Enough already 343

This ridiculous 50% win percentage system you have in place is an absolute joke and it’s what is killing this game. It’s so blatantly obvious that teams are purposely unbalanced to guarantee a win or loss and it’s the exact same system you have in MCC. Almost every single match is a blowout on either side and it makes the game either an absolute snooze fest when you win and incredibly frustrating being on the losing end. I constantly check the KD’s of players after a match and there’s always a team with multiple onyx players while the other team has multiple scrubs.

Just balance teams based off of the players KD’s and be done with it otherwise this game will be completely dead if it’s not already.

5 Likes

I think it tries to match teams on their average MMR so that both teams should have a 50% chance of victory (rather than trying to make sure one team definitely wins which would just be rubbish for the losing team).

Overall this means players should be winning 50% of their matches and should experience sweaty games 100% of the time.

If you’re experiencing blow-outs a lot it sounds like their algorithms are not working correctly or the population is very low, but I’m not sure though.

1 Like

It can be related to multiple players on one team simultaneously having significantly above or below average performances per their usual performance. Player performances can naturally fluctuate match-to-match based on numerous variables.

It’s also possible that someone or multiple people on a particular team quit leaving unreliable bots in their place. This can quickly cause a cascading effect to the competitiveness of the match.

Of course population issues are often at the forefront of these types of problems. As the matchmaker struggles to locate similarly skilled players within a reasonable search period and acceptable connection therefore it’ll allow compromises in order to ensure people are at least given a match.

Nah man it’s the same thing in MCC, if you check everyone’s KD after the game they consistently put 2-3 high positive kd players on one team vs a team with 3-4 negative kd players.

3 Likes

This has been my experience with Halo since Halo 2 vista.

1 Like

I’ve felt it since Halo 2 on the XBox myself, although I’m not sure if it was as noticable as this. I always presumed though it was matchmaking in general that gave this feeling towards me, since I never exprienced this is games before that just used custom server browsers.

The match maker is doing it’s best to create evenly matched teams.

The 50% win rate is a natural consequence of that. There is no part of the algorithm that specifically decides the result of any particular match.

The downside of a lower population is that matches are being created with a wider range of ranks. The downside of this is that the top players have to carry, the bottom players have to survive, and that the team structure is precarious (leading to blow outs if someone quits or the team doesn’t gel on the objective tactics).

KD’s are such an unreliable statistic.

And keep in mind that in an unbalanced team those “scrubs” still have an expected contribution. It may be negative (expecting more deaths than kills) but they can still contribute by going less negative than expected. Often the result of these games comes down to which team makes better use of the bottom players.

If you mean off their actual rank - it’s trying to. We just need a bump to population. Fingers crossed for Forge / Season 3 to provide that impetus.

-scribbles down “ridiculous 50% win percentage” onto the already lengthy list of the ‘only reasons’ this game is dying-

For me the 50% win percentage system is fine. What’s really frustrating is the distribution of CSR. You shouldn’t lose a lot of CSR if you did well on the game (good K/D and objective score).

2 Likes

We’re gonna need a bigger pad.

1 Like

The distribution of CSR score depends on three main factors;

  1. The result; up with a win. down with a loss.

  2. The size of the upset (your skill compared to theirs); the bigger the upset the bigger the jump (to a max of 15 points).

  3. Your MMR; remember that your CSR is basically trying to chase your MMR - so it jumps towards it, but resists moving away from it.

When you see team-mates go up or down differently - it’s because points 2 and 3 are different for them.

Personal performance in a game influences your MMR - not your CSR.

And the metrics are KPM and DPM (kill and death rates).

KD and objective score have no bearing on your rank. The former because it is a rubbish stat and the second because they found using it added no accuracy to the ranking system.

You can see how you went on the Waypoint app by looking up your expected kill and death counts for the game and comparing it to your actual values. You can find them by hovering over the game on the graph of last 20 results.

Any MMR change is reflected back onto your CSR in the next game (via point 3 above).

1 Like

Thanks for your reply.

Still doesn’t make sense to me that objective points are not used for weighing. I don’t understand why there has to be a hidden MMR.

If that’s what matters, why not get rid of CSR and just use MMR for ranks and make it public

Playing the objective is important for the win - and winning is the most important part of ranking up. So it’s already included. As for who does it - it doesn’t really matter from the team perspective (don’t forget it’s a TEAM game) - but the better teams share the repsonsibility.

When Microsoft ran the data they found objective scores specifically didn’t increase the predictive ranks. Neither did KD, KDA, assists, or medal counts. So you can’t use that metric to rank up. If you did you would just start to lose a lot, be miserable, and rank straight back down.

I don’t understand why it’s hidden either. Josh Menke used to generate print outs for people who asked. So there is no real reason to “hide” it.

You have to be prepared for what it is though. It’s isn’t an actual number. It’s a range of probable ranks. The number oft quoted is the average of the curve - but you need to keep in mind that there is a 50% chance your actual rank is lower than this, and the curve can be quite wide!

The other factor about the MMR is that it jumps around a bit. Probably more than we expect. And this allows the system to deal quickly with changes in form (which fluctuates more than you would think - up to 150 points on any given day).

This was probably why Josh rarely just gave anyone a number. He usually gave them a graph of your MMR over the last 100 games or so.

Your CSR is just a way of describing your MMR. And by having a slower trajectory (15 points max per game) it’s a smoother journey.

People get really upset if their CSR drifts down 5 to 10 points. Imagine if you saw your MMR drop by 4 to 5x that.

The key is to relax. Play. Have fun. Let your MMR do it’s thing in the background. And smile (or jump around excitedly) whenever your CSR hits a personal best.

The issue is the quickplay plays like ranked. I’m so tired of this excrement show of a game.

I’ll bite.

How would you like it to play?

Specifically what win rate and K/D fulfils your desires?

Some sort of nap is in order.

Just unistalled halo infinite for this very reason! i am so sick of playing with 3 players that cant look up or down correctly let alone place well during the match. yet on the opposing team there are 4 fully aware individuals that seem to understand every aspect of halo… halo ce- halo infinite. the definition of insanity is doing the same thing expecting different results . does anyone feel like that sums up matchmaking in halo perfectly ?

Your Halo Tracker doesn’t look too bad?

210 games with a 51% rate. You’ve won 11 out of your last 25!

And I think you are being a bit hard on your team mates. You have a KD just over 1 (which is all anyone can ask) - but you have gone negative in 11 of your last 25 or so games.

But anyway. Hope you enjoy whatever you decide to do next.

1 Like

Just match people on the rank that they have currently achieved and everything will work out.

I’m in the philosophy that everyone should start at silver and plat and work up. But… You should play people at your rank. I shouldn’t be in diamond and only play onyx players because I did good in mm a year ago… Makes me not want to play

I don’t bother playing ranked anymore cause I just no longer care about it so I’ll remain unranked.

1 Like