Early Look at SOS

In my previous life I worked the longest time (almost 19 years) for one company. This company changed names a few times and at the end it was owned by a German corporation. A HUGE German corporation. I often joked that I hated the Germans but that I had never personally met a German I didn’t like, which was very true. I suppose the NCAA is the same way. I dislike a lot of what the NCAA does but I’ve never had a personal interaction with an NCAA member that I didn’t like. I bring this up because the NCAA changed their statistics website over the summer and there are things that I’m use to that are missing. My contact at the NCAA has been really helpful getting things put back and this made me think of my German story.

Another German story from my previous life was I walked into a meeting where my American counterparts were discussing the floor plans of our new building. The presentation showed a large lab on the top floor with the words, “Eagle’s Nest”. I sort of paused and looked around to see if it was a joke. With no one laughing I again looked around to see if anyone else was looking around while wearing the same look of panic I had to be wearing. Failing that, I sat down, listened for a few minutes and then interrupted to explain the significance of the term “Eagle’s Nest” in German culture with respect to World War II. Soon after they decided to name the labs after colors.

No more German stories today. The NCAA site has a report called “Toughest Schedule” that only recently started working (sort of). It has the Strength of Schedule (SOS) and future SOS listed for each team (currently only SOS is working). SOS is typically used as an indication of which teams have the toughest schedule. With that said, I would be remiss if I didn’t remind you that like with any statistic it can’t be used by itself to tell a complete story. SOS is based on the teams you play won/loss record and we should all know that not all wins and losses are created equally.

With three weeks in the book most teams have had their out-of-conference play completed with the bulk of the teams having started conference this past weekend. With games through this past weekend, the leader in SOS is United States Merchant Marine Academy (SOS of 0.784)! This is actually an excellent showcase on how this statistic can be abused. The Mariners (makes sense) currently have a 2-3 record having played teams in the New York and New England region. These teams are:

  • Rutgers-Newark (Won) Record 8-2
  • Pratt (W) Record 3-5
  • Montclair St. (Loss) Record 5-4
  • Hunter (L) Record 7-2
  • Coast Guard (L) Record 12-1

After I quickly do the math I can see that the NCAA website still has problems because this works out to an SOS of 0.714. Sigh.

Regardless, this example still works because this is still a really good SOS. For comparison sake, I like to see an SOS in the 0.585 to 0.600+ range come selection time. So, we have this team with the “best” SOS in the country and by looking at their opponents we can see that we don’t have teams that are world beaters. Both Hunter and Coast Guard were regionally ranked at times last year but pretty low in large regions. They certainly don’t match up with the teams Cal Lutheran has played to start the year and they have an (unconfirmed) SOS of 0.646. The trouble gets to be that if we were picking at-large teams for the NCAA Tournament, the Merchant Marine would have the SOS advantage over every other team in the nation! And, based on my case studies that I do following selection time, we know that SOS is probably the most important criterion the NCAA selection committee uses! This is why I’ve always advocated a deeper look into the criteria (Art versus the Science of Selection).

Anyway, that isn’t even what I wanted to do with this post. To summarize, I’ve given you not one but two German stories, found problems (again) with the NCAA statistics and shot holes in the NCAA selection process before I’ve written a word on the topic at hand!

What I did want to do is look at the early SOS calculations with respect to our teams in the West Region. We now have to keep in mind that the SOS numbers I list aren’t entirely updated but why should fake numbers stop anything I’m doing.

Out in the ASC our four top SOS leaders are:

  1. UT-Dallas (0.650 SOS) 6-4
  2. Sul Ross State (0.618 SOS) 2-5
  3. LeTourneau (0.571 SOS) 5-5
  4. UMHB (0.565 SOS) 6-1

McMurry, one of two undefeated teams in the West Region has an SOS of 0.500.

We can derive from this that UT-Dallas and UMHB, two leaders in the conference, are playing competitive schedules and still performing well. With this small sample size, you have to remember that every win you get actually hurts your SOS because it’s a loss for your opponent, which feeds directly into the calculation. [See comments below on why this statement is incorrect.] We can also see that Sul Ross State may be better than we thought (picked last in the ASC West) and that LeTourneau may be a tougher team than we thought. McMurry on the other hand may be heading for some difficulty and unless teams they beat start winning (a lot) they will have a very poor SOS come selection time.

Out in the NWC our four top SOS leaders are:

  1. Pacific Lutheran (0.630 SOS) 9-1
  2. Puget Sound (0.594 SOS) 5-5
  3. Lewis & Clark (0.466 SOS) 7-2
  4. Whitworth (0.464 SOS) 8-3

Pacific, a team I picked to be second in conference, has an SOS of 0.433.

So, it looks like Pacific Lutheran is the real deal this year. No surprise. Puget Sound may actually be better than I thought to start the year and then we have a big drop-off in SOS. Lewis & Clark has done wonderfully to start the year but their SOS is saying it wasn’t against winning teams for the most part. I was concerned about Whitworth to start the year and they surprised me to this point but, like L&C, their opposition hasn’t been the best. We’ll see the NWC improve in SOS because they will play each other and their records (for the most part) are fine but it appears that there needs to be an effort to improve the quality of their opponents to start the year.

