Social media numbers for “The Voice” semi-finals

Back from a hiatus on the Big Island to race and relax, it’s time once again to review the social numbers for this week’s “Voice”.” Past weeks have shown a relatively poor correlation between buzz on Twitter and actual results, but as the pool of contestants has decreased, the correlation has improved. This can of course be due to the increased probability of being correct simply as a function of smaller potential outcomes, but let’s see what the numbers say regardless.

For the period spanning last night’s live performances on the East Coast through midday today, we see the following:

the voice overallCompare this with only positive mentions, and we see a similar distribution:

the voice positive

The good news is that there’s a clear separation between the top two and the remaining three, so it appears as though Amber and Danielle are clearly safe. Two will be sent home tonight, though, and while it appears as though Sasha will be one of them, the difference between Michelle and the Swon Brothers is not very large. Both also seem to have a very loyal following that has kept them on the show even when the buzz numbers were relatively low, so at this point it’s just a guess as to which of the two will remain after tonight’s show.

Here I’m going with the Swon Brothers, purely because the novelty of a duo going to the finals vs three female soloists might have gotten the vote out better.

As this season has shown, though, the social numbers have been wrong before, so don’t take it to Vegas. :-)

Hawaii 70.3 — a hot time was had by all

By some measures, I should be disappointed in my Hawaii 70.3 (“Honu”) result — I got 9th in the M50-54 age group as opposed to 7th last year, I failed to break five hours, and my body betrayed me in the run, which is supposed to be my best event:

honu 2013 result

But I like to take an analytical approach to my racing. Let’s compare some numbers:

2012 Top Male:

Name Swim Bike Run Finish
Armstrong, Lance 00:23:22 02:01:46 01:22:29 03:50:55

2013 Top Male:

Name Swim Bike Run Finish
Alexander, Craig 00:25:11 02:13:59 01:23:08 04:05:43

Now, Alexander is not the cyclist that Lance is, but he is a much better runner, and is a three-time winner of the full Hawaii Ironman to boot. So a winning time almost 15 minutes slower than last year is perhaps an indication that the conditions were more difficult this year.

Let’s compare the same athletes year over year:

Name Year Swim Bike Run Finish
Wee, Bree 2012 00:27:01 02:29:10 01:32:51 04:32:45
WEE, Bree 2013 00:28:29 02:30:31 01:48:08 04:51:05
Beckmann, Holger 2012 00:32:01 02:33:22 01:33:48 04:44:48
BECKMANN, Holger 2013 00:40:41 02:38:19 01:38:56 05:05:46
Doi, Keish 2012 00:34:52 02:37:38 01:31:37 04:48:58
DOI, Keish 2013 00:36:43 02:35:39 01:42:30 05:00:07
Hersey, Ian 2012 00:38:37 02:40:27 01:47:22 05:13:30
HERSEY, Ian 2013 00:37:40 02:36:16 01:51:05 05:11:33

This isn’t a complete analysis of everyone who competed in both years of course, but it does look at a female pro and a couple of top older male age groupers, and then at me. A couple of things emerge from the numbers:

  • The swim was slower in 2013
  • The run was slower in 2013

It’s not clear to me exactly why the swim was slower, but they did reverse the direction of the course this year, which might have been a factor. I was faster this year than last, but I’ve been swimming a lot faster in workouts this year, so should have been several minutes faster. For the run, it’s crystal clear why it was slower: it was HOT! Last year’s race had some very strong winds, which kept the temperature down.

My own run performance came down to the fact that my run training has been next to nothing since mid April when I strained my hamstring, forcing me out of my planned Big Sur Marathon and making the prospects of even getting through the Honu run iffy at best. Fortunately, the hamstring healed after four weeks, but the damage was done to my run fitness. But that isn’t an excuse — I came with great overall fitness and confidence, and was ready to race.

