Page 1 of 1

D-1 Outdoor best PV height histogram, by sex

Posted: Mon Mar 30, 2015 11:18 pm
by Divalent
[Edited to add: I did redid with whole country (not just east regional) and I did change the permissions so that everyone should be able to see. So the link below is to the new copy. Let me know if you have trouble accessing.]

I think you might find this interesting. From the TFRRS website I copied the PV data from the 2014 D-1 outdoor data base then bin-sorted to create a histogram.

The attached image is a histogram of the best height achieved by virtually all D-1 vaulters in the 2014 Outdoor regular season (prior to the NCAA regionals). Shown are both men (blue) and women (purple).

Note: I truncated best height down to the nearest 0.05 meters (5 cm), and the data point represents the number of vaulters whose best vault that regular season was between that height and 0.04 meters higher. Also, the best and worse 10 or so data points for each sex are the actual count for that height, but between the two extremes I did a box average of 3 to reduce the noise.

By and large the difference between Men and Women is about 1.1 to 1.2 meters (a bit more than I would have expected)

[It appears that uploading of any attachment is not allowed in this forum (I was informed of this only after separate attempts were reject for the wrong file type and then for too big of an image). So on to plan B ...]

Hoping this link will do it: https://drive.google.com/file/d/0B-oBpKKKOJQpMkgxVW9tYTFESjQ/view?usp=sharing
(it seems to work in preview mode)

[Edited after KirkB's post: I changed the settings on the file to be viewable for all. So it should work now. Sorry about that.]

Re: D-1 Outdoor best PV height histogram, by sex

Posted: Tue Mar 31, 2015 12:41 am
by KirkB
Divalent, I'm very interested in your analysis, but your link popped up this Google Drive page:
You need permission

Want in? Ask the owner for access, or switch to an account with permission. Learn more

You are signed in as [my-addy]@gmail.com.

Request access Switch accounts

So I clicked the Request access button. I expect that it will send you an email asking you to allow me access.

While I'm sure that you won't mind doing this for me, you probably want to set the file to be public. And (unlike myself) I doubt that everyone has a Google Drive account.

If you set the file to public read-only, anyone should be able to read it without a Google Drive account, and without you having to give each person access (time consuming, and causes delays). I haven't tested this, but I assume it will work. Be careful not to grant write access to your Google Drive account to anyone --- all you want them to do is look at your file.

Becca, can you fix the issue of not being able to upload attachments on PVP? Then we wouldn't have this hassle. (In another thread, PVStudent has experienced this too.) Thanks.

Kirk

Re: D-1 Outdoor best PV height histogram, by sex

Posted: Tue Mar 31, 2015 7:23 am
by IAmTheWalrus
Cool graph, and nice work!

Do you think it would work better as a summed graph? I think it's a little weird where the graph dips down around 17' or so and then spikes up again. Clearly anyone who has jumped 17'6 has the capability to jump 17', and so I think that if a data point counted for its own height as well as every height below, it would give you a more accurate picture and probably allow you to find the percentiles.

Thanks for sharing!

Re: D-1 Outdoor best PV height histogram, by sex

Posted: Tue Mar 31, 2015 12:48 pm
by Divalent
IAmTheWalrus wrote:... Do you think it would work better as a summed graph? I think it's a little weird where the graph dips down around 17' or so and then spikes up again. Clearly anyone who has jumped 17'6 has the capability to jump 17', and so I think that if a data point counted for its own height as well as every height below, it would give you a more accurate picture and probably allow you to find the percentiles. ...


I think you are asking if I can integrate it, so that it starts at 100% at the lowest bar (everyone clears at least that) then declines down to 0% as the bar rises. If I got the time ... although I think the information is all there.

I'm not sure what causes the Spikeyness, (there are others in the main pack that are obscured by my smoothing in that region), but it could be due to the normal progression intervals used in meets, or jumpers late in the season trying to reach the qualifying mark and/or trying to get over the 5.2 mark (i.e., psychological). (The data is what it is.)

BTW, I edited my initial post here to update the image (it now is the whole country, not just the eastern regional) and I think I have the permissions set so you don't have to ask me to see it. New link is:
https://drive.google.com/file/d/0B-oBpKKKOJQpMkgxVW9tYTFESjQ/view?usp=sharing_eid

Re: D-1 Outdoor best PV height histogram, by sex

Posted: Tue Mar 31, 2015 12:50 pm
by IAmTheWalrus
LOL!!

Yes I am speaking of an integral. I don't know how I didn't realize that. Thank you for interpreting my gibberish.

Re: D-1 Outdoor best PV height histogram, by sex

Posted: Tue Mar 31, 2015 4:11 pm
by KirkB
Permissions are working fine now. Thanks.

In addition to the men's spike at 5.20m, I noticed a spike for women's at ~4.20m too.

Kirk

Re: D-1 Outdoor best PV height histogram, by sex

Posted: Tue Mar 31, 2015 9:22 pm
by Divalent
KirkB wrote:... In addition to the men's spike at 5.20m, I noticed a spike for women's at ~4.20m too.
Kirk


I think the spikeyness is mostly an artifact of sorts due to commom progression intervals, as it is particularly noticable in the "average" section of the womens histogram. Here is the unsmoothed histogram: https://drive.google.com/file/d/0B-oBpKKKOJQpV3Vja0xTeGdINlk/view?usp=sharing That mens spike doesn't seem so significant anymore, does it?

Notice that the frequency of spikes is 15 cm, the typical progression interval, and likely represents the fact that near the end of the season (when most vaulters are doing their best and acheiving PRs) they generally have a better idea of what height is gonna be needed to quality for each regional (top 48), so meets will set opening bar so that with 15 cm progressions, it will be landing right about at that height or a cm or 2 higher. (Why have a height just below what you know will be needed, then one well above it; that won't help the competitors in your meet qualify for the Regionals).

Of course, a progression designed to get the better vaulters qualified determines what the lesser vaulters get to attempt. My theory (which I can go test by looking at the data) is that the reason the women's histogram appears spikier is that the minumum qualifying height was roughly the same in both the east and west, whereas the men's was 5 or more (but not 15!) cm different. So Women were in sync, men were out of sync.

[Okay, I looked at the data: women's minimum qualifying heights were about 1 cm apart (3.97 vs 3.96) in the two regions, whereas Mens were about 10 cm apart (5.18 ver 5.07). So for women, a good progression in a late season meet would land on 3.97 to 3.99 in either region. For men, in the east you would probably be looking to land on 5.10, and in the west on 5.20. So that supports my wild a** theory :D ]