editor's blog
Subscribe Now

A Conference Presentation Figure of Merit

In this business, I end up going to a lot of conferences on a lot of different topics. Now, I was trained as an engineer and have worked closely with technology for a long time. But it’s also been many years since I was a practicing engineer. And it’s been even longer – college, mostly – since I had to do much of anything with analog circuits or transistor device physics or any of dozens of specialized topics that come together to make our industry possible. And no, Mom, after all these years, I still can’t fix the radio.

Now, some conferences have presentations that are little more than slightly watered-down sales pitches. I won’t name names; you know who you (they) are. At the complete opposite end of the spectrum are ones like IEDM and ISSCC, which don’t even have exhibits – they’re that serious. For those of you that don’t know those shows, IEDM is about transistors and ISSCC is about circuits; I just use them as examples. But each discipline has its equivalent. It’s hardcore or gtfo.

Having sat through several of these sessions over the years, I was musing on how hard it can be for a non-specialist like me to keep up with these presentations. And I suddenly realized exactly what it feels like.

At the beginning of a talk, you walk up behind a pickup truck with a rope tied to the ball hitch and someone gives you the other end of the rope to hold on to for dear life.

And then the talk starts: the pickup starts moving, and you start trotting along behind it. As the pace picks up, you go faster and faster. At some point, try as you might, your footing fails, but your grip holds true – and you’re being dragged along behind, sort of keeping up, but only in theory, and it hurts. As things progress, eventually you can no longer maintain the charade, and your grip fails, and you crumple in a glorious cloud of dust as the truck proceeds without you.

That’s fine as far as it goes qualitatively, but it made me wonder if I was the only one experiencing this. Presumably, the rest of the audience is better suited to the material. But it occurred to me that, as a good engineer, and in an attempt to rehabilitate my pride, I could fashion this into a Figure of Merit (FoM) for conference presentations. Here’s how it works.

We define four specific events in a presentation as experienced by the attendee. Two are obvious: the start and end of the presentation. The third is that moment when your footing goes and you aren’t running any more, but you’re still holding on. We’ll call this the Skid time. The final event is when you give up entirely; we’ll call this the Bail time.

Now, give everyone in the audience a little wireless box called a Bail/Skid (B/S) Recorder. It has two buttons: one marked “Skid” (or “S” or, if it’s for a Unix topic, then something completely unrelated like “^”) and one marked “Bail” (or ibid.). Now, if and when each listener experiences a personal skid or bail moment, he or she can record it by hitting the appropriate button, which relays the event wirelessly to the Session B/S Aggregator. (A pad with a gyro sensor would probably need to be included, attached to the forehead so that the unit can note whether the attendee fell asleep and not misinterpret the lack of clicks as comprehension.)

The Aggregator can then assemble a Bail/Skid Quotient for each audience member. This is defined as a two-ratio measure. The first number represents at what point in the presentation you bailed; the second gives the point where you went into full skid. Specifically, if C is the commencement time, S is the skid event time, B is the bail event time, and E is the session end time, then define the duple:


So, for instance, if a presentation starts at 9:00, you skid at 9:10, you bail at 9:20, and the session ends at 9:30, then your B/S Quotient would be [67,33]. Simple.

So what can we do with this? Lots!

First, if I feel like I’m the only one in the room experiencing this, I can test this theory by comparing my personal B/S Quotient to the average of the room – this is my B/S Rank. If it aligns well with the rest of the group, then I did OK – and then, if the ratios are low (indicating early loss of traction), we know that it was due to a poor presentation, not my poor brain.

The larger the average B/SQ, the better the presentation – and also the best tailored to the audience.

We can also tease out the cause of poor comprehension. In general, it could be because it was a poorly crafted or targeted presentation, or, alternatively, the material might have been very good, but it’s not uncommon for the presenters to have reasonably thick accents that simply make the communication of that accent different. They shouldn’t be penalized for this ; heck, these guys are engineers and hardly have to talk to anyone anyway. They did some good work, and as punishment for that, they’re made to stand in front of an audience to talk for half an hour. Nice motivation…

If the former were true, then the B/SQ distribution should show a single mode occurring early in the presentation: everyone got lost pretty quickly. On the other hand, if it were the second case, then we can rely on a very likely assumption to help determine this: it’s very unusual for someone to have an accent that no one in the audience can understand well. So you’ll end up with a bimodal distribution, with a large mode dropping out early, but a small later peak that represents the few that make it most or all of the way through just fine– for instance,  those few other surfers in the audience that were able to parse the “brahs” and “gnarlies”.

You can also play with the difference between the skid and bail times to learn things. Does the bail happen immediately after the skid, or do you end up getting pretty scratched up before giving up? Does that reflect presentation timing, listener bravery, or listener stubbornness?

And might some pre-game analysis by a vindictive presenter assembling a talk allow him or her to craft a well-placed skid, followed by an easing up to maximize skid pain? If skillfully done, the audience might even get to the point of thinking it could recover from the skid, at which point the speaker would speed up just slightly to keep them off their toes.

Yes, a good time could be had. And it would allow much better post-talk analysis to understand in better depth just what when wrong where and why, which isn’t actually useful except in the unlikely event that the same presenter presents again another year. But you know, you can never have too many metrics.

I’m really excited about this new approach. It may need some refinement, like an overall qualifier as to the general alertness level going into the session (call it the Denali Party Level), but I think I’m on to something. In fact, I think I’ll submit it as a paper. Hmmm… which conference?

Wait, did I hear someone bail already? Please, come back!

Leave a Reply

featured blogs
Aug 18, 2018
Once upon a time, the Santa Clara Valley was called the Valley of Heart'€™s Delight; the main industry was growing prunes; and there were orchards filled with apricot and cherry trees all over the place. Then in 1955, a future Nobel Prize winner named William Shockley moved...
Aug 17, 2018
Samtec’s growing portfolio of high-performance Silicon-to-Silicon'„¢ Applications Solutions answer the design challenges of routing 56 Gbps signals through a system. However, finding the ideal solution in a single-click probably is an obstacle. Samtec last updated the...
Aug 17, 2018
If you read my post Who Put the Silicon in Silicon Valley? then you know my conclusion: Let's go with Shockley. He invented the transistor, came here, hired a bunch of young PhDs, and sent them out (by accident, not design) to create the companies, that created the compa...
Aug 16, 2018
All of the little details were squared up when the check-plots came out for "final" review. Those same preliminary files were shared with the fab and assembly units and, of course, the vendors have c...
Jul 30, 2018
As discussed in part 1 of this blog post, each instance of an Achronix Speedcore eFPGA in your ASIC or SoC design must be configured after the system powers up because Speedcore eFPGAs employ nonvolatile SRAM technology to store its configuration bits. The time required to pr...