Why Aren’t There More Tape Rounds?

Yesterday, clarinetist Richie Hawley posted an interview on his website about prescreening recordings.  He discussed many aspects of “tape rounds,” including why they’re becoming increasingly more prevalent in the classical music world.  While I definitely agree with Richie’s assessment that tape rounds are a fair process, I find myself asking a simple question:

Why don’t MORE orchestras make tape rounds an integral part of their auditions?

First, let’s look at how orchestras use tape rounds.  In my orchestra (the Nashville Symphony), when more candidates apply than we could possibly hear in two or three days (this is usually the case), the committee separates the candidates into two categories:  “please come to the audition” or “please make a tape, ” and once we review the tapes, we invite the candidates we want to hear to the live audition.  Most orchestras use a similar screening process, but some others approach tape rounds a little differently.  For example, in 2008, the New York Philharmonic did the entire preliminary round for its Principal Clarinet audition by tape, meaning only a small number of candidates were invited to the live audition.  A couple years later, the Chicago Symphony had an “optional” tape round for one of its Principal Clarinet auditions, where candidates could send a tape and receive a recommendation from the committee whether or not they should come to the audition (you have to wonder if anybody actually sent one of those).  When that audition didn’t yield a winner, the subsequent audition had a tape round (not an optional one) – that was the audition in 2011 that Steve Williamson won.

One orchestra that really has the tape round thing figured out is the Boston Symphony.  For their Second Clarinet audition in 2007, there was a tape round even for candidates who one might think wouldn’t be asked to make one (like the eventual winner, Michael Wayne, who was playing Principal Clarinet of the Kansas City Symphony at the time).  The tape round helped the committee narrow the applicant pool down to somewhere between 30-40 candidates for the live prelims, and those who advanced (9 candidates) were invited back three weeks later for the semifinals and finals at the expense of the BSO.

The Boston Symphony got things right at every stage of this audition.  First of all, they narrowed the field enough that they were able to conduct the prelims all in the same day, thereby eliminating any problems that might stem from committee fatigue.  Also, having only one day of prelims (as opposed to two or three) allowed the BSO to hold the audition on a Monday, which is the preferred audition day for candidates with jobs in other orchestras.  Paying for candidates’ travel expenses so they could return three weeks later for the semis and finals was also a smart move.  People might say that only the wealthiest orchestras can afford to do this, but that might not be true, and here’s why:  Here in the Nashville Symphony, it would cost about the same to pay one committee member for one day of auditions as it would to pay for one semifinalist’s travel expenses.  So if you can use a tape round to eliminate one day’s worth of candidates, and then narrow down the number of semifinalists to the same number of people you have on the committee (we have 9 for principal auditions), it’s pretty much a wash financially.  Even if it would cost a little more to do it this way, there’s a ton of benefits for all parties involved.  First of all, if the candidates advance from both the tape round and the live prelims and are then reimbursed for travel to the semis and finals, they’ll have more confidence that the committee really wants to hear their playing and will therefore prepare more thoroughly for the later rounds.  Also, the committee will have more confidence in the semifinalists – the committee can safely assume that these candidates are fully prepared for the audition and are therefore representing themselves more accurately.  As a result, the committee listens more realistically, and the odds of a no-hire are lowered.

After studying the way orchestras have successfully used tape rounds, there are still a few myths about tape rounds that need debunking.  First of all, some people might feel that making a tape is an expensive process that candidates shouldn’t be expected to undertake.  This might have been true twenty years ago, but nowadays there is plenty of widely available technology that candidates can use to make their own tapes.  All it takes is a computer, some sound editing software (some of which, like Audacity, is free), and a relatively inexpensive microphone (such as this one, which I have).  If you can’t record straight into a laptop, you can use a portable recording device (I use a DAT player – if you want something less antiquated, there are many options).  With the increasing ubiquity of tape rounds for schools, summer festivals, and orchestras, young clarinetists would be wise to invest in this equipment and cultivate the skills necessary to use them.

Another myth is that it would be more “fair” or “generous” for orchestras to allow EVERYONE to come take their audition.  Nothing could be further from the truth.  If a committee reviews your tape and decides that you aren’t what they’re looking for, they’ve saved you hundreds of dollars in travel expenses.  Also, when a candidate makes a tape, he or she has an infinite number of attempts to nail the excerpts and to fairly represent his or her best playing.  It’s a privilege to have that kind of control over what you’re presenting to a committee – it’s certainly preferable to spending a bunch of money in travel and lodging just to have a crappy audition, or even worse, a great audition that’s not what the committee is looking for.   If your goal is to “gain experience” by travelling to an audition and warming up in a room full of other candidates, there are plenty of small orchestras that will afford you this opportunity.  Don’t forget that you can always try to simulate this process by having mock auditions with your friends in your home location.

All in all, tape rounds are an easy way to solve a lot of the inherent problems of auditions.  They make the process cheaper, they make it less grueling, and they give the candidates a better chance to represent themselves fairly, all of which increase everyone’s faith in the process.  So the next time you send a tape and the committee decides you’re not invited, maybe YOU should be the one saying “thank you” at the end!

