Quote from: Dave 'd-mac' McHugh on March 10, 2018, 03:54:00 PM
Scranton with one loss and .589 SOS was going to host over a 4-loss and .658 SOS from Tufts (at some point, the SOS number is just too... gaudy when you have multiple losses).

This is 2018.  We are in the era of data science.  Evaluating teams based on SOS and winning percentage is not only old fashioned, but it is prone to bias and confusion. 

NCAA decision-making should never be based on these metrics.

Mathematically, these metrics have big problems, including: 

• When an elite team beats a low-quality team, the elite team's SOS goes down although the win tells us nothing new about the team.  What if Bowdoin played another 30 games this season against teams that never won a game?  Suddenly, its SOS would drop, and its near-perfect record would lose its meaning -- for no good reason.
• Another problem with using SOS and winning percentage is that doing so requires analyzing two numbers.  There is no good way to compare a team with a high SOS and low winning percentage to a team with a low SOS and a high winning percentage.
  

We need a single metric that estimates a team's quality.  There are many algorithms out there that do this.  One example is the Elo rating system used for chess: https://en.wikipedia.org/wiki/Elo_rating_system.  Chess and basketball are different in important ways -- such as that chess doesn't have seasons -- so Elo wouldn't fit basketball perfectly.  But there must be a good solution out there.

Quote from: amh63 on February 28, 2018, 11:41:50 AM
Nice class year...fine year for Amherst sports, if memory serves.

2007: 

• The year Coach Gromacki started at Amherst.
• The first year the men's team won a national championship.  Lost forever are those Andrew Olsen no-look passes and half-court alley-oops. The closest I have returned to that great day in Salem vs VA Wes was when one of the '07 captains -- now a doctor -- spilled a drink on my wife at reunion last year. I'll take it.
  

It was an important year for Amherst basketball.  Can't speak to other sports.

Quote from: Dave 'd-mac' McHugh on February 26, 2018, 10:36:36 PM
Yeah - I don't see how a four-loss Tufts is "ranked" ahead of Scranton. Even other data I don't think changes that. Scranton appears to be the higher seed in that grouping.

Losses to Amherst shouldn't count.   

Quote from: Panthernation on April 01, 2013, 04:48:45 PM

Quote from: amh07 on April 01, 2013, 04:27:59 PM
It is not a coincidence that those exposed to the real world embrace fuzzy logic whereas those who are not don't.

Why do so many of you act like you know what the real world is and we don't? You know nothing about our lives. We never talk about your lives; we talk about your arguments.

Sorry, that was a cheap shot because I don't actually know to what extent you are in the real world.

But it was an attractive shot because I remember how I thought in college.  I realize I felt that every question could be answered convincingly with the application of reason.  I also remember how older people would tell me, "Don't think you know so much."  And I didn't really understand that until I entered the real world, made investments based on ideas I arrived at using valid reasoning, and failed.

Quote from: Panthernation on April 01, 2013, 04:41:13 PM

Quote from: amh07 on April 01, 2013, 04:27:59 PM
Epistemology is a really interesting thing.

By definition, arguments using formal logic (or, not, etc) are sound only when everyone knows their propositions (their underlying Ps and Qs) are true.  In sports, there is usually disagreement on the truth of the arguments' propositions (e.g. that PER is an accurate measure of player value).  Therefore, in sports, formal logic is unable to produce a sound conclusion (one that every reasonable person would agree is correct).

That is why fuzzy (or probabilistic) logic is a better fit for sports (and the real world).  We make observations and use Bayesian reasoning to identify propositions that are probably true.  Then, using fuzzy logic, we can draw conclusions that may or may not be true.  Fuzzy logic is humble; it admits that it can't reach conclusions that everyone will agree with.  That is why it (and not formal logic) is suited to sports.

It is not a coincidence that those exposed to the real world embrace fuzzy logic whereas those who are not don't.

The law of non-contradiction is formal logic. Without it, our discourse can't get anywhere because the opposite of any truth claim can be simultaneously true. Why confine its application to rhetoric and math?

Lefrak made a simple proposition that when many well-informed people believe something is true, that fact increases a rational person's expectation of the truth of the thing.  This is a valid argument using Bayesian reasoning and fuzzy logic.

I bet that if you named any supposed self-contradiction that someone made on this board, then I could argue using formal logic that it is not a self-contradiction.

In the fuzzy real world, Barack Obama is a great president and a terrible president.  Aaron Toomey is the best player and the second best player.

Quote from: Panthernation on April 01, 2013, 03:38:35 PM

Quote from: amh07 on April 01, 2013, 03:13:59 PM
Applying formal logic to the real world is like eating jell-o with chopsticks.  You will look fancy, but you won't pick up anything.

Formal logic is only good for rhetoric and math, not sports.

Yes. Sports dialogue would be well served if no argument could be checked by any objective standard. Unfortunately, too many sports fans feel this way. (As Old Guy put it, the "your team suck's worse!" variety.)

Epistemology is a really interesting thing.

By definition, arguments using formal logic (or, not, etc) are sound only when everyone knows their propositions (their underlying Ps and Qs) are true.  In sports, there is usually disagreement on the truth of the arguments' propositions (e.g. that PER is an accurate measure of player value).  Therefore, in sports, formal logic is unable to produce a sound conclusion (one that every reasonable person would agree is correct).

