NXN At-Large bids? Who get?

Another CA Guy wrote:

Not sure what you're looking at but Saugus ran at NXN last year and finished 12th. Newhall = Saugus.

Doh! You're completely right, sorry. I forgot to translate from "Newhall" in my database. Still have the problem that they finished well back of Xavier and New Trier. (Of course, they can counter argue that they've got a longer, richer history).

SW vs Calif wrote:

Doh! You're completely right, sorry. I forgot to translate from "Newhall" in my database. Still have the problem that they finished well back of Xavier and New Trier. (Of course, they can counter argue that they've got a longer, richer history).

Seemingly a minority opinion in this thread, but I think the only thing that should matter is performance this year. How a team or a region raced at NXN last year or over many years shouldn't affect the selection; attention should be focused entirely on this year's races.

Of course, human nature and committees being what they are, it is certainly possible that past performance could be raised in a discussion of the "most worthy" of potential at-large teams.

Other factors might be introduced as well. For instance, last year I suspect Sarah Baxter's name was mentioned when the final at-large selections were made.

Another CA Guy wrote:

Here are Bill Meylan's speed ratings for many of the girls teams he mentioned and a few others.

Pine Creek - 154 / 126 / 126 / 124 / 114 / 104 / 97 - 129.2 (highest)

Xavier Prep - 144 / 130 / 124 / 122 / 119 / 113 / 112 - 128.8 (highest)

New Trier - 142 / 133 / 132 / 114 / 111 / 105 / 104 - 126.4 (highest)

I believe the ratings under-estimate the NXN-MW race, given the windy conditions (officially at 25-30 mph, but the start finish of the course is on top of hill, and racers ran straight into headwind for both looong straights).

Consider:

1. At NXN-MW, Bloom's US#4 Carmel was rated at only 122.5 and #9 Naperville North at only 122.3 (New Trier was rated 123.9; above must be IL State, although should be 126.6).

2. Ratings at NXN-MW generally are down from each team's state meet.

Naperville N's rating went from 126.7 at IL State to 122.3 at NXN-MW; Mason's went from 126.82 at OH State meet to 119.72 at NXN-MW. (Indiana ran the NXN-MW course, minus hay bales & wind, for State.) These suggest about a 5 point underestimation at NXN-MW.

3. Tully has ratings on 42 girls who ran both IL state and NXN-MW. Exactly one of the 42 got a higher rating at NXN-MW (she ran 30 sec off PR at IL-State, recovered well for NXN-MW). Tossing out 3 girls who obviously bombed at NXN-MW (and keeping in the 1 who bombed at State), the average drop in Tully's ratings was a full 7 points (plus change).

So, I'd boost New Trier from the currently rated 123.9 at NXN-MW to more like 128.

Another CA Guy wrote:

Seemingly a minority opinion in this thread, but I think the only thing that should matter is performance this year. How a team or a region raced at NXN last year or over many years shouldn't affect the selection; attention should be focused entirely on this year's races.

In theory, I completely agree.

However, it's inherently difficult to compare across regions when teams don't compete against each other. Even speed ratings can be biased due to high starting point in some region vs low in others, until the regions compete against each other. For most schools, the NXR and NXN races are the ONLY chance they get to compete against other regions. So, those meets can show which regions have been over-rated and which under-rated, in past years (hopefully that will lead to some corrections, but not sure it does).

@BMeylan Interesting, I heard they were actually adding more races to remove congestion. With the Open races, 350 runners seems near improbable for both the finish line, and the 1000m bottleneck.

Would make for an interesting race, but I am sure many coaches would complain because some of their top talent would simply be slowed down by the compression. Put too many kids in a race, and it is very hard to pass, which dictates pace. Frustrating if a NE team is holding up a NY team.

Will be interesting to see how things shake out. I have been keeping my own record of At-Large potential, but I never like to say anything on the record before all the eggs are in the basket.

The speed ratings for the Midwest girls are accurate and consistent with the methodology applied to all NXN regions.

It is common for many runners to perform at a lower level at post-season races following their State Meet such as an NXN regional race … Along the same lines, I always caveat my NXN National projections by saying that most teams will perform at lower levels at NXN Nationals as compared to races that got them there.

