I used to think river gravels were boring stuff (as compared to collecting fossils or neat mineral specimens) until I learned more about old river terraces (floodplain remnants) and exposed gravels on hilltops and hillsides. In these settings, the gravels tell us about how much a long-existing river can migrate laterally over time.
These gravels are on the opposite side of the peninsula from the old raceway grandstands. The peninsula lies within an incised meander of the Chattahoochee River and the old river channel cuts across the hill. Meanders are usually present in areas of low-gradient, usually on coastal plains, closer to the ocean. That it is an incised meander suggests that the meander system was locally established (as just described), then with a rapid drop in base-level, down-cutting preserved the meanders.
Where there are sedimentary rocks in the drainage basin, sometimes you can find petrified (permineralized) wood in river gravels. I used to find small rounded pieces of petrified wood in some of the old Rio Grande gravels west of El Paso.
I hope to return to the park to check out the large chunk of rock pictured just below the tree roots. From the surrounding river pebbles, you can see that this chunk is far larger than the rest, suggesting that "there is more to this story". It would take a great deal more energy to move this size rock, as opposed to the adjacent pebbles. In the exposed gravels along the edges of this peninsula, I don't recall having seen other clasts this large.
There are several possibilities.
 A possibility is that this chunk of rock fell into the river from a nearby (now eroded) bluff or rolled down a slope into the river.
 Another more interesting possiblity is that this is a chunk of permineralized wood, i.e., a chunk of wood that was buried in the river gravels and mud and was permineralized by silica in the groundwater after burial. I considered that when I shot the photo, but sundown was approaching and the park was about to close, so I didn't remove it from its position to more closely observe the characteristics. A chunk of wood that large would be light enough to be moved along with the gravels, prior to its permineralization.
If it was permineralized wood, "woodn't" that be neat! [Sorry.] As that sort of thing is not that common on the Piedmont, I would take some more photos before dislodging it from the vertical face. Something like that might be worthy of a short scientific paper, perhaps if any internal cell walls are preserved, a paleo-botanist might be able to determine the genus of the wood, based upon the internal structures.
 One more possibility is that on the nearby exposed shoreline, amid the weathered, saprolitized metamorphic rocks, there are areas of hard quartzite. This chunk of rock could simply be a chunk of quartzite that was "ripped" from the river floor (you can see the scoured bottom below the gravels) by a flood and "rolled" to its present location. It is easier for fast-moving water to roll an object than it is to carry it.
[There, without your realizing it, I have "walked you through" the concept of "Multiple Working Hypotheses".]
I have proposed three possible ways that this large chunk of rock (maximum dimension is perhaps 1.5 feet) came to be placed within these smaller pebbles in the old river bottom. There could be other possibilities that I have not yet considered. Speaking from my bias as a field scientist, this is why we learn to "brainstorm" with other scientists and when properly trained, we engage our own imaginations. If the chunk of rock is not permineralized wood, then that Hypothesis is discarded and the others subjected to greater scrutiny. That is the way science works. And because of the ravages of time and erosion, there may not be a single "best" answer.
Unlike lawyers and politicians, scientists should understand that our initial opinion may not be the correct one.