The swim started well — I placed myself on the right side of the field and swam un-punched and un-kicked for the entire time. The change that the race organizers made to start the age group women seven minutes after the age group men really helped. Fewer bodies, fewer arms and legs. It also made a big difference on the bike, because in years past I’d have to pass a lot of women who were fast in the water but clogging up the bike course (no offense intended — I know I could also solve the problem by “swimming faster”). The only tough part of the swim was after the final turn buoy, heading back to shore — that’s where the wind was coming from, and the chop was noticeable. The sun was also right in our faces, so it was difficult to see where the swim finish actually was.

Once out, though, I noted the time (a little disappointed, but you don’t dwell on a minute or two in such a long race) and headed up the ramp for T1. Though I needed help from a volunteer first to get my swimskin unzipped – the zipper was stuck. She had trouble as well, but eventually got it to unzip. I sprinted up the hill and found my bike, and made one of my faster transitions, other than fumbling with the shoes I already had in the pedals (the course goes sharply uphill right out of transition, so it’s tricky to jump on the bike without your shoes already on your feet — something to practice). Once off and going, I started passing guys (and a few really fast female swimmers) pretty quickly. The wind was going pretty fiercely and was mainly a crosswind, which requires a bit of nerve to ignore and just stay down in the aerobars in.

I was holding around 210-220 watts for the first hour, and averaged about 22 mph. Then the climb to Hawi came, which involves not only a long uphill with a few steepish parts but also a pretty good headwind. At some point near the turnaround, a fellow Wattie Ink guy passed me, but I caught him on the ensuing descent, which was blissfully free of the dreaded crosswinds this year, which meant I could cruise pretty easily in the aerobars at 30+ mph, regaining a lot of the average speed I had lost on the uphill. The Wattie guy and I were exchanging positions every so often, as I would push the uphills a little harder than he did. He turned out to be a younger guy named Dillon, whose dad John was also a racing Wattie (John passed me in the run — it would become a theme).

On the section back to Kawaihae and the Queen K, I continued picking off people, including a few female pros and one or two female age groupers who were swimming and riding like pros :-) , but I stopped myself short of really hammering — that’s what makes the power meter such a great tool if you know how to use it. I knew I was on my way to a good bike split, however, and sure enough I ended up notching a four-minute PR for the course. I also noticed very few bikes in the transition area, especially anywhere near my number, which meant I was doing pretty well in the age group (turns out I was 9th off the bike). But now I was venturing into the unknown — the longest run I  would have done since February or March.

It started off well enough, though immediately I could tell that it was substantially hotter than last year. The key was to keep my hydration and electrolytes going, as cramping in the heat is pretty much my standard MO. I had to reset my Garmin because my Auto Multisport mode apparently didn’t have transitions included, so the watch had stopped recording. I use the Garmin for pacing — mainly to hold myself back early on. The first mile was still a 7:0x, but it felt easy enough. That would change pretty quickly once I hit the sections on the golf course, where the spongy grass sucked all speed from the legs. Somewhere in mile 4, the first little leg cramp happened, but I took care of it pretty quickly with some water and an electrolyte cap. I had taken roughly 6-7 of those on the bike, and in hindsight that probably wasn’t enough.

Saw a few Wattie teammates out on course: first, the young Dillon that I’d been playing leapfrog on the bike with passed me in mile 3, and I didn’t think there was any way he was coming back to me. Then, by surprise, his dad John passed me in mile 4 and introduced himself. I kept John in sight for a while, and even re-passed him in mile 7 (after having my first solid set of leg cramps, which stopped me dead in my tracks). When I went past him, I joked that my teammates call me “a tenacious bastard.” Just before mile 9, as we headed out on the long, desolate out-and-back section of road, I cramped hard. I tried walking backwards, sideways — anything to use different muscles and work out the cramps. John went past me at this point and told me I’d need some of that tenacity — he was spot on.

Finally the cramps got worked out and I resumed running, until an even worse set happened 20 yards before the next aid station. It was so bad that I had to ask the volunteers to bring the water to me — I couldn’t move. I took my time, took a couple of electrolyte caps, and gradually I could start moving again. It was odd: for a few minutes I could run 7:20 pace, then all of a sudden I was stopped dead in my tracks, then the cramps subsided and I could run again. That’s pretty much how it went for the rest of the race. The last quarter mile I was determined to keep running no matter what, and almost every muscle in both legs was twitching, ready to lock up. But I finally made it over the finish line, and a PR is a PR, no matter how ugly parts of it were.

