NESCAC 2014

[Becks]

Comments on Sat, Oct 11 games:

Hamilton 1-0 Bowdoin - Upset win by Hamilton, but stats showed they deserved it. Hamilton outshot Bowdoin 16-10 and out SOGd them 7-3.

Trinity 2-1 Middlebury - A bit of an upset, but this time, the stats indicate that Middlebury was the better team on the day and Trinity was lucky to win. Midd outshot Trinity by 20-7 and out SOGd them 6-3.

Amherst 3-2 Colby (OT) - Unexpectedly tight came. Stats show that Amherst deserved the win and that the game was more defensive than the score suggests. Amherst outshot Colby 13-5 and out SOGd them 7-5.

Williams 4-0 Wesleyan - A bigger margin than expected and stats show the margin was deserved. Williams outshot Wes 21-2 and out SOGd them 12-2.

[Jump4Joy]

Becks, as usual, helpful info. Thanks.
I do note that Midd's loss to Trinity is another outcome based on "luck" when they've been the "better" team but lost. The same was said when the Panthers lost to the Jeffs at the season start. When does luck weigh less and underperformance weigh more? 

[jc2]

I would say that looking at how many shots a team takes does not tell the complete story. There are quality shots and there a prayer shots. In the case of Middlebury, I suspect we might be seeing the impact of the graduation of their All American GK. Midd won a lot of one goal games last year and this year have been on the losing end of a lot of one goal games. On the other hand, maybe they were just  lucky last year and not so much this year.   

[Becks]

I would say that looking at how many shots a team takes does not tell the complete story. There are quality shots and there a prayer shots. In the case of Middlebury, I suspect we might be seeing the impact of the graduation of their All American GK. Midd won a lot of one goal games last year and this year have been on the losing end of a lot of one goal games. On the other hand, maybe they were just  lucky last year and not so much this year.

Stats suggest that poor shots and weaker goalkeeping "may" be factors in Midd's struggles. Last year, Midd had a team save's percentage of .886 in league games (3d best in the league behind Amherst and Wesleyan). This year it is .692., which is worst in the league. This could be the result of some combination of weaker goalkeeping and bad luck. Just as bad, however, Midd's league opponents' saves percentage is a phenomenal .915, which is also worst in the league (last year it was .829), which could be the result of some combination of taking worse shots (eg more shots from distance) and bad luck. I note that luck can play a big part because when we are talking about only 5-15 goals, a lucky couple of goals can shift that stats quite a bit.

[jc2]

Excellent points as usual.

Of note, in NESCAC games,  Midd takes the 2nd most shots per game (16) but has the worst conversion rate scoring on only 6% of their shots. I would argue that this is more likely the result of poor shot selection than bad luck. It might explain why statistically they appear to "dominate" a game (by taking more shots) but end up losing. I still contend that putting a lot of stock in the total number of shots can give a false impression about whether a win or loss was "justified".

[Ocean 1]

amh63: great dedication to wind your way up to Midd to support the Jeffs on the gridiron!
To add to Amherst's success yesterday: field hockey also claimed victory.
Interesting that a struggling Trinity (winless in conference) took top-o-the-table Bowdoin to OT. Maybe weather was a factor.
Way too much schedule left to speculate on the qualifying 8 for the 'CAC post-season. Meanwhile, the Polar Bears have home field advantage in their crosshairs. Can Williams blemish their perfect conference record?

As we know now, Williams did indeed help to blemish the Polar Bears conference record...although not without one heck of a performance by both teams.

While the Ephs SOG dominated the Polar Bears by a 2-1 margin, the real story was in the OT periods where the Ephs outshot them by a 3-1 margin (i.e., 9-3) and the game winner in the 108th minute.

It was one of the best 'CAC games I've seen in recent years and the Polar Bears "D" was as advertised...they just ran up against a very tenacious offense...and what is turning out to be a pretty good defense as well now that they made the adjustment from a 3-back to a 4-back scheme.

[Ocean 1]

Anyone see the Mid v. St. Lawrence game yesterday...NESCAC v. Liberty League?

