Pointing out that, this is something you can do - you questioned it, I stated what would be fair. I was agreeing with you on part of what you were saying but also trying to explain the complexity of statistical figures which is why I like to see the raw data and come to conclusions for myself.
Also try this taking the highest and lowest numbers away doesn't give you the exact same average, however, it may be close I would have to look at the sample number breakdown for each data point and see if we truly have an outlier. Lets use some simple numbers for our friend here? Sample population your test scores (35/100, 42/100, 53/100, 85/100, 45/100). Add these up and divide by 5 - you get a 52 and in the college you attend that translates to an "A+" - good job. Now lets look at the outliers, 35 versus 85. the 35 is how many standard deviations away from the average of 52? In this case it is one. Now look at the average of 52 and compare it to the outlier of 85 - how many standard deviations is that away? (Answer: approximately 2) So in this case, I could make the argument of throwing away just the 85 and keeping everything else in my data set. By doing this, you would have a 41.25 or for you, because we would round, a 41 average - in this case the college would award you an (A-). Now even if I subtract the highest and lowest as I suggested is something you could do but it is not something you have to do - would get this for an average 46.667, or 47 which where you got to school would be a solid "A". And if you still have questions please re-read the post by
itgoeslike. He was providing you with some good data but also explaining to you what the significance of some of the lower value data represents. I am, like all of us here are, suggesting that we may be able to look at these last two years as an anomaly.
I was also pointing out that I don't know all the variables in deciding why some games had lower attendance or higher attendance numbers, but people throw out numbers like they are 7 years old and they can finally count to a hundred. I am suggesting that we have some data but I hinted that we need more. I also hinted at the fact that we need to understand the variables that lead to the outcome of each data point. I also stated that I wasn't looking to get into the statistics in a lengthy complicated discussion. But throwing stats around like a headline marginalizes the truth IMO. With a lager population of data, we may be able to consider a variable by season like who was coach for that particular season.
Ever gamble?
Here is one, like black jack, roulette, or Craps, well of the three which gives you the best odds. Answer: Craps. - but look it up yourself. Point being, like as in metaphor, these are just games, and I can tell you that when you start out, you have better than a 50/50 chance at winning at the Craps table (roll a seven or an eleven - you win, and so long as you roll these you can keep wining, but role something else, and you need to re-roll that number without rolling a seven or an eleven otherwise you lose. Now using stats, roll a seven, what is the likelihood that you roll a seven again - it is less than 50% see each time you roll the odds of you rolling that combination go down each time. So you need to know the data - the variables, the first time you roll a seven the odds are statistically in your favor, but the next time they are not. This is what I was alluding to when it comes to attendance. What are all the variables??? I could go on, but I wouldn't want your head to explode.