Also, mathematically this is very hard to explain. I know we're not talking precisely about coin tosses, but for the most part the game, starting with two players using equal salary, is largely a 50/50 proposition. A "good" season is 90 wins, which means winning five of every nine games throughout a full season. WWLWLLWLW. It's hard to pick out patterns when going 5-4 is "crazy winning" (.556 win pct).
Statistically, with coin tosses, there will be hot streaks and cold streaks. The gambler's fallacy is the desire to bet against a streak in the mistaken belief that a streak is unlikely to go once more in the same direction, when in fact the odds are still 50/50 for each toss.
The observation here is that hot streaks and cold streaks at least APPEAR often to be balanced out by "corrective moves" in the opposite direction. If a hot streak like 20 straight happens once every 2,000,0000 trials, I can handle that. But back to back 1 in 2 million shots??? That is hard to explain.
Unfortunately, one of the few things that WOULD explain it happens in the black box of the game engine.
Let the conspiracy theory arguments begin.