I will say this now for the record and you can reference later on in the season. I will not be surprised if the Bills end this year in A territory, how high can they go I have no idea. I think the limit of the move allowed by the system to 2 steps up or down is basically placing a limit of 2 std deviations up or down. This, of course, and provided a normal distribution of all possible outcomes (which is not necessarily always the case) fully encompasses 95% of all possible outcomes. This is basic statistics after all.
I mention this because this leaves 5% of all possible outcomes going outside the limits both up or down. Again this assumes a normal distribution of all outcomes. However, as it happens in the market, the skew of the distribution curve does not necessarily follow the "normal" distribution. In truth a normal distribution curve is just one specific case in the family of distribution curves. Amplitudes, skew and kurtosis may vary.
What I mean to say is that sports events have enormous numbers of not easily controllable variables involved. These will affect the shape and location of the overall distribution curve for any particular year. The change does not have to be marked, but it will be there. So, what I would be interested in is the 5% of all outcomes left out of the curve by the 2 std deviation limit. Can we go farther up or down than the 2 std deviation limit? My answer is yes, although doing this would not be a common event and require a special set of circumstances.
Therefore allow me my own fantasies about the Bills ending the season as an A team (-, +, or just flat A) by the end of the year.
I mentioned the market earlier in this post. Take the SP 500 as an example. When the SP gains money for a year, the distribution curve skews to the right with the amplitude and kurtosis remaining the same. When the SP loses money for a year, the distribution curve skews left with the amplitude and kurtosis remaining the same. This will work exactly the same way with your own portfolio. But now, here comes the interesting part. Depending upon the degree of diversification of your portfolio, you will approach the amplitude and kurtosis of the SP distribution curve as the degree of diversification of the portfolio increases. On the other hand if you decide to develop a non diversified, concentrated portfolio, the distribution curve for the year will be a lot flatter (less amplitude) than the SP's, and the degree of kurtosis will change to give a much wider distribution curve with much larger tails both up or down. You may not have heard this ever before, but it is the way it actually works for the market in the real world. Modern portfolio theory aims to highly diversify portfolios according to a complex set of rules, creating portfolios with characteristics approaching the SP's. This will make your portfolio approach, but never exceed, the performance of the SP. On the other hand, if you concentrate instead of diversify your portfolio you will be playing for longer tails in its distribution curve, which will allow greater gains or greater losses than the SP's for any given year.
Most of you may find this pretty boring but I find it fascinating, both as it applies to the market and as it may well apply to sports. A winning team will see its performance distribution curve skewing to the right, a losing team will skew to the left. Again, in order to beat the system, a team has to promote changes that will give a distribution curve of performance results with long, fat tails. I happen to think this is what is likely to happen with the Bills this year, allowing us to exceed the 2 sd limit of your model and get our performance level into the 5% tail area of the distribution curve.