A very well written exposition. There are a few things that I don't agree with but they are minor points: 1) The title. It is catchy and probably not meant to be taken literally, but I can't help it. To me, at least two more things are missing or utterly misunderstood, particularly on the CBMM side: [b]Statistics.[/b] People throw in terms like median, mean, average, bracket, bucket etc. without fully, or at all, understanding the underlying math and the nature of the phenomena it describes. As a corrolary, [b]skill.[/b] People seem to assume that skill is normally distributed, with there being zero evidence to support that argument, and numerous studies pointing to much more skewed distribution. There is also the assumption that the only way to implement SBMM is to have skill brackets. This model suffers from low population numbers in the tail ends of the distribution, as correctly pointed out by CBMM supporters. An alternative would be to have percentile-based brackets. Top 20%, above-average 20%, average 20%, below-average 20%, bottom 20% for example. This model would significantly reduce connection problems (by increasing population pools) at the tail ends without greatly exacerbating skill variance induced problems. The biggest skill variations will be in the tail-end brackets. However, the top 20% bracket is more likely populated by players who are willing to invest time to improve (as indicated by their "git gud" remarks,) and the bottom 20% are in the shallow end of the learning curve, and shouldn't have a really hard time climbing out of their bracket. Furthermore, some quote IB population numbers, but that is purely a vanity metric since it does not accurately model any meaningful business outcome. A much more reliable metric would be [b]player retention[/b] and [b]sentiment[/b]. There are numerous ways to measure the former, but for the sake of this particular argument it would suffice to measure the number of IB games played after pinnacle bounties have been completed. As for the latter, if Bungie isn't running sentiment analysis on these forums, Twitter and Reddit subs, they should really catch up with the times. [b]Psychology.[/b] The average forum poster probably doesn't have more than a rudimentary, instinctive, and quite likely incorrect understanding of psychology, further skewed by their own biases. As a corollary, they have no grasp of the mechanisms of [b]learning.[/b] Some, particularly the "git gud" crowd, obviously have no idea how learning works. Anything from dopamine response to learning curves and anything in between is trivialized to effort / time put in, with complete ignorance - willful or otherwise - of the underlying mechanisms that actually enable learning. [b]Game theory and game design.[/b] Apart from the occasional game industry insider, most people here have no clue about the underlying principles of game design and the psychology and math behind them. Granted, many seem to grasp intuitively when things like reward loops or balance are broken, and probably have good examples of those working in other games, but neither side can present more than layman arguments to defend their position. For example, how and why outcomes tend to follow a power law distribution even when opportunities seem to be equal. Speaking of outcomes, the one objection I have to your actual post is your implied definition of "result." You seem to be equating it to score or W/L outcome, whereas I would argue that in a loot-based game, the outcome is [i]gear.[/i] Since IB end-of-match drops and bounty progress for pinnacles are tied to your actual performance (which is an acceptable proxy for skill in a very simplified model) then IB is in fact a competitive mode. As such, based on your own arguments, SBMM should be enabled for IB.