You don't get any Match Score change for drop calling. Yet we all know that can turn the tide of a match - for better or worse! Match Score isn't the final metric for player performance so a system which exclusively used that, without recognising the overall result, would not reflect the true value of each player in the context of a team game.
The current system favours winning - whatever you did, you helped your team more than it hurt your team. If you under-perform massively, maybe it's because you were the first to go scouting and your team won because you called a target or popped a UAV before getting wrecked.
Basing the ranking up/down on a fixed Match Score number has another obvious problem: players will drift to either extreme of the scale, because they tend to play above or below that level. Bottom of Tier 5 and top of Tier 1 will have a cluster of players with almost nobody through Tier 4, 3 and 2.
Setting a fixed Match Score target suitable to each tier would work. Upon entering Tier 3, say, 300 could be the target for raise/lower progress along that bar. Enter Tier 2 and perhaps it's 400. Tier 1 could be 500. Perhaps 100 for Tier 5 but make it quite a 'long' bar so they have enough actual experience before mixing with Tier 4 players.
Tiers are useful so players are generally drawn from 1 level above or below, where possible. This helps games be more consistent. Making the bar fairly 'long' within each Tier also dampens the effect of getting a very good or bad run of luck.
It's also worth mentioning the upward creep is quite subtle. It gives an overall sense of progress to players, encouraging them to keep playing and keep trying to improve without everyone reaching Tier 1 after a month of play.
For context, I'm 1.25 years in and have been middle of T2 most of 2019 so far. That's where I seem to stay as that's my true level, so the upward creep doesn't flatter players beyond their actual ability over this timescale.
Over very long timescales, such as several years, perhaps it does get under-performing players like into T1. Even then, as shown in the chart, they aren't being complete potatoes to break even or rank up with each team win. Also, if they have played that many games they will have seen everything before and that gradual build-up of actual game experience is mirrored quite well by a general upward creep 'XP bar' mechanic, imho.
With a fairly small FP population, Tiers might not work but the overall team win/loss should still affect how each player is ranked for that game. Win as a team, lose as a team...this is even more true in FP from what I've seen.
Solaris ELO is a frustrating experience because there is so little history. It swings back and forth by huge amounts when new players start badly then get better. The lack of a general upward trend can be demoralising - it's fair after 25 games but before that, a bit of a lottery. If FP uses a separate system from QP, such big swings might make it meaningless and cause the games to be extremely imbalanced at first.
Edited by BenMillard, 19 March 2019 - 01:25 AM.
4
OZHomerOZ, on 18 March 2019 - 08:24 PM, said:
