With that said, how can you allow there to be such things, for :months: on end, like the stomping ground of early T-34s? In 51 games in the T-34 '41 I lost a single one, and I have many friends who have similar stats. Obviously, with the advent of the British tree things have changed, but it's these glaring inconsistencies that allow people to assume other things such as temporary imbalance, or even people just having a bad experience with a type of vehicle, are bias when in reality they aren't.
Surviving and winrate is one of the stat, Kill/Death ratio, amount of Hits/critical hits are example of different stats.
Also, it requires to understand what exactly the match/kills/deaths were in particular case - it may happen, that some of rivals were from prior BRs.
It is really not possible to answer such question - as I haven't personally seen even one of those said 41 battles. It can be matchmaking, luck, cautious play, skill, and even that this is exaggeration. From my experience I am usually better than other players in both my and enemy team, very rarely not in top3 in match, but that is happening on both US and USSR tanks (I don't play German tanks). But any single user is not statistics, and even statistics differs between average player (that is averaging best and worst players) and best players (say, top 100k players overall).
It is very important to rely on statistics and it is very important to "reality check" statistics with your own experience.
The last thing provide great difference in both attitude towards certain vehicle and ways to balancing vehicles. Naturally, most of the games (and game developers) are trying to make game mpre friendly (i.e. balance the game in case of multiplayer) to average player, not to most hardcore players. The reason is that hardcore players will exploit all weaknesses/strengths anyway, that is what make them hardcore.
Unfortunately, people will assume bias, it is unavoidable. In any competitive game hardcore gamers assume developers to be kinda evil (and/or stupid), partially because any game is usually not 100% focused on most hardcore players, and partially because it is easier to blame developers, than mistake and better competitors skills.
But in reality there is completely NO point in having ANY bias, it can't be positive for anything. Each developer can have ego issues or some bias, it's human nature, but rarely they share one same bias and developers as a whole, as a company, can not benefit anyhow from such bias, especially it is hard to benefit from contradicting biases (i.e. both Russian for English community and US/British/German for Russian community), even if it would be possible to have both (it is actually somehow possible, if we consider pure evil developers. We could, for example, nerf/up vehicles based on a language selected in game).
Even if someone is "creating" vehicle to be more overpowered, than it should be, than there will be BR adjustment, which is done by other people. Imagine, you are responsible for adjusting BRs - how could you persistent justify incorrect value, not based on actual statistics? Well it could be done for some initial period of time, when statistics is not definitive, but afterwards it would require conspiracy with almost all management, and conspiracy require evil will, not only evil, but also stupid - i.e. all management should think that it makes more sense to have such bias than to have more benefits, such as earn more money or popularity or both, which is clearly even stranger than to think that the whole world is secretly controlled by some masons for their benefit.
(Of course, there is also no point in being stupid, but usually we, humans, believe we are smarter than all other people, while most often that belied is due to lack of knowledge. So I would rather not discuss it.).