9 or 19? | Sports Information Blog

Two Division III men’s soccer polls came out yesterday. In the one released by the National Soccer Coaches Association of America, Muhlenberg was ranked No. 19, and in the one released by D3soccer.com, the Mules were No. 9.

How could there be such a difference? One reason is that it’s still early in the season, and it takes a little while for the teams to sort themselves out. Another is that the two polls are conducted in very different manners, both with their quirks.

The NSCAA poll sticks very closely to the regions. This makes sense because Division III competition is regional in nature, and there are few inter-regional games during the season from which to form a basis of comparison. Teams are ranked 1-10 in each of the eight regions, and in general the top teams in each region are 1-8 in the poll, the second teams are 9-16, etc.

There are some exceptions: The No. 2 in the Mid-Atlantic on the women’s side, Johns Hopkins, is ranked No. 7 nationally, ahead of two regional No. 1s. By the way, the Mule women’s soccer team plays at Johns Hopkins on Saturday night.

The down side of this system is pretty obvious – teams jump around wildly in the national poll based on their regional ranking. Four teams that were ranked below the Mules last week are ahead of them this week, not because voters suddenly think they’re better, but because they moved up to No. 2 in their regions. As long as it is No. 3 in the Mid-Atlantic, Muhlenberg won’t be able to move up much past No. 15 or so.

The D3soccer.com poll works more like a national poll. And with national polls, it’s sometimes hard to break in, but then once you’re in it can take a while to fall out.

Muhlenberg was ranked 15th in the preseason poll, which was based in large part on last year’s final poll. Since the preseason poll, teams ahead of Muhlenberg have suffered losses that caused them to drop, while the Mules have inched up from 15 to 12 to 9 with their five straight wins to start the season.

This poll is not without its oddities. When a team loses, voters are often inclined to move it down only so many spots. Here’s an example from the women’s poll: Last week, Rochester was ranked 13th and Union received votes. Union beat Rochester, 2-1, but this week Rochester (5-1, No. 16) is still ranked ahead of Union (6-0, No. 17). The NSCAA poll has it more in the way you’d expect based on that result: Union is No. 2 in the region and No. 11 nationally, while Rochester is No. 3 and No. 18.

So are the Mules really No. 9 or No. 19, or none of the above? Only time will tell. Two things are for sure, though: the polls are fun to watch, and it’s better to be in them than out of them!

Great blog post! You nailed the difference in the polls.

I have little to add to what you say about the NSCAA polls beside to suggest that their national ranking should NOT be called a poll. Their polling happens regionally not nationally. When they throw all the regional #1′s into the national top eight and so on, they are creating a national ranking without polling on a national basis. Thus it’s a misnomer that the NSCAA calls it a national poll. It’s a national ranking, and a poor one because of the lack of national polling.

The D3soccer.com Top 25is a true national poll and thus has the potential to be much more accurate and relevant. As to the “oddities” of the D3soccer.com Top 25 poll, I agree that voters seem to actually under-react to new results which does create some real head-scratchers at times. On the surface of it, I don’t find the specific example you mentioned to be so odd because the evaluation of games needs to take into account much more than the mere score and two teams’ relative position needs to be based on their full seasons not just their isolated head-to-head result. However, I do think there have been plenty of other better examples of voter unwillingness to acknowledge what the latest week’s results indicate. Since there is (a) no real national exposure to allow voters to see most of the teams with their own eyes, (b) minimal inter-regional play, and (c) minimal head-to-head play among top teams, doing a national rankings requires a fair amount of guess-work and voters would do well to remember and acknowledge that as it would free them to be more responsive to new results and move teams more from week to week.

In closing, both rankings have flaws. However, the NSCAA flaws are inherent and thus fatal. On the other hand, the D3soccer.com flaws are not systematic but rather rest with the voters’ approach and thus the ranking can improve.

One thought on “9 or 19?”

Flying Weasel on October 4, 2011 at 2:31 am said:

Great blog post! You nailed the difference in the polls.

I have little to add to what you say about the NSCAA polls beside to suggest that their national ranking should NOT be called a poll. Their polling happens regionally not nationally. When they throw all the regional #1′s into the national top eight and so on, they are creating a national ranking without polling on a national basis. Thus it’s a misnomer that the NSCAA calls it a national poll. It’s a national ranking, and a poor one because of the lack of national polling.

The D3soccer.com Top 25is a true national poll and thus has the potential to be much more accurate and relevant. As to the “oddities” of the D3soccer.com Top 25 poll, I agree that voters seem to actually under-react to new results which does create some real head-scratchers at times. On the surface of it, I don’t find the specific example you mentioned to be so odd because the evaluation of games needs to take into account much more than the mere score and two teams’ relative position needs to be based on their full seasons not just their isolated head-to-head result. However, I do think there have been plenty of other better examples of voter unwillingness to acknowledge what the latest week’s results indicate. Since there is (a) no real national exposure to allow voters to see most of the teams with their own eyes, (b) minimal inter-regional play, and (c) minimal head-to-head play among top teams, doing a national rankings requires a fair amount of guess-work and voters would do well to remember and acknowledge that as it would free them to be more responsive to new results and move teams more from week to week.

In closing, both rankings have flaws. However, the NSCAA flaws are inherent and thus fatal. On the other hand, the D3soccer.com flaws are not systematic but rather rest with the voters’ approach and thus the ranking can improve.