That is why fuzzy (or probabilistic) logic is a better fit for sports (and the real world).  We make observations and use Bayesian reasoning to identify propositions that are probably true.  Then, using fuzzy logic, we can draw conclusions that may or may not be true.  Fuzzy logic is humble; it admits that it can't reach conclusions that everyone will agree with.  That is why it (and not formal logic) is suited to sports.

It is not a coincidence that those exposed to the real world embrace fuzzy logic whereas those who are not don't.

Quote from: Panthernation on April 01, 2013, 02:10:19 PM

Quote from: lefrakenstein on April 01, 2013, 01:25:13 PM
While formal logical certainly says that agreement and disagreement have no bearing on an argument's validity, back in the real world, when the overwhelming majority of smart, knowledgeable people on a topic disagree with an argument relating to that topic, it's not really a great sign for that argument's credibility.

But formal logic exists for a reason and has bearing on reality. Large groups of intelligent people are often wrong. The topic of evaluating athletes' value is not an exception. That you are appealing to this argument is not a great sign for yours.

Applying formal logic to the real world is like eating jell-o with chopsticks.  You will look fancy, but you won't pick up anything.

Formal logic is only good for rhetoric and math, not sports.

Quote from: Panthernation on March 27, 2013, 01:18:08 AM

To demonstrate that point, in games that Amherst has lost (2) or won by single digits (6) Toomey is a collective 31-94 (33%) and only had one shooting performance that truly contributed to a win (his 34-point performance on 9-12 shooting in the Rhode Island College game).

If eight games is too small of a sample, when the same parameters are extended to last season the trend holds. In single digit wins (6) or losses (2)* Toomey shot 36-102 (35.3%). In single digit wins or losses over the last two seasons, Toomey is a combined 67-196 (34.2%).

*Toomey did not play in one of Amherst's 3 losses last season

Conversely, in games that Amherst has won by double digits this season (21) he is 128-267 (48%).

What was the shooting percentage of players other than Toomey in games when Amherst lost or won by single digits (excluding Kaasila and Pollock)?  Only if it is statistically significantly higher than Toomey's is this argument somewhat persuasive.

Notwithstanding that, any statistical argument claiming a player performed worse than another must compare those players' performance in similar circumstances.  So can anyone do the following analysis?

Compare team points scored for time when Kalema was on the floor without Toomey versus time when Toomey was on the floor without Kalema.  To make the analysis even better (and give a nod to Panthernation's point), you could control for total points scored by Amherst in the game.

My guess is that this analysis won't yield statistically significant results showing Kalema is better.  If it did, Kalema's superiority would probably be obvious to most Amherst fans and to Coach Hixon.

Quote from: amh07 on March 09, 2010, 07:54:01 AM

Quote from: wbbhoopsfan on March 08, 2010, 05:16:39 PM
if she is back and Beryuti is out at Kean, look for Amherst to win the title.  Anyone feel there is a better team? 

I'm concerned about Hannah Munger at George Fox, who is listed as tall as Troy Whittington at Williams (6'5'').  Amherst will certainly benefit from another tall body to take fouls.

Some things come back to haunt you.

The Amherst seniors have been an inspiration to me for the past 3 years I've watched them.  They are role models to me and to everyone else for their dedication, persistence, focus, and confidence.

My impression at an arm's length:

- Lem Atanga McCormick is balance, flexibility, and quiet self-assurance, coolly transitioning to be a 3-point sharp-shooter when circumstances called for it.

- Jackie Renner is the Warren Buffett of D3 women's hoops, smart when others are impulsive, holding her resources in reserve until just that right moment when everyone else goes left and she goes right. 

- Shannon Finuncane is the epitome of emotional intelligence.  Her rapid low clap will invigorate anyone, not to mention her steals and her going to the floor.  Did you see her hand on Kim's shoulders after Kim fouled out?

- Kim Fiorentino is consistency and focused energy, always doing her job at the highest level even when she isn't the center of attention.

- Caroline Stedman is the picture of mental strength and "deer-like" grace, always leading the team at critical times.  I am at a loss for finding another example of a person outside of athletics who consistently steps up the way she does.

- Livia Rizzo: although I've seen less of her, she is apparently a very likable person, a great athlete, and a great basketball player.

The traits that these women possess are among the most important qualities that a person can have OFF the basketball court.  As they leave the Amherst bubble, the seniors should be aware that these qualities will place them steps above whomever might be their next competition.

Thanks, NH Hoops Fan for pointing out the awesome article.  What an exceptional person.

http://www.ncaa.com/news/basketball-women/article/2012-03-15/atanga-mccormick-shows-many-sides

Thoughts on the George Fox and Amherst match-up?

Amherst with 65 rebounds against Norwich!  Is Amherst a better rebounding team this year than last year?

Megan Robertson the new Jaci Daignault? 

Savannah Holness the new Sarah Leyman?

I was glad to see Wash U take out George Fox yesterday.  For Amherst to take on Hannah Munger in offense and the almost impenetrable George Fox zone defense would be scary (for me).

On the NCAA site, when I click on the Amherst-Williams game, I get the Brandeis Mens game.  What's going on?