Some rules for good oral presentations

• Assume the audience knows nothing about the topic. Typically, aiming the talk at a group with less knowledge gets it right (if you speak to experts in the field, address the talk to non-experts, if you speak to faculty, aim at graduate students, ...). Even if you speak about a topic at a place that is famous for work on this topic, it is usually only a handful or less people that actually have worked on the topic. Generally people like it if they hear something they know, but they hate it if they don't understand.
• Keep it interesting and CUTTING EDGE all the time. No general, textbook, lecture, introductory stuff. No history (except to lighten things up deliberately). Of course the basics need to be explained. The art is to do this in simple words as needed
  along with forefront science. As a rule of thumb, each slide should have some forefront science, even if it also explains some basics.
• Convey excitement. Be excited. Say it explicitly if something is interesting or exciting, otherwise the audience will not know.
• No Bullets. At least that is the ideal talk. Find a way to visualize all points you want to make
  with a graphics, picture, etc. Why? Bullets can be read by the audience in a 1/10th of time the speaker needs to read them. So if you read the bullets its really boring, if you don't read them they are useless anyways. In essence, bullets are not suitable for a visual presentation - you could as well hand out your bulleted list for reading. No need to give a talk about it. There are a few possible exceptions such as summarizing the main message of the slide along with graphics, the conclusion at the end of the talk etc. But most slides should be bullet free!

  In particular do not use the format where bullets are on the left/right half of the slide and a picture is on the right. Rather make the picture big and relate text lables to features in the picture directly. If this is not possible you probably did not pick the right picture to make your point.

• No definitions, introductory explanations etc at the beginning - explain everything at the time it is needed.
  Better repeat explanations of unfamiliar symbols, definitions.Do not assume just because you said something, from now on everybody knows it.
• Keep the time under all circumstances.
• Create a story line from beginning to end. Do not just add thing after thing, each transition to the next slide should be motivated by a storyline.
• Each transparency should have a clear message and a clear purpose within the story. Sometimes its good to explicitly add the punch line in words to each slide, especially if slides get posted or distributed.
• Do not switch topics in the last 1/3 of the talk.
• Use equations only if absolutely necessary. Do not show them to demonstrate "how you did something" but only if they serve a paedagogical purpose (for example, to show that a quantity depends exponentially on another one, or to show the ingredients needed to calculate something). In theory talks, higher level explanations than the detailed equations are needed - ideally through some visualization. Showing lots of results, also from intermediate steps is also a good strategy (the same is of course true for experimental talks: don't show electronic diagrams or data acqusition code but higher level concepts). If equations and derivations are essential use the blackboard. This will get the timing right.
• Experimental talks should show a picture of the setup.
• Answer questions briefly.
One or two sentences are preferred. Do not go back to the slides unless absolutely necessary.

back to PHY983