I met a few other Watties in the finishers area — here’s me with Mickey McDonald from Bend, OR, who crushed the bike with a 2:29 split:

watties honu

My Team Sheeper buddies Mike and Steve finished together, also having tough runs laden with cramping, so I wasn’t alone in my suffering. :-)  Here’s the happy crew:

sheepers honu

Now it’s on to my Kona prep — if nothing else, I learned from this race how far I still have to go to get my run fitness and my hydration up to the task of double the distance and more than double the suffering. Bring it on!

A big thanks to all of my sponsors: TrainingPeaks, which I use religiously; Wattie Ink, which represents a great set of product sponsors and athletes (and has a very cool-looking kit); and, last but not least, Team Sheeper, the greatest combination of training program, training partners and friends a guy could ask for.

Another Week, Another Set of Live Rounds of “The Voice”

This week on “The Voice,” we’re down to 10 contestants, another two of whom go home tonight, so if nothing else our odds of predicting who’s going home are naturally improving week to week. But let’s look at the social media numbers again. First, overall “share of voice”:

thevoice-overall 2013-05-21 at 10.07.39 AM

Then, positive “share of voice”:

thevoice-positive 2013-05-21 at 10.07.19 AM

For the first week ever, both rankings produce the same bottom two, Amber and Kris.

If we look at week-to-week follower count growth, however, we get a different ranking:

Twitter handle Followers 5/14 Followers 5/21 Delta Percentage
@theswonbrothers 14750 19588 4838 33%
@michellechamuel 21305 27370 6065 28%
@josiahhawley 40894 50517 9623 24%
@dbradbery 45723 55430 9707 21%
@hollytmusic 13454 16273 2819 21%
@ambercarrington 14520 17149 2629 18%
@kristhomasmusik 12130 13966 1836 15%
@sarahsimmusic 25314 28828 3514 14%
@sashaallenmusic 16674 18985 2311 14%
@judith_hill 39953 43041 3088 8%

So the question is whether buzz or follower count (or neither) correlates better with actual results. So far, follower count has correlated a lot less well, so this week I’m going with buzz — especially since overall buzz and positive buzz line up this time.

But social numbers haven’t been right yet this season, which I’m sure make Kris and Amber happy to no end. :-)

What Did We Learn from This Week’s “The Voice”?

Yesterday I looked at a number of different cuts of social media data and found a number of candidates for the Bottom 2 on “The Voice.” The interesting thing about the actual results was that none of the data I had access to correlated with the outcome, at least in a discernable pattern that you could build any sort of rule or model off of. I could try using a stats tool like R to find non-intuitive features that better predict an outcome, but that is probably like using a bulldozer to pound in a nail — there aren’t enough samples to make that approach meaningful. So I’m left with hypotheses:

  • Twitter isn’t a good proxy for voting behavior in the case of this show due to either a demographic skew or to the fact that the audience gravitates in general to other channels of interaction.
  • I haven’t found the right features yet. I used growth in Twitter followers as one potential proxy for “momentum,” and that correlated with one of the results — Vedo had by far the lowest percentage growth in followers week to week (he started of course with the highest number, so it gets harder to maintain the same growth rate), but in the case of the other ousted contestant, Garrett Gardner, we saw the highest percentage growth in followers, so that piece of data doesn’t correlate at all. Garrett, in fact, by any of the numbers I had access to or derived, should still be on the show. (Though I personally found his vocal performance on Monday weak, despite an interesting arrangement.)
  • The vote tallies among the bottom half of the contestants are actually fairly close (which the counts I had showed as well), so until we get to a point where we see either much greater volume or much greater differences between contestants, the best we’re going to do is random guesses over a larger pool of contestants with similar numbers. In other words, absolute rankings don’t work until you start to see larger gaps.