Here is something I wrote in an article last year … "Most top teams around the nation focus on peak-performance at their State Meets (California included) ... But California is the only NXN region where designed peak-performance coincides with the NXN qualifying race ...... The California State Meet is contested on a standard California speed course that is familiar to the runners ... For NXN regional races, many teams have to travel to out-of-State courses which are unfamiliar which results in performances somewhat lower than their respective State Meet performances (not to mention they wanted to peak at their earlier State Meet)."

I first noticed the decline in post-season races when evaluating Footlocker regional races years before the advent of NXN - a few runners have carrier-bests, a relatively small segment race normally, and a larger than normal segment race poorly … that same decline is also apparent in NXN regional races and must be incorporated into the speed rating methodology because it does actually happen.

Thanks for the reply.

You would know better than me if they're consistent across regions. I only looked here. And I can definitely believe some drop-off. But 41/42 runners? That suggests a huge dropoff, and seems unlikely. (2/3 of runners would ring true to me). So, if that's true in other regions, then OK.

Last year, New Trier had higher ratings at Nationals than NXN-MW, by about 2.7 pts. (Carmel by about 0.5 pts). So that also suggested to me the MW course was under-rated. But, again, I didn't check other regions.

From Another MW guy: "But 41/42 runners? That suggests a huge dropoff, and seems unlikely. (2/3 of runners would ring true to me). So, if that's true in other regions, then OK.

Last year, New Trier had higher ratings at Nationals than NXN-MW, by about 2.7 pts. (Carmel by about 0.5 pts)."

You make some good points ... Last year, NXN Portland Meadows was relatively dry and fast compared to most years ... I think that helped the performance level of many runners compared to prior years ... the cold, wet, windy and slippery conditions of Portland Meadows can affect runners negatively, so maybe the better conditions in 2011 were a positive for a change.

"41/42 runners" does seem extreme to me as well ... I have noted elsewhere that speed ratings can never be precise because they are based on data points that are imperfect and variable (namely human beings and their performance) ... "Nationally based" speed ratings typically have a higher margin of error than ratings specific to more local regions (e.g. within a single State) ... The "41/42 runners" might be explained by:

.. (1) The ratings for the Illinois State meet were too high by 1 to 2 points.

.. (2) The ratings for NXN Midwest were too low by 1 to 2 points.

.. (3) A combination of the above (just one point each might fix some of the 41/42)

.. (4) The race was just slow within various segments of the race due to a slow pace for various reasons ... I find this generally uncommon for an overall race, but it does happen on occasion (especially within a segment of the race) ... A few years back, the Fayetteville-Manlius girls went out intentionally slow at the NXN regional and it "skewed" the speed ratings ... The resulting ratings were correct statistically, but a poor indication of the ability of some runners.

So team average speed ratings should be viewed in the context of margin of errors ... not to mention that one runner alone can skew a team average (by being excessively high or low).

All true. The average difference was 7 pts, even throwing out the larger outliers, but not the smaller. Could be your points 1 & 2 (at the high end) plus also the letdown that you mentioned earlier, altogether adding up to 7 pts.

I don't mean to be a big complainer. Believe me, I understand how difficult it is to do what you do, let alone get every single person or meet right. I am being MW-partisan here, and want to make sure the MW teams get "full credit" for their races. If I find time later I'll check some other State-to-region comparisons to see if they look similar

Maybe North Central and those two Kamiakin runners trained through Borderclash? I think their coaches are more concerned with how they do at NXN than a fun race like Borderclash. Kamiakin had 5 runners in front of Hermiston's Alejandro Cisneros at NXN NW but he managed to place top 10.

Cisneros also didn't have a good race at NXN-NW... he was about 30 seconds off of where he should have been, given what he did at state and in previous meets.

Maybe you should take a look again at NXN-NW vs. BorderClash results...

Summit's usual #2/3 finished 9th and 11th overall at NXN-NW.

