December 10, 2014

# A Look at Speed by Height

The table and charts below are based on all Regular Masters runs from 11/1/2013 to 11/1/2014. Analyzing this data allows us to answer the question; “what is the average speed at each height”. The histogram charts below show the distribution of speeds for each height. (Note: the charts may take a few moments to load)

## Average YPS at Each Height

Height # of Dogs Avg STD YPS Avg JWW YPS
8″ 1062 3.20 3.92
12″ 1555 3.45 4.31
16″ 1795 3.67 4.66
20″ 3263 4.00 5.08
24″ 1486 3.76 4.70
26″ 199 4.65 5.77

## Distribution of Speeds Across Each Height – Jumpers With Weaves

All 6 Jumpers With Weaves distribution charts use the same horizontal axis range. This allows you to visually compare the yard per second distribution between heights.

## Distribution of Speeds Across Each Height – Standard

All 6 Standard distribution charts use the same horizontal axis range. This allows you to visually compare the yard per second distribution between heights.

## Episode 347: Interview with Penny Leigh, Director of the AKC Agility League

Episode 347: Interview with Penny Leigh, Director of the AKC Agility League

## Episode 346: Interview with Senior Open World Champion Dudley Shumate

Episode 346: Interview with Senior Open World Champion Dudley Shumate
• Deb Joseph says:

Sarah, it would be interesting to see the same averages filtered for dogs that earned more than 500 points (or, qualified for Nationals). -Deb

• Deb Joseph says:

Yes. I agree. Thanks.

• Deb Joseph says:

Well, I think there is a problem with the presentation of the data in the histograms. I could explain mathematically what’s triggering this belief, but its a bit complicated. Perhaps the easiest way to see it, is to simply slide the cursor horizontally across the histogram. The width of each peak should be the same (for example, a peak might count the number of dogs with yps between 4.0 and 4.4 yps, the next peak 4.5 to 4.9), but when I slide the cursor across the histograms I don’t see even widths.

• Sarah Fernandezlopez says:

I think it’s a rendering blip since the chart is rather squished. I can provide a link to the charts in a separate window where they have more room to breathe ðŸ˜‰

• Deb Joseph says:

I’m currently teaching discrete probability, which means I look at all such data very carefully.

I’ve got a couple questions. The histograms are not fully labelled, so its not quite clear what is being plotted. The vertical axis varies from plot to plot, so even visualizing differences is difficult.

I am guessing that the histograms are the averages for each dog that ran in the time period. If that is the case, a dog that ran only once very fast has a high average. While a dog that ran a thousand times and had an even faster run will show up with a slower average. Is that the case? I’m trying to understand how to interpret the shape of the histogram and the outliers.

Also, since the average of the averages is not the average, what’s the meaning of the chart. For example, did you take all JWW runs by all 8″ dogs and find the average? Or, did you find the average for every 8″ dog that ran JWW, and then take the average of these averages?

Thanks, Deb Joseph

• Hi Deb!
Second question first, I did an average of the averages so that I’m getting an average speed for an 8″ dog rather than an average speed for an 8″ run (different, as you point out).

The histogram is showing the distribution of AVERAGE YPS. So each dog has an average YPS and this chart then shows the distribution, so showing the number of dogs that have the same average YPS range. I left the vertical access alone for the charts (number of dogs), since I’m not as interested in the size of the class and it would cause the smaller classes (26″ for example) to not visually show the variation across the jump height. But I did line up the horizontal access so that you can compare SPEED across heights.