One thing is for sure — pure follower count is meaningless in this show (as it was in last week’s “American Idol”). A lesson for all who use follower count as a proxy for influence.

Social Media Numbers for “The Voice” Top 12

Last night’s “Voice” episode featured the Top 12, and for the first time this season, audience votes and downloads are the only determining factor for who goes home tonight. Like last week, I’ll be looking at Twitter numbers as a proxy for voting behavior so that we can see the extent to which opinions expressed there correlate with results. The voting methods as explained on the show were phone, text, Facebook and iTunes download, so Twitter is an independent factor in this equation and has no direct bearing on the outcome.

Last week, the Twitter-derived numbers correlated reasonably well. Not perfectly, though. For Team Adam, I hadn’t set up the topic in time, so had about 1/5 of the numbers I had for the Team Blake and Team Shakira contestants. So there my numbers were just plain off. The other factor was the judge’s discretion, and Blake elected to save the Swon Brothers rather than Justin Rivers. Welcome to the big leagues!

I personally enjoyed many of last night’s performances, but what’s important is what those who might vote thought. I took a couple of different cuts at the data. The first is simply looking at “share of voice” from the period of the airing of the show (starting in the Eastern time zone) through late this morning Pacific time. Here’s what we get:

TheVoice-2013-05-14

Simply taking the ranking from most mentions to least during this time period, we end up with a Bottom 2 of Kris Thomas and Michelle Chamuel. We could look at the data another way, though: if we filter by positive sentiment — the idea being that people who express positive opinions are more likely to vote (or perhaps “you vote FOR someone, not AGAINST someone else”) — then the rankings change a little:

TheVoice-positive-2013-05-14

 

Now Amber Carrington and Holly Tucker make up the Bottom 2. However, note that the overall numbers aren’t very high when we apply this filter (sentiment analysis isn’t an exact science by any means, and the software used here is biased towards precision over recall, so is somewhat conservative.

Confounding the data further is iTunes, which doesn’t give precise data but does provide a “popularity” meter. If we look at each of the above sets of Bottom 2, we do see that Amber and Michelle’s performances did not max out the popularity meter, whereas Kris’ and Holly’s did.

One other data point we can look at is a week-to-week difference in Twitter followers. If we rank by percentage gain first, then by overall number, we get:

Twitter handle FollowersLastWeek FollowersThisWeek Delta Percentage
@garrettgardner2 16893 24426 7533 45%
@michellechamuel 14872 21305 6433 43%
@dbradbery 32352 45723 13371 41%
@josiahhawley 29100 40894 11794 41%
@theswonbrothers 10644 14750 4106 39%
@hollytmusic 10232 13454 3222 31%
@ambercarrington 11274 14520 3246 29%
@kristhomasmusik 9721 12130 2409 25%
@sarahsimmusic 20826 25314 4488 22%
@sashaallenmusic 13735 16674 2939 21%
@judith_hill 36078 39953 3875 11%
@vedothesinger 59217 61606 2389 4%

Based on this, it looks like Michelle picked up a lot of followers, so I’m going to have to go with Kris and Amber as the ones going home, with a possibility that Holly ends up somewhere in there. But I’m not taking it to Vegas — the differences aren’t large. Tough call this week!

Enjoy the show!

It’s Finals Season Once Again — Prediction Time for “The Voice”

I was planning to enjoy a few months as a professional triathlete — well, that’s what my wife calls my vacation from work — but a little issue somewhere in the left glute / piriformis / hamstring has curtailed my running, so I find myself back in the world of social analytics, but totally for fun. And what’s more fun than finals season on “American Idol” and “The Voice”?

I’m often asked about (and asked to speak about) the predictive power of social media analysis, and I always tell people it’s more of an art than a science. Also, the predictive power of social media numbers varies according to what you’re trying to predict. Events that are rooted in popularity, though, correlate pretty well with numbers you see in social media, and that’s especially true where the events have a voting system that isn’t just one person, one vote (like political elections). Hence, televised singing contests where the audience votes.