Kamiakin's #1 finished 13th (re: behind Summit's usual #2 and 3)

North Central's #2 finished 8th (re: one spot ahead of Summit's usual #2, three spots ahead of Summit's usual #3)

So should it be a surprise that Summit's top three were right with North Central's #1/2 and ahead of Kamiakin's #1? No, they were a decent run from their usual #1 away from doing that the week before.

BMeylan wrote:

So team average speed ratings should be viewed in the context of margin of errors ... not to mention that one runner alone can skew a team average (by being excessively high or low).

Just to follow up a bit, with very incomplete comparisons. Again, I don't want to be too annoying - I greatly respect what you do, I'm just trying to add to the available info.

For review, 41/42 IL girls had lower ratings at NXN-MW (average over 7 pts lower, AFTER throwing out the ones who bombed at NXN-MW). But I know I'm being partisan, so looking into other states from MW:

OH runners: I only found 12 (sorry, not from there, and wasn't being exhaustive). All 12 had lower ratings at NXN-MW, average 7.3 pts lower.

MI runners: only found ratings for 3 who did both MI state and MW: One girl got a rating 0.5 pts higher at NXN, the other 2 were 13.5 and 5.4 pts lower at MW than OH. So, that's 57 girls, and only 1 (who at state was WELL off her ordinary time) who rated more than a point better at regional.

Since IN girls run same course at State, I'm not evaluating (Carmel was about equal, ratings wise).

Checking into SW ratings for top 4 teams, plus Monarch, found ratings for their state meets plus NXN-SW for 31 girls. On average, they got ratings 3.93 pt HIGHER at NXN-SW than in their state meets. 20 had higher ratings at the regional, only 11 had higher ratings at State. I was only matching individual girls, not team averages. Surely a few of the girls were down (or running team races) at their State meet, but others were presumably down at regional. But it's a small shift, actually favoring regional over state.

So, if there's a letdown from State to Regional, it seems to disproportionately affect Midwest runners.

I know the IL course is a fast, flat 3m, and this year had perfect weather. Maybe it was over-rated. And maybe OH, I don't know. Or, maybe the MW meet is just a little under-rated.

Key considerations that NXN uses for at large spots:

* Only results from this year's races (not past history)

* Regional races get weighted much more than anything. Early races are not weighed

* Individuals (baxter et al) have no role in team selections

* Rankings are not used. Thank God - Milesplit rankings blow... until they start to copy others.

At NXN Midwest, 5 of the 7 Carmel girls ran at or above their speed ratings compared to the Indiana State Meet ... At the top end of NXN Midwest girls:

1 - Ashley Erba (Indiana) set a speed rating PR

2 - Madeline Perez (Illinois) less than 1 point off her speed rating PR

3 - Taylor Werner (Missouri) 6 points better than her State meet win

4 - Mary Kate Vaughn (Ohio) set a speed rating PR (2 pts better than State meet win)

5 - Courtney Ackerman (New Trier IL) 1 point off her speed rating PR

6 - Bobbie Burgess (Indiana) equals speed rating PR

One factor that I should have mentioned when comparing the Illinois State meet to NXN Midwest is the difference in distance, and more importantly, the speed ... For the Illinois girls State meet, my race adjustment was 144 seconds (I added 144 seconds to the final times) ... the adjustment for the girls NXN MW was 42 seconds ... Major difference ... And looking back, I probably should have scaled the Illinois girls ratings, and that would lower nearly all them ... scaling is usually more necessary for girls than boys, but I only do it 4K races when necessary.

NXN Southwest typically has teams perform at a higher level (speed rating-wise) as compared to their State meets , especially Colorado and Utah ... the Fort Collins CO girls are a good example ... and the NXN Southwest ratings for the American Fork guys would typically put them on the podium at Portland Meadows ... Each region has its own characteristics which need to be evaluated separately.

After doing group comparisons of the five NXN regions that have already raced (comparing them directly to each other), I would not change any of my posted speed ratings ... I believe they are accurate within an error margin of two points.