Posted in General | 5 Comments

Instant Runoff Voting: A Fix for the Future?

Here’s a story we’ve all heard too often over the last few years:

A national-level orchestra announces auditions for a principal position, to be held in three months’ time at the orchestra’s home venue. Applications and resumes are sent and screened; letters of invitation are returned. 100 musicians, many of whom have advanced degrees and a decade or more of professional orchestral experience, prepare for months. They travel to the orchestra’s home city at personal expense, and following four days of intensive auditions, word comes down:

Three people in semifinals. One made it to finals. One other musician was auto-advanced.

Nobody was hired.

The audition presents such a high barrier to entry that only the most outstanding musicians will succeed. But in an era in which playing ability is at a higher level than at any time in the past, shouldn’t we see more candidates advancing to semifinal and final audition rounds? Why did recent auditions in Houston, Los Angeles, Chicago, and elsewhere not only fail to hire anyone, but were only able to yield preliminary round advancement rates around 3-5%?

“The Committee reserves the right to immediately dismiss Candidates not meeting the highest artistic standards”

Each committee member has a unique idea as to what makes the ideal musician, but “highest artistic standards” is so subjective that it’s essentially useless in an “advance” or “not advance” situation. We’re left with the problem of trying to please more than half of the panel in order to advance, when nobody – including individual committee members – has a clear idea as to what the committee’s collective preferences are.

What if we could find a way to accurately establish the committee’s overall preferences during the audition, and design the audition to always reflect these preferences in the outcome? What if we could set up an audition where candidates always advance at a rate of 20%, where the process always yields the highest-quality musicians, where the audition always ends with a result, and where the music director always retains the final say?

The problem isn’t with the players and it’s not with the committees. The problem we have is structural – a flaw in the methods used to determine whether or not candidates advance to the next round. This problem can be solved if committees are willing to make changes to their voting protocols.

Let’s look at a group of musicians who’ve just finished playing a preliminary audition round:

The only question a committee member attempts to answer here is, “do I vote to advance this candidate?” It’s a hard question and the answer is highly subject to committee fatigue and contamination from previous rounds. It does not provide any useful information for comparing candidates who did not enjoy majority support, and by attempting to satisfy the “highest artistic standards” criterion with a majority vote requirement, the committee eliminates far too many candidates, paralyzing the audition by the time it gets to the semifinal round.

Now let’s consider what happens when committee members ask a different question, “which of the players in this group do I prefer the most, and which ones the least?” by using another voting protocol called Instant Runoff Voting (IRV).

IRV works by asking the voters to rank the members of a given group in order of most-preferred to least-preferred.  The audition is 100% screened, no voter opt-outs are allowed, and the vote is taken without discussion.  If no majority is found, the lowest-ranked candidate is eliminated and his votes are redistributed – which means that even if a voter’s first choice is eliminated, his input still contributes to the final decision.

Image source: Wikipedia

Here are the same musicians in the prelims with IRV rankings:

In the straight up-or-down vote, Voter F was the oddball, but here, his input is still useful. We don’t have a majority winner yet, so we hold a runoff – Ed is eliminated because he has more “least-preferred” votes than any other candidate. With Ed eliminated from the race, our results look like this:

Runoff Voting Round 2
vote 1: (Charlie) (Alice) (Dan) (Bob)
vote 2: (Dan) (Bob) (Charlie) (Alice)
vote 3: (Bob) (Alice) (Charlie) (Dan)
vote 4: (Charlie) (Alice) (Dan) (Bob)
vote 5: (Alice) (Bob) (Dan) (Charlie)
vote 6: (Dan) (Bob) (Alice) (Charlie)
vote 7: (Charlie) (Bob) (Alice) (Dan)

Alice=1 Bob=1 Charlie=3 Dan=2

Alice and Bob are up for elimination at this point and one must go.  Bob has more “least-preferred” votes (on the right column) and is eliminated.  Continuing on:

Runoff Voting Round 3

vote 1: (Charlie) (Alice) (Dan)
vote 2: (Dan) (Charlie) (Alice)
vote 3: (Alice) (Charlie) (Dan)
vote 4: (Charlie) (Alice) (Dan)
vote 5: (Alice) (Dan) (Charlie)
vote 6: (Dan) (Alice) (Charlie)
vote 7: (Charlie) (Alice) (Dan)

Alice=2 Charlie=3 Dan=2

Here, Dan and Alice are tied but Dan has more “least preferred” votes and is eliminated. Onward…

Runoff Voting Round 4

vote 1: (Charlie) (Alice)
vote 2: (Charlie) (Alice)
vote 3: (Alice) (Charlie)
vote 4: (Charlie) (Alice)
vote 5: (Alice) (Charlie)
vote 6: (Alice) (Charlie)
vote 7: (Charlie) (Alice)

Alice=3 Charlie=4

Charlie will advance to the semifinals. Alice, who initially appeared to be a long shot in the rankings, turns out to have been competitive within this group of candidates. In the consensus-based poll taken earlier, she and Charlie had the same number of votes, which tells us two things: one, non-advancing candidates who do well under the old system will still do well in the new one; two, the IRV rankings yield more finely-grained information on committee preferences given the same group of players. This does leave the question as to what happens when two or more players are highly preferred and would advance out of the same group under consensus voting, and I have a few ideas about that – but for now I’ll leave that particular question as an exercise for the reader. (Comments are open…)

By design, IRV eliminates the problems of committee fatigue, low rates of advancement to semifinals and finals, the spoiler effect, and no-result auditions. One person from each group will always advance, because this is a tournament-style competition rather than a free for all. Instant Runoff Voting is perfect for this, because given a committee and a group of candidates, IRV accurately calculates the committee’s overall preferences.

A 100-candidate tournament-style audition using IRV in groups of five with a mandatory single advancement in each group, randomly selected prelim and semifinal time slots, and a single auto-advanced candidate will always look similar to this:

To be clear, this goal of this audition structure is not to find an outright majority winner in the final – there is no elimination voting, just preferential voting. The goal is to find the most highly-ranked candidates to play trial weeks so the Music Director can then make a decision. If the Director doesn’t like what he hears in the trial, the orchestra holds another audition – and this doesn’t mean the audition failed, it only means that a job was not offered after a trial.

IRV auditions will always end with the committee’s most preferred candidates taking trial weeks. They are designed to have no other outcome.

Compared with traditional majority-vote auditions, 100% screened IRV auditions enjoy the following advantages:

  1. A larger field of candidates are guaranteed to advance to later stages, giving committees access to a deeper candidate pool, particularly in the semifinals.
  2. Committee fatigue does not affect candidate advancement.
  3. Candidates will feel better about the system since the odds of advancing are always at least 20% in each round and at least two candidates will take trial weeks.
  4. It’s nearly impossible to rig this audition, even if the entire committee has gone rogue.
  5. No-result auditions are impossible, making the process much less time- and resource- consuming.
  6. The Music Director retains the final say on hiring after trial periods are complete.

The IRV audition doesn’t solve all of the problems facing orchestra auditions, and it has a few issues of its own. But we’ve all been told for years that the audition is the best method yet invented for tackling the problem of finding a great musician to fill a spot in an orchestra. Assuming that’s true, and we’re going to use auditions to identify suitable candidates, shouldn’t we be discussing realistic ways to correct the flaws inherent to the audition process?

Auditions are difficult, expensive, and time consuming for everyone involved, and “no result” is the worst possible outcome. Incorporating IRV into screened auditions will leave candidates and committees more satisfied with the process, and orchestras will be far more likely to find great musicians to add to their ranks.

For more information, I’ve laid all this out – along with specific suggestions for audition guidelines – in a Google Presentation here.

Posted in Guest Authors | 29 Comments

Houston Symphony: The Bigger Picture

Yesterday, when I heard that the Houston Symphony had finished their principal clarinet audition and not chosen a winner, I wrote a post focusing on a troubling element of the audition, specifically that the anonymity between the audition committee and at least one semifinalist had been compromised.

I wrote that I was frustrated when I heard this, and I realize now that I should’ve been more clear about why I was frustrated.  I was frustrated that in hindsight, some of what happened in the audition can be viewed as unethical and potentially influential over the audition’s outcome.  My frustration had nothing to do with the fact that I didn’t win the audition.

I want everyone to know why Clarinet Jobs was founded.  I founded Clarinet Jobs in 2007 as a means for clarinetists to exchange useful information about auditions.  A lot has changed in the orchestra world in the past six years, but one thing remains true:  We’re all students of the audition process.  No two auditions are exactly alike, and I enjoy cataloging the fascinating experiences of my friends and colleagues, both on Facebook and here on this website.

In yesterday’s case, I used the Clarinet Jobs platform to call into question certain aspects of this audition and to spark an intelligent dialogue.  I was willing to take the heat that comes along with putting the information out there and to face any potentially harmful consequences.  I don’t take these risks to get attention – I do it because I firmly believe in the audition process.  Regardless of what side of the screen I’m on, I advocate for the highest ethical standards possible, because the higher those standards are in an audition, the more likely the audition is to yield the results everyone wants:  A winner.

Yesterday’s post has been removed.  It was not meant to harm anyone – it was meant to inspire change for the next go-round and to remind us all that, whether we like it or not, we represent the orchestras that employ us.  Hopefully, as this dialogue continues, we can discuss some broader questions:  How is it that so few advanced from the prelims?  How do we explain the increased frequency of auditions that result in non-hires?  And, perhaps most interestingly, what will the Houston Symphony do next?

Posted in News | 19 Comments

Welcome to Clarinetjobs.com!

We’re finally here!  We have plans we think are exciting.  Bear with us as we learn the ropes.

We’re not replacing the Facebook page – we’re just looking to expand Clarinet Jobs to include some other content we think the clarinet world will find useful.  Stay tuned!


Posted in News | Leave a comment