Out in the SCAC our four top SOS leaders are:

  1. Dallas (0.683 SOS) 6-3
  2. Trinity (0.588 SOS) 10-2
  3. Schreiner (0.570 SOS) 1-12
  4. Centenary (0.564 SOS) 6-7

Colorado College, my #1 team in the nation, has an SOS of only 0.522.

The Dallas SOS shocks me. One item to mention is that I’m not sure if the non-DIII opponents are removed from this SOS calculation. I would think they should be removed but I know Dallas has played two non-DIII teams and a number of their opponents have built up winning records by playing other non-DIII teams. Still, kudos to Dallas. Skipping Trinity for a second and we find Schreiner with a good SOS but you have to remember that those 12 losses feed into a good SOS. More time is needed to see how this shakes out. I wasn’t sure how Centenary would fare this season but this is a good sign for them. The Trinity versus Colorado College comparison is interesting. Everyone would favor Colorado College when they play in a couple of weeks, right? Well, this is a red flag to that assumption. I know those 11 wins Colorado College has hurts their SOS but Trinity has 10 wins.

Out in the SCIAC our four top teams are:

  1. CMS (0.700 SOS) 9-2
  2. Cal Lutheran (0.646 SOS) 6-6
  3. Occidental (0.600 SOS) 3-4
  4. Pomona-Pitzer (0.570 SOS) 3-6

Chapman, a team on the rise this season, has an SOS of only 0.472.

The CMS SOS just shows how impressive this team has been to start the year. That’s 0.700 with no Hunters or Coast Guards thrown into the mix. Cal Lutheran has played a really tough schedule and just need to make sure they get some wins off of weaker teams moving forward. Occidental and Pomona-Pitzer are probably saying we need to hold off on judgment right now until we get more matches and results. With Chapman, we just need to be wary. They’ve won what’s in front of them and that’s all you ask but if we see them start losing some SCIAC matches then we can point to a weak schedule to start the year.

To round things out, UC Santa Cruz has an SOS of 0.471 and Mills College has an SOS of 0.476. Neither teams is over 0.500, which probably means more of the same for these two teams moving forward.

UPDATE – It’s not surprising that my viewpoints on SOS are similar to Ricky Nelson’s viewpoints. He is, after all, my leader. Here’s a post he wrote this morning on SOS and selection.

Advertisements

6 thoughts on “Early Look at SOS

  1. “… you have to remember that every win you get actually hurts your SOS because it’s a loss for your opponent …”

    I’m not sure of the methodology of the NCAA stats SOS, but the opponent winning percentage component in the SOS used for selection purposes excludes the “results against the team in question.”

    This post got me thinking. Slightly different words were used, but I posted the annual plea on my blog.

    Like

  2. Here is the “quality” opponent count by west region team. Not your QWI points but how many teams in the Pablo top 100 you have played. If a school is not on the list they have not played one.

    School / League / quality matches / wins / Losses / pablo
    Trinity University (TX) SOCOLL 10 10 2 7
    Cal Lutheran SOCAL 9 6 6 53
    CMS SOCAL 9 9 2 11
    Schreiner SOCOLL 8 1 12 316
    Puget Sound NOWEST 7 5 5 96
    Texas Lutheran SOCOLL 7 5 7 110
    Texas-Dallas AMSW 7 6 4 15
    La Verne SOCAL 6 5 4 54
    Pacific Lutheran NOWEST 6 9 1 6
    Whittier SOCAL 6 5 5 63
    Centenary SOCOLL 5 2 9 222
    Colorado College SOCOLL 5 10 0 2
    Mary Hardin-Baylor AMSW 5 6 1 14
    Pacific University NOWEST 5 6 4 34
    Southwestern SOCOLL 5 3 5 35
    Austin SOCOLL 4 7 3 95
    Concordia Texas AMSW 4 5 5 90
    George Fox NOWEST 4 5 4 98
    Linfield NOWEST 4 5 5 146
    UC Santa Cruz INDEPN 4 3 10 205
    Whitworth NOWEST 4 8 3 59
    Chapman SOCAL 3 10 1 51
    East Texas Baptist AMSW 3 6 5 118
    Johnson & Wales-CO SOCOLL 3 0 5 338
    LeTourneau AMSW 3 4 5 224
    Lewis and Clark NOWEST 3 6 2 131
    Occidental SOCAL 3 3 4 106
    Pomona-Pitzer Colleges SOCAL 3 3 6 159
    Texas-Tyler AMSW 3 7 1 39
    Whitman NOWEST 3 5 5 124
    Willamette NOWEST 3 4 6 185
    Cal tech SOCAL 2 1 8 291
    Dallas SOCOLL 2 6 3 247
    Hardin-Simmons AMSW 2 8 3 121
    Sul Ross State AMSW 2 1 5 292
    Louisiana College AMSW 1 0 6 357
    McMurry AMSW 1 9 0 60
    Mills ACAA 1 1 10 411
    Redlands SOCAL 1 6 4 165

    Like

  3. I heard back from the NCAA. Non-DIII matches are excluded from this SOS as are the results of the team you are looking at. So if Team A beats Team B, who was previously 9-0 (and now 9-1), Team A’s SOS is 1.000 and not 0.900. The loss doesn’t factor in. If Team C had previously lost to Team B and played no one else then their SOS would be 0.888 (based on an 8-1 record).

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s