We’re well into “Idol” right now — we’re down to the Top 3, one of whom gets booted this week, and next week is the finale — so I’ve decided to take a look instead at “The Voice,” which is several weeks away from its finale and still has, as of this writing, 16 contestants. Tonight, four will go home — one from each judge’s team. The way the process is supposed to work, the top two vote getters on each of the four teams advance, and then each judge gets to choose among his/her bottom two for who advances and who goes home. Therefore, audience votes get you in the top two, but then it’s up to the judges — not the best scenario for showcasing predictive power of social media.

Nevertheless, let’s look at the data we have. First, the contestants, their Twitter handles, their current follower count, and their Klout score:

Contestant Twitter handle Followers Klout
Danielle Bradbery @dbradbery 32352 66
Holly Tucker @hollytmusic 10232 65
Justin Rivers @justinrivers 58910 63
Swon Brothers @theswonbrothers 10644 63
Sasha Allen @sashaallenmusic 13735 68
Garrett Gardner @garrettgardner2 16893 63
Kris Thomas @kristhomasmusik 9721 65
Karina Iglesias @karinaiglesias_ 9024 64
Caroline Glaser @carolineglaser 40132 68
Judith Hill @judith_hill 36078 73
Sarah Simmons @sarahsimmusic 20826 66
Amber Carrington @ambercarrington 11274 63
Josiah Hawley @josiahhawley 29100 68
Michelle Chamuel @michellechamuel 14872 66
Vedo @vedothesinger 59217 72
Cathia @cathiasings 10527 66

Certain contestants “punch above their weight” when you compare their Klout score to their follower count; Klout’s metrics are proprietary but place a greater emphasis on engagement (e.g., replies, retweets, etc.) vs. pure potential audience size.

Just as a comparison point, we if we look at the three remaining contestants on “Idol,” we see the following:

Contestant Twitter handle Followers Klout
Angie Miller @angieai12 143664 79
Candice Glover @candiceai12 78983 78
Kree Harrison @kreeai12 74503 74

So even if “The Voice” is beating “Idol” in the ratings, the “Idol” contestants have a greater social media presence by several measures than do the “Voice” contestants. An interesting side note: all “Idol” finalists — including the ones that have long since gone home, have “verified” accounts on Twitter (meaning that Twitter considers them celebrities), whereas none of the “Voice” contestants do. This suggests to me that a deal was brokered between the “Idol” producers and Twitter.

Then we have actual buzz on social media. In my Attensity Media account, I set up a “Voice” topic and also created specific “entities” for all of the contestants that conflated their names, Twitter handles and hashtags so that I could get their counts in one place. I did set this up pretty late on Monday, after the show had aired in prime time, so the “Team Adam” and “Team Usher” contestants (who performed that evening) will have lower numbers than the “Team Blake” and “Team Shakira” contestants, who performed last night. Anyway, here’s what the live dashboard looks like:

The Voice May 6 1:49 PDT

So what do we have? We can compare two sets of numbers: general social media popularity and current week’s “buzz,” keeping in mind the grouping by judge. If we do that, we get:

Team Adam

Caroline Glaser 40132 68 2690
Judith Hill 36078 73 2500
Sarah Simmons 20826 66 1238
Amber Carrington 11274 63 1101

Team Blake

Justin Rivers 58910 63 4671
Danielle Bradbery 32352 66 10598
Swon Brothers 10644 63 4011
Holly Tucker 10232 65 5167

Team Shakira

Garrett Gardner 16893 63 5388
Sasha Allen 13735 68 5515
Kris Thomas 9721 65 4192
Karina Iglesias 9024 64 2439

Team Usher

Vedo 59217 72 1516
Josiah Hawley 29100 68 1784
Michelle Chamuel 14872 66 1728
Cathia 10527 66 1125

Again, the buzz numbers are artificially low for Team Adam and Team Usher contestants since I set up the topic late. There’s also the judges’ discretion in which of the bottom two vote-getters each judge decides to eliminate. That said, we’ll see tonight how predictive the social numbers are. Enjoy the show!

