw crohm's diagnosis. A member of my family was diagnosed a few years ago. It can be pretty rough for sure. Hope it doesnt impact on you too much. Good luck to you and the rest of the vexed guys.

Share this post


Link to post

Snip

I will gladly put my name to anything required to get you all access for Worlds

 

A petition or one of those "(x) number of retweets" things. Whatever helps or whatever is needed.

  • Upvote (+1) 1

Share this post


Link to post

I hated watching EnVyUs when they were winning tournaments.  They didn't play objectively smart Halo.  They made a ton of plays that -relied- on execution, where, if it worked, it worked great, but if it failed they were at an even bigger disadvantage.  But when Huke and Mikwen were going off, they were winning.  So it worked.

 

I love watching this Splyce roster because, while it seems like they're just flying like mad monkeys...they're still being very cognizant of what space they're controlling and where their priorities on the map are.  It's fast, high-execution, but safe Halo.  If they mis-execute, they're not screwed

I know what you mean, Splyce are incredibly clean when it comes to positioning and timing of pushes.  TBH i think its as much the fault of H5 in that it makes all these tools available and Splyce have finally nailed the most effective way to exploit them which is i think more what bothers me.  Really interested to see if they can maintain this standard going into worlds though and if any teams can find an effective counter.

Share this post


Link to post
Guest
This topic is now closed to further replies.

×

Important Information

By using this site, you agree to our Terms of Use.