Wasn't expecting much from St. Lawrence but ultimately realized they play a pretty good brand of soccer in the Liberty League where William Smith is typically head of the pack. Bennett Rankings now has the Liberty League as the #2 power conference with NESCAC following at #3.

Based on how BR has rated teams compared to Massey, I am now a much bigger fan of Massey. I think it is much more nimble and more quickly reflects the strength of teams in any given year. Heck, BR still has Wesleyan and Trinity ranked ahead of Conn, which is ridiculous. Based on Massey rating, NESCAC is still clearly stronger than the Liberty league. The median NESCAC team (Middlebury) has a Massey rank of 344, while the Median Liberty team (Union) has a Massey rank of 452. Massey has the UAA as clearly the top D3 women's conference, with a median rank of 299.

Both BR and Massey were predicting a 1-goal Midd win over St Lawrence and, based on the game stats, that or a tie would have been a fair result. Midd outshot St Lawrence 17-11 and out SOGd them 7-5.

Becks...your point about why BR would have Conn ranked below Wes and Trinity was well taken. In fact, today BR has Trinity (6-4-1) ranked #52, Wes (1-8-1) ranked # 74 and Conn (10-1-0) ranked #89.

Seemingly, the answer lies in part with Bennett's explanation today when they introduced their Offense and Defense Rankings and methodologies where Strength of Schedule (SOS) plays a major role in the derivation of ranking position. Interestingly, SOS for Trinity is 38, Wes is 13 and Conn is 60. Apparently, despite Conn's superior W/L record, SOS plays a major role and appears to be a key reason for their #89 ranking. Would be instructive to understand to what degree Massey utilizes SOS in their rankings.

An explanatory excerpt from the Bennett site today:

"This week, we're excited to introduce BennettRank Team Offensive and Defense Rankings. We've calculated BR-O and BR-D in the background all season, but it wasn't until this week they stabilized enough for us to share. Our hope is to provide players, coaches, and fans with a metric deeper and more accurate than a simple goals for/against list that they can use to preview matchups, understand opponents, and audit their own team's performance. As with everything we do, we want these rankings to be as fair and objective as possible, so in the spirit of transparency, we figured we should take the time to explain how they work.

The calculation starts with goals scored, for and against. Scoring goals will help a team's offensive ranking, allowing goals will hurt their defensive ranking, as you might expect, but only to a point. Any goals scored beyond four won't help the team scoring them or harm the team allowing them. We figure a blowout is a blowout. After four goals we know an offense is good, and a defense is bad, so maximum awards and penalties are handed out. Any goal scored beyond the fourth is like pouring another cup of water onto a soaking-wet towel — it's not going to get any wetter. Call it the victory saturation-point.

We then take the raw goals scored/allowed numbers and modify them with opponent BennettRank. Thus, BR-Offense and BR-Defense are Strength of Schedule-adjusted rankings. Like the classic BennettRanks, BR-O and BR-D will rank each team from top to bottom across the division; e.g. 1-332 for D1 womens soccer.

We make no apologies for weighting Strength of Schedule heavily in our algorithm. Goals scored against a strong team are more impressive than goals scored against a weak team. Goals allowed to a strong team are more understandable than goals allowed to a weak team. Strength of Schedule is a major component of our overall ranking, and will be a major component of our offensive and defensive rankings as well.

Most (if not all) BR Top 25 teams have a reasonable balance of offense and defense. It is rare for a team to reach the BR Top 25 without that balance. As the season progresses, look for these rankings to stabilize as we see these teams prove who they are with more data."