Thanks for the reply. Not exactly the answer I wanted, but glad you checked. As I said, I respect your work and am grateful for your efforts --I know what a time sink it is!!! I also noticed that the front end of the pack aligned well... but what got me started wondering about this is my daughter, who wasn't near the front. (Not the back either). She ran, relatively, awesome at NXN-MW. Beat absolutely every girl (11/11) who had beaten her by 5-25 sec at IL state (and a few who ran faster but obviously were off at NXN-MW); and all (literally) people who were close behind at IL were 25-30 sec back at NXN-MW. So, great, right? But her speed rating went down.

So, ratings (at least for my daughter) for one of the two races were off... I'm willing to accept it was the IL meet, rather than NXN-MW. (Although my cursory examination of OH showed same effect). I think perhaps there were bigger discrepancies at the middle and back of the pack than up front. This could easily be due to the tough conditions. Those near the front have run many tough races and are going for glory (at both meets); many of the mid-packers are just running hard, and when they hit that long final straight with a 30 mph wind in their face... well, suffice to say I saw some painful final straights. As Bill Cosby would say, "the bear grabbed me, then handed me a refrigerator..."

So I guess I could say my daughter ran tough, but not necessarily fast!

PS Can I assume that when you say you "added xx seconds", that you scale that? I.e., add less to fastest times than to slowest? [I have data on % time diff as well as raw time diffs b/w IL & Midwest ... either way my daughter looks good ;-) ]

SW vs Calif wrote:

Another CA Guy wrote:

Seemingly a minority opinion in this thread, but I think the only thing that should matter is performance this year. How a team or a region raced at NXN last year or over many years shouldn't affect the selection; attention should be focused entirely on this year's races.

In theory, I completely agree.

However, it's inherently difficult to compare across regions when teams don't compete against each other. Even speed ratings can be biased due to high starting point in some region vs low in others, until the regions compete against each other. For most schools, the NXR and NXN races are the ONLY chance they get to compete against other regions. So, those meets can show which regions have been over-rated and which under-rated, in past years (hopefully that will lead to some corrections, but not sure it does).

"In theory, I completely agree," was exactly my first thought. Furthermore, it is impossible for the committee to be unbiased (not intentionally, just human nature). Evaluating teams and region strength is beyond difficult and the current system hasn't been fair in my opinion because it's almost impossible to get it right when comparing teams from all corners of the country. Although not perfect, this is why I like the idea of rewarding regions with success (multiple top 10 team finishes) the year before.

In Cal state, Simi Valley finished =3rd in Division I, almost 100 pts behind Great Oak, 40 behind St.Francis. Buchanan 4th

Saugus (Div II) yet to run

From Watchout's blog (http://watchoutxcrankings.blogspot.com; thank you, sir), here are the regions (from 2007+), followed by # of top 10 ten finishes, followed by number of at-large bids, and my own notes:

California 7, 8 (9 of the 18 teams have finished 15th or lower)

New York 3, 0

Northeast 8, 4 (5 of 14 teams have finished 18th or worse)

Southeast, 0, 0 (highest placing team = 17th, 4 of 5 years SE has had last place team)

Southwest 8, 1 (10 of 11 teams have finished in the top 11)

South 7, 3 (most top 4 finishes, but also 5 15th or lower)

Midwest 8, 4 (at-large teams have placed 4th and 5th, lowest ever 14th place)

Heartland 3, 0 (7 of 10 have finished 14th or lower)

Northwest 6, 0

Perhaps it proves nothing, but it may illustrate that some regions have been unfairly (not intentionally) awarded at-large bids where other regions have been too often overlooked.

California needs to merge, but everyone else is done... who gets the boys at-large?

NW - Summit?

SO - Lewisville Marcus - probably not?

MW - Xavier? - probably

HL - Edina?

SW - Los Alamos?

SE - tie for 2nd, but historically weak region?

NY - favorites advance; gut say no at-large

NE - 1 at-large; Don Bosco too far back for 2nd at-large

CA - looks like it'll be close for 2nd and 3rd with merge; historical at-large preference for CA schools

So, NE, MW, CA each get one. CA is debatable, but almost guaranteed to get 1 at-large. Final bid is between Summit, Los Alamos, Edina. SE looks impressive with tie, but the SE is arguably the weakest region. Xavier and Los Alamos looked similar in terms of regional performances, so I'll go with Los Alamos for final at-large.