There were a few things I was not happy with during the release of H4 MM. I would like to see some improvements for H5. Most of these also affect MM throughout the life of the game:
1. Skill ranking.
Skill rank (however it is done) should start at the median. Should you choose to continue with CSR for skill matching, you have lots of data from H4 on what the median rank is that playlists tend towards, with players with small sample sizes eliminated. That’s where it needs to start. NOT AT ZERO (or one).
If you start at the minimum, then low-skill players have to wait for everyone else to rank up before getting matches with other equally-skilled players. If you start at the median, then low-skill players drop while higher-skilled players simultaneously increase. This makes the skill gap more apparent much more quickly, and allows better matches in a shorter time. See my post here for a more detailed explanation, if desired.
2. Playlist searching.
Please see this topic for details. It describes a matching method that would make matching faster, is more robust to population size changes, gives users more choice, and allows the option to prevent lobby backouts. I have actually simulated this with a model in the R statistical programming language, and the benefits are quite real, especially when playlist population is low.
Importantly, it removes the need to take any gametype out of matchmaking.
3. Playlist organization.
(A) I do think there should be a social / ranked split. (B) I think the social options should be limited and not include Objectives (I have no intent to debate the exact makeup here). (C) There needs to be a true Classic playlist option from the very beginning (no sprint, AAs, or perks). The popularity of such a playlist cannot be gauged based on H4, as by the time something like that actually came around, most of the population was gone. If Suggestion #2 above is taken, the playlist organization becomes much easier and simpler than what we have in H4.
While I like them, they need some more thought than they received in H4. I have some detailed suggestions in this thread, but the long and the short of it is that getting commendations should never involve sacrificing the win. For commendations that are unrelated to winning - but still cool enough to have - they should be relegated to a permanent, rotating commendations playlist.
If retained, specializations should reflect your playstyle and should not give perks as a reward. So, Wetwork means you assassinate a lot of people. And so on. It does NOT mean you get a Stealth perk. I do not agree with unlocks for weapons/AAs/perks. Unlocking armor, emblems, skins, and so on is cool. I like that. But unlocking gameplay-critical items should be left alone. More details here.
This may be a pipe dream, but I would like to see map packs considered a core part of the game - covered by the initial purchase. Instead, I would like to see DLC revenue come from non-gameplay items, like armor, skins, avatars, and other cosmetic items. I have yet to buy a map pack (I refuse), but I have purchased an armor pack because I wanted my helmet.
Small issue, but I don’t see SR as adding value. Progression ranks with few (or preferably zero) bennies don’t really make sense, especially as they are capped. For folks who like progression ranks, why not just have an XP total on their gamercard instead? Simpler and no cap. Instead, gamercards should display skill ranks in ranked playlists, and no ranks at all for casual ones.
It would be nice if there could be some sort of partnership with the regular Forge communities such that a “Forge” playlist could be supported that would appear in the regular playlist lineup (though stats would still count as custom stats). The HC / FH / BC management could link a certain number of maps per week that would appear in the playlist. Anyone could join - it would operate just like a regular playlist, but with maps chosen by those 343 partnered with. After playing, the players would be asked to rate the map and optionally add comments. I think this would provide a better beta test for community maps - and allow more and better quality maps - into matchmaking than currently happens.
It would also benefit the Forge communities. As the maps need to be selected by the respective sites’ management, it would promote Forgers participating in the test lobbies to get their maps noticed. It would also remove some of the overhead at 343 to try to sort and test community maps for suitability in matchmaking.