[Becks]

Ocean1 - The only explanation I can see for BR ranking Trinity and Wes above Conn is that their algorithm is still giving weight to prior years' results. I think Massey's algorithm flushes those out pretty quickly. I don't see how SOS explains it, because if you eliminate non-league games, the teams' SOS's are probably quite similar and Conn's results have been much better. BTW, Wes's SOS is due to plummet -- they play their one real cream-puff game of the season mid-week - Elms (BR #309). Worth noting that Massey only has them as 2-goal favorites. [insert emoticon of guy bugged-eyed in horror]

[jc2]

Ocean1 - The only explanation I can see for BR ranking Trinity and Wes above Conn is that their algorithm is still giving weight to prior years' results. I think Massey's algorithm flushes those out pretty quickly. I don't see how SOS explains it, because if you eliminate non-league games, the teams' SOS's are probably quite similar and Conn's results have been much better. BTW, Wes's SOS is due to plummet -- they play their one real cream-puff game of the season mid-week - Elms (BR #309). Worth noting that Massey only has them as 2-goal favorites. [insert emoticon of guy bugged-eyed in horror]

BR claims they make a mid-season adjustment to account for teams whose performance is significantly different than the year before. Seems like Conn would get a boost by this adjustment. Anyone know if this has happened?

[Becks]

BR claims they make a mid-season adjustment to account for teams whose performance is significantly different than the year before. Seems like Conn would get a boost by this adjustment. Anyone know if this has happened?

BR says they make the adjustment between weeks 5 and 6 after about 8 games have been played. http://www.bennettrank.com/collegesoccer/d1-womens-soccer/mid-season-one-time-adjustments/ We're past that now and the only team that looks like it might possibly have gotten an adjustment is Trinity, which dropped 25 places between weeks 6 and 7, but it's possible that that was just a reflection of their loss to Bates.

[jc2]

Anyone know how the NESCAC all-conference awards are determined? Who votes? What criteria are used to select the All NESCAC teams, player of the year, coach of the year, etc? Is there a formula to distribute the slots among the teams based on their winning percentage (the team with the best record, probably has the best players)?  Is the coach of the year the one with the best team or are there other criteria?

Any insight?

[Becks]

Anyone know how the NESCAC all-conference awards are determined? Who votes? What criteria are used to select the All NESCAC teams, player of the year, coach of the year, etc? Is there a formula to distribute the slots among the teams based on their winning percentage (the team with the best record, probably has the best players)?  Is the coach of the year the one with the best team or are there other criteria?

Any insight?

I think the coaches nominate players on their team and all the coaches vote. Successful teams generally get more of their players on the all-FCIAC teams, but I don't think there is any formula. Coach of the year generally goes to the coach of the top team or at least one of top teams. I've never seen it go to a coach who took a bad team and made it respectable. Player of the Year is a bit of a misnomer because it seems to go to the senior who has had the best college career, even if their senior season was rather mediocre.

[Ocean 1]

Anyone know how the NESCAC all-conference awards are determined? Who votes? What criteria are used to select the All NESCAC teams, player of the year, coach of the year, etc? Is there a formula to distribute the slots among the teams based on their winning percentage (the team with the best record, probably has the best players)?  Is the coach of the year the one with the best team or are there other criteria?

Any insight?

I think the coaches nominate players on their team and all the coaches vote. Successful teams generally get more of their players on the all-FCIAC teams, but I don't think there is any formula. Coach of the year generally goes to the coach of the team or at least one of top teams. I've never seen it go to a coach who took a bad team and made it respectable. Player of the Year is a bit of a misnomer because it seems to go to the senior who has had the best college career, even if their senior season was rather mediocre.

Concur with Beck's assessment. Also, player stats have much to do with the choices and impact of players on their respective teams. Offensively, total points in a given season and/or career is a clear metric. Defensively, it gets a little nebulous but GAA gets a team recognized and from there the coach's make their individual cases.

[jc2]

Game of the week looks to be Bowdoin v Conn. Massey is projecting a tie.  Any thoughts or predictions?

[amh63]

Amherst got a little more experience playing on turf last night at home under the lights.  Amherst won over Keene State 4-0.....led at the half 2-0.  Keene state had good possession time I n the first half and was aggressively attacking late in each half...info from announcers since I was switching between contests...the other being the field hockey game...away game.  A large number of young players inserted by Amherst in the last eight minutes.  Keene State is the defending LEC and went to the post season last season...is in 2nd place in the LEC presently behind UMass-Boston.  Only lost in conference is to ECSU who beat Amherst last week and beat another CAC team last night in an away game.  Amherst has two critical games remaining...CONN and WES that will impact seeding.  Cardinals this Sat in Middletown.