Start BCS Problem

• WWRR •
• БАСКЕТБОЛ •
• ХОККЕЙ •
• ТЕННИС •
• ФОРМУЛА 1 •
• CF •
• БЕНДИ •
• ЭКОНОМИКА •

Гостевая

271314@mail.ru

ФУТБОЛ

ШАХМАТЫ

У.Е.НОТА

ХОККЕЙ

ТЕННИС

ЭКОНОМИКА

ОЛИМПИАДЫ

I start BCS problem with:

BCS "THINK TANK" ENTRANCE EXAM by Mr. Hartman.

Question 1: Three computers produce rankings for all Division 1A college football teams. Team A receives rankings of 2, 4, and 6, while team B is rated 3, 4, and 5 (in a different order). No one computer is to be valued more highly than any other. Which team should have the better computer average?

a) Team A
b) Team B
c) Teams A and B should have the same computer average.

My answer to this question - variant d) simply is not present. And what you will reply what area of a rectangular if its width of 20 inches?

Question 2: We seek to rank the strength of schedule (SOS) of each D-1A football team. We agree that our rankings will be based solely upon the won/loss records of our opponents and not upon the score or location or date of a particular game. Suppose that team A's record is 10-1, while team B's is 9-2. We, as outside observers, look at the list of opponents played by team A and by team B and discover that each team played the same set of opponents. Our conclusion about the SOS rankings for teams A and B should be that:

a) A played a more difficult schedule than B
b) B played a more difficult schedule than A
c) They each played the same set of opponents, so each should have the same SOS ranking.

And again I can not answer. I do not know to WHOM have particularly lost team "A". If the team "A" has lost one match to the outsider, and the team "B" would lose two matches - to two leaders it is one estimation. If the team "A" has lost one match to the leader, and the team "B" would lose two matches to outsiders - absolutely other estimation. So my answer is variant d).

Question 3: A golf club has a pitching contest in which each participant strokes two golf balls, one yellow and one blue, toward a flag 75 yards away. The players' initials are on each ball. An accurate measuring tape is available to measure the precise distance each ball is from the flag. In addition, distance markers have been placed on the ground at one-foot intervals from the flag. The shorter the total distance of a player's two balls from the target, the better. Once all contestants have hit their two balls, which of the following would seem the appropriate way to proceed with the judging?

Re-position the yellow balls as follows: move to the 1-foot marker the yellow ball which was closest to the flag, move the second-closest yellow ball to the 2-foot marker, and move the third closest yellow ball to the 3-foot marker. Proceed in like fashion with all remaining yellow balls. Leave the blue balls in their original locations. For each contestant, now add the current distance of each yellow and blue ball and reward the player with the lowest total.
Re-position the yellow balls as just done above. Then re-position the blue balls in the same manner. Now add the current distances and again reward the player (s) with the lowest total.
Why re-position the balls at all?! Simply add the exact distances that were originally established when the balls were struck and reward the low total.

If you are going to define the best ballerina it is not necessary to use methods of measurement heights of a jump accepted in track and field athletics are long also. Though in both cases we speak about height and length of a jump. So, I choose variant d. which was not present in the question.

All finished? Let's look at your results. Did you answer "c" to each of the questions above? If so, I've got good news and bad news. The good news is that, from a logical standpoint, you answered the questions correctly. Congratulations! Oh, the bad news is that you flunked the exam! For to ace this exam, you had to miss every question! In fact, if you had wished to be "in step" with the "logical" procedures used in the original BCS process, you would have needed to answer each of the above questions with "a"!

I shall now give you a short quiz with one questions. This question closely mirrors actual situations that have arisen in evaluating teams in the BCS process. Let's see if you have what it takes to be a candidate for the BCS "think tank"! (Hint: the answer to this question may strike you as being so obvious that you suspect a trap. Just relax, go with what seems inherently reasonable, and you'll be OK.)

Question 1: you have got in court of a primitive tribe. The judge follows laws of this tribe. But from your point of view these laws are absolutely illogical and unfair. The judge at all are not able to count up to ten. Your actions?

a) You try to learn to count and follow them to laws of logic
b) You study their life and will adapt to their laws
c) You study their life and translate their laws on laws of logic

All finished? Let's look at your results. Did you answer "a" to the question above? If so, You can continue attack BCS Formula but You will never get an agreement with it

Did you answer "b" to the question above? If so, You agree with BCS Formula but You will never satisfied with it

And if You select answer c) You let us continue

So, I've got good news and bad news for Mr. Hartman who create this article. The good news is that, he start research this problem. Congratulations! Oh, the bad news is that he use wrong tools to research it.

The main purpose of this website is to objectively and logically address the many egregious shortcomings that have been present in the BCS implementation. Examples of such flaws, which will be more thoroughly discussed momentarily, include the following:

In the BCS formula, the human poll values that are used should be the exact mathematical averages of the pollsters' ratings, rather than the whole numbers 1, 2,..., 24, 25. The current method invariably introduces inaccuracies via a distortion of the pollsters' actual balloting.

Voting of people (trainers, journalists, fans etc.) on my firm belief in general should be excluded in that kind as it is done now. The person simply is not the objective tool for measurement of sports force.

As to more exact account of voices, it about the same what to call for use of nuclear hours at preparation of fried eggs.

The computer component has far too much importance. Each computer "voter" should be given the same weight as a human pollster. Currently a single computer carries the clout of roughly 24 sportswriters!

In my opinion only the computer is capable to estimate teams objectively. The offer to balance voting one person and one computer program at all has no sense. You see changing parameters you easily can create thousand variants of ranking.

The strength of schedule (SOS) calculation is also faulty. At a minimum, it should guarantee that any two teams which play IDENTICAL schedules (that is, the same slate of opponents) are assigned the same SOS rating. Presently this is NOT the case!

The strength of schedule is automatically taken into account at calculation of teams ratings. The separate account of this factor results simply in the double account of it.

The dubious "quality wins" factor that was added prior to the 2001 season has the potential to actually HURT a team's rating, rather than reward it, for defeating a "quality" opponent! Like all the other examples listed above, this too is extremely easy to correct.

Any additional amendments only lower objectivity of ranking when the question is the championship. Only results of matches in this case should be taken into account.

The initial algorithm used to average the computer rankings was so ill-conceived that individual rankings of 1, 2, and 3 would not yield an average of 2!

Above it was already spoken that ordinal numerals it is impossible to average the same as both quantitative. And the result of "averaging" of ranks 1, 2, 3 is not necessarily equal 2.

To learn much more about the BCS system and its problems, please begin your tour. Well, let us continue revision of Mr Hartman's article by clicking here