Another Race in Paradise

My wife’s a travel writer and covers Hawaii extensively, and one of the considerable perks I enjoy by virtue of being married to her is the ability to tag along on some of her work trips. And so it was that I found myself on Maui last weekend during the Maui Oceanfront Marathon festival of races. There were quite an array of races all happening on the same day: a marathon, a half marathon, a 15K, a 10K and a 5K. The half and 10K were on  an out-and-back course starting and finishing in Lahaina, and the rest were point-to-point affairs, requiring a shuttle bus to get to the start. All races finished in the same place in Lahaina town. We were staying up in Napili, right next to the Kapalua resort, which made my choice of the half work pretty well logistically.

Not that I had bothered to bone up too much on the logistics – this was strictly a “fun race,” an early season test of running fitness. I was hoping to go under 1:25, or just slightly faster than 6:30 pace. We’d had a nice dinner at Merriman’s Kapalua the night before, complete with wine pairings, which I heartily recommend — other than perhaps the night before a half marathon. :-) Woke up to some slight GI distress, which I won’t go into detail on, but I wasn’t feeling that race ready.

Got to the start line at 6:25 a.m. in plenty of time for a 6:45 a.m. start. However, it turns out that the start was at 6:30 a.m. (did I mention I hadn’t paid too much attention to the logistics?), so I jumped in near the front of the field and figured I’d do my warmup in the first mile. The horn sounded and we were off.

I was running pretty relaxed, and in the first mile I was probably in about 10th place overall. I knew that there was at least one other race going on at the same time — a 10K — but I wasn’t sure if there was a 5K as well. So you really couldn’t tell who was in which race. Plus it was pretty dark out at that hour — I ran with my sunglasses in my hand until there was enough daylight to put them on.

I started reeling in runners after mile 1, which I passed in a somewhat leisurely 6:35. The first female was my first passee, then I came up on a group of three guys running together. I went past them and surged as I did, just to discourage anyone from sitting on my wheel. One older guy in a “Yukon” singlet did sit on, then surged past me, which I thought was an interesting move, so I tucked in for a little while. Mile 2 was 6:19, so the surging had definitely picked the pace up. I was trying to stay relaxed, though — there was still a long way to go.

“Yukon”‘s breathing was pretty labored, and I could sense him slowing, so I surged past him again, this time for good. Next up ahead were three other runners, and I was starting to close in on them. I hit mile 3 in 6:27 (there was a bit of uphill in that mile), and all of a sudden, the three runners 20 yards a head of me turned at the 5K cone. I thought I was supposed to go on for another 3.5 miles to another turnaround for the half, but the road ahead of me looked closed — there were red cones lined up on the shoulder. So I second-guessed myself and thought that maybe it was a two-lap out-and-back course or something, so I turned back around to follow the others.

I had lost some ground to them during my hesitation, but started reeling them in much more quickly. Mile 4 came in 6:22, right at the point I passed a guy in a Laguna Niguel singlet; then all I could see were two guys together right up ahead of me. I went past them pretty quickly, and now there were only runners coming the other way on their way out. One woman high-fived me and said I was in the lead. That didn’t seem quite right, but I was just focused on staying relaxed and dealing with occasional rumblings from my gut — nothing severe, but I was a little worried about them in the second half of the race.

I hit mile 5 in 6:25, so I was still on goal pace, and behind me I could hear someone coming up on me. It was one of the last guys I had passed, and he looked as though he was making his finishing surge in the 10K. He pulled even with me, and I looked at him and gave him a “good job” nod before letting him go — I still had 7+ miles to go.

Or so I thought. As I came up to the start/finish line, it appeared that there was only a finish chute, not a place to turn around and go back out, and furthermore I was being announced as the 2nd-place finisher! The official time was 40:24 — not the 10K time I would have liked on my permanent record, but oh well.

At least I got to get first dibs on the free post-race massage. :-)

Lesson for the day: if you can’t be bothered to read the race instructions closely, don’t get bummed out when things go awry. Besides, as one of my friends pointed out, “you’re still on Maui after all.”

It's all good

It’s all good