FORCE PLATE DATA
&
THE DANGERS OF GENERALIZED DATA
One of the easiest, and at the same time hardest things about using a lot of technology, is making sense of the data. Reading studies that others have put out is fairly straightforward. They've done the work and published the results. You can then use those results to help shape how you go about training athletes. However, the one thing I've found is that when results get published, they tend to be generalizations. While generalizations can be useful, they often don't mean much for the individual.
This is where the dangers of generalizations come into play. Looking at the average correlation's previously, it would've been easy to conclude that there is no correlation between ground forces and velocity. As you can see in this example, ground force data may be very important for this individual. I'm hesitant to say it definitely is because these are limited data sets. Since it's just 5 throws, the trend may not hold up over hundreds of throws. The data could become more similar to the generalization. However, if the trend were to stabilize with a correlation around a 0.45, there's still something here. Everyone generates velocity differently. It's up to those working with pitcher's to figure out how and then use it to help the individual develop!
Over the last few months, I've looked into Ground Forces quite a bit. I've seen studies that suggests that there's not much correlation between Ground Forces and Pitching Velocity. However, with a lot of discussion surrounding the significance of the front leg's ability to stabilize recently, some believe that front leg forces are very important.
With the help of a Wii Balance Board and the installation of a program to measure force and rate of force (http://www.rehabtools.org/strength.html ), I had to dive in an experiment for myself. For the experiment, 7 pitchers were used and they each performed the same drill while Force and Rate of Force (RFD) were tested on each leg. Each guy threw approximately 5 pitches (more throws were added when mis-reads occured) and averages were taken for each. Force is measured in kg and RFD is measured in kg/sec. When all was said and done, the results yielded results that were somewhat expected based off of other studies I had seen.
| Velo | BL Force | BL RFD |
| 79.533 | 81 | 280.567 |
| 78.56 | 160.46 | 1087 |
| 74.85 | 79.8 | 251.25 |
| 80.85 | 135.025 | 679.325 |
| 79.2667 | 77.8333 | 275.7 |
| 77.9 | 101.75 | 621.3 |
| 76.35 | 108.7 | 679.5 |
| Correlation with Velo | 0.180611864 | 0.046609374 |
| Velo | FL Force | FL RFD |
| 77.1 | 137.667 | 865.4 |
| 78.08 | 141.68 | 867.22 |
| 75.55 | 129.65 | 1075.75 |
| 79.3667 | 174.9 | 1496.87 |
| 80.333 | 80.2 | 664.433 |
| 75.95 | 99.475 | 743.975 |
| 74.06667 | 113.3 | 803.3667 |
| Correlation with Velo | 0.1023267 | 0.197671198 |
As you can see, all 4 of these correlations are very weak. That doesn't necessarily mean that the results aren't useful. Look at the 2 different guys highlighted above in red & yellow. They have a pretty similar max indoor mound velocity (90.8 vs 89.9). They both generated pretty similar velocity, force and rate of force with the back leg for this test. The front leg tells a much different story. One stabilizes fairly well (red). The other doesn't (yellow). Now to dive even deeper into how this data isn't a waste and show an example of the dangers of generalizations, let's take out one of individual data sets from above (highlighted in grey).
| Velo | BL Force | BL RFD |
| 75.2 | 144 | 921.5 |
| 77.5 | 167.8 | 1180.9 |
| 82.7 | 167.4 | 1247.8 |
| 78.5 | 165.2 | 1075.5 |
| 78.9 | 157.9 | 1009.3 |
| Correlation with Velo | 0.687653072 | 0.770633418 |
| Velo | FL Force | FL RFD |
| 71.9 | 138.5 | 737.2 |
| 78.8 | 148.3 | 815.1 |
| 80.8 | 143.2 | 1034.5 |
| 77.7 | 142.9 | 744.5 |
| 81.2 | 135.5 | 1004.8 |
| Correlation with Velo | 0.1201869 | 0.793923004 |
This is where the dangers of generalizations come into play. Looking at the average correlation's previously, it would've been easy to conclude that there is no correlation between ground forces and velocity. As you can see in this example, ground force data may be very important for this individual. I'm hesitant to say it definitely is because these are limited data sets. Since it's just 5 throws, the trend may not hold up over hundreds of throws. The data could become more similar to the generalization. However, if the trend were to stabilize with a correlation around a 0.45, there's still something here. Everyone generates velocity differently. It's up to those working with pitcher's to figure out how and then use it to help the individual develop!
Some things I want to look into further on this subject are how these numbers look when adjusting for the individual's bodyweight and whether or not looking into the differences between back and front leg forces can paint a picture of how well the athlete transfers energy! For a fun experiment, 5 of these 7 guys have hit 89+ mph in games. The other 2 are low 80's. Try to figure out which 2 are low 80's from this data!!
To conclude, there are benefits to generalizations. They just don't necessarily mean anything for the individual athlete. While ground force data may not yield significant correlations with velocity, the individual data could help paint the picture of how the individual generates velocity. Everything can be useful. It just may take a lot more digging than you hope for!
