The RP logic is easy to understand when it comes to the numbers gained or lost. First: What the winning team' s players gain in sum is exactly what the others lose.
You may check your stats in recent games and compare with the other teams points. Than you'll notice.
I Just Fed Chat with some data from our last games.
EASHL RP Distribution – Match Data
Before the game
Opponents: 21145 / 23331 / 18828
Team average: 21101.33
Us: 25128 / 25361 / 25059
Team average: 25182.67
Exact team average difference: 4081.34 RP
After the game
Opponents: –194 / –260 / –144 → Total: –598
Us: +200 / +194 / +202 → Total: +596
Difference between total gain and loss: 2 RP (very likely caused by rounding of hidden decimal values)
Observations
Total RP gained and lost is nearly identical.
Highest-rated opponent (23331) lost the most (–260).
Lowest-rated opponent (18828) lost the least (–144).
On our team, the lowest-rated player (25059) gained the most (+202).
The larger the RP advantage, the harsher the penalty for losing.
Large rating gaps create a high-risk / low-reward scenario for the higher-rated team.
An upset win against much higher-rated opponents results in a very large RP gain.
Consistent with a near zero-sum MMR system with individual weighting based on rating and rating gap.