Not necessarily. Your MMR could have stayed the same during your win streak, but each win still gave you +1 CSR, so your CSR was going up albeit slowly. So you got a 15 game win streak, where you went +1 for every win after your CSR converged with your MMR, and if your MMR didn't move then your CSR would have been 15 points higher than your MMR after the streak. Then upon the loss, you lose 15 CSR because your CSR was over your MMR. But it doesn't mean your MMR went down; it could have stayed the same, or it could have even gone up a little bit (but not at the rate of +1 per game over 15 games).
It is possible. But I'd argue that scenario is also suggestive of the same problem.
I'm not ruling out the possibility of something being wrong; you could very well be one of the exceptions like Juror. But while you may have felt that your performance was deserving of an MMR increase, isn't it possible that you didn't actually meet the expectations Trueskill2 had of a higher skilled player? You say your record best K/D during the streak was 16-1, which sounds very nice. I can't find that game in your history; could you link it? I'm wondering what the CSRs of your opponents were.
Probably easiest format to see stats from a lump sum of my recent games: http://halodatahive.com/Player/Record/the%20sig%20fig#///Stat summary
W: 22 L: 1
K/g: 11.95 D/g: 7.91 A/g: 6.45 K/D: 1.51 (tossed stats where other team quit early game)
1st place: 10, 2nd place: 12, 3rd place: 1
Lowest kills in a game: 9 (1 time)
Approximate average opponent rank (did my best having to approximate CSRs)
: Diamond 4
MMR may be greater than CSR if they haven't grinded yet. So Diamond 3/4 should be considered the low boundary.Average team score margin
: +8.32 (first to 25)
And yes, my MMR should theoretically flatline. My skill is my skill. The average rank of my opponent is slightly less than my current rank, so I should be expected to win more than I lose. I probably should not be expected to go 22-1 with a similarly ranked teammate and I probably should not have those stats.
Here's the thing: I am becoming increasingly confident that Juror is not the exception. I haven't seen a ton of evidence that there is much difference between mid diamond and onyx players, based on what I have seen from my opponents and available stats. Heck, we went 2-0 against Juror, the exception, this season. Perhaps onyx players have some common variable in their game history that avoids the "Juror quirks." I can only guess.
If I had to make a first guess as to why higher end players (diamond/onyx) drop in MMR over time, it'd be an overshoot on the expected kill rate. I can easily envision a scenario where MMR stagnates or drops against lesser opponents because the other team camps it out to varying degrees, as often happens in doubles. It can be a slower game, and it wouldn't shock me if even in games we are winning comfortably, we are going down in MMR because maps are still large and we can't find action, or we do force action and the game is closer than expected because you are jumping into poor situations. It is completely plausible that I had multiple games in that streak where we spent too much time hunting and therefore "under-performed."
The gold standard example of this problem was Mikwen's HCS game history, where the he dropped significantly in a 100-0 win because the other team hid. Menke's remark suggested he believed this was a problem limited to HCS pros- and as I've said in the past, I don't think he appreciates how frequently those sorts of games happen beyond that ultra-elite tier.
I still suspect that there are variables that are less applicable on this end of the curve. Which is fine- just bringing more attention to it in the event it can be explored and tuned.
I also acknowledge that according to the game, I may indeed be a diamond 6 player. But, and forgive me as this will undoubtedly sound like sour grapes, if 22-1 and not ranking up is the game working as intended- then I'm just less inclined to put time into ranked. I want to test myself. I want to win. I don't want to force myself into a super aggro, potentially selfish, and not necessarily better play style just to tackle some weird metric. It's just not a good feature and seems to be behaving more strongly than just a "are you being carried indicator." And heaven forbid the game freezes in the pre-game lobby and gives me a DNF (happened twice this week)- it'd undo 3-4 hours of winning games.