Performance Notes: Transcendental Etude No. 11 in D-flat Major (Liszt)

This is a video of me performing Liszt's Transcendental Etude in D-flat Major at Levine, the eleventh in the set of twelve Transcendental Etudes.  The Transcendental Etudes are studies, each of which focuses on a set of specific technical issues.  For example, the fifth in the set focuses on double notes and evenness, the twelfth focuses on tremolos, namely rapid oscillations between two notes or two clusters of notes, and the eleventh one,  which I perform in the video above, focuses on chords and leaps across the keyboard.  The technical issue of interest is heavily incorporated into the writing of the etude; for example, in the twelfth etude, the melody is stated over constant tremolos, whereas in the etude in the video above, the music is full of chords and passages where the pianist would need to rapidly displace his hands across half the length of the keyboard.  The intuition behind this is that through learning and performing the etude, the pianist perfects the execution of the technical difficulty of interest through sheer repetition.

But also, these etudes are studies disguised as concert pieces, as opposed to dry technical exercises by composers such as Czerny and Cramer that do not pretend to be anything but.  The technical challenges in each of these etudes are written in such a way that they add texture and color and therefore become integral to the music.  In the etude above, for example, the leaps in the central section (5:27 through 6:26 in the video) allow the pianist to seem like he is simultaneously playing in different registers, creating the impression of layers of sound; this, together with the large chords, gives a massive, orchestral feel to the piece.  Thus, successful performances of pieces like these not only require the technique, but a degree of musicality as well.

During my Spring 2013 jury, after a performance of "Alborada del Gracioso" from Ravel's Miroirs, one of the teachers evaluating me suggested that I do some Liszt in order to work on creating a fuller, more orchestral sound on the piano; he felt the large chords in "Alborada" were a little thin-sounding.  I brought this suggestion back to my teacher, and she gave this particular piece among a list of suggestions.  Given that I needed to work on getting a richer, more sonorous sound out of the piano and on wide leaps across the keyboard, I selected this piece, thinking it would be perfect for me.  Plus, out of all of the Transcendental Etudes, as well as the piano repertoire in general, this has been among my favorites ever since I listened to it for the first time when I was seventeen years old; it was something about the exotic harmonies and modulations, the way that the three main themes (0:47 through 1:10, 2:40 through 3:25, and 3:40 through 4:40 in the video) are interwoven, and the way it satisfyingly builds to its climax (6:27 in the video) and then fades away.

I was able to learn the notes and bring this piece up to performance standards in about four months.  This etude was not nearly as technically difficult as I thought it would be.  It was not exactly something that I could sight-read at tempo, but it certainly was not one of the more difficult of the Transcendental Etudes; the consensus among most pianists indicate that honor goes to either the fourth or fifth of the set.  Working on this piece was extremely useful for me; chords have never been my strong suit, as I often have been unable to play chords with the full sound they need, and I would often tire in dense chordal passages.  I can tell in more recent repertoire that having learned and performed this piece has helped; I noticed that I have been able to execute thick chords with better sound, and I do not tire as easily when I need to play a sequence of them.

Learning the Piece

I found it helpful, both for learning the notes to the piece and for mapping out the trajectory of the music, to divide the etude into seven sections:

Measures 1 through 37 (0:00 through 2:39): A slow, tonally ambiguous introduction where an introductory theme is developed, amidst fragments of the primary theme, which will make its first full appearance in the second section.

Measures 38 through 57 (2:40 through 3:36): A flowing section marked "poco piu mosso" (a little faster) that serves as the first full exposition of the primary theme.  The theme is stated in right hand chords over a rising and falling figuration in the left hand outlining open fifths.  Liszt instructs that the theme is to enter pianissimo, with the soft pedal depressed; the melodic line reaches higher and higher as Liszt instructs the pianist to release the soft pedal and the music builds to a fortissimo high point that abruptly fades to bare B octaves.

Measures 58 through 79 (3:37 through 4:50): An calm arioso that serves as the first full exposition of the secondary theme.  A rhythmically free statement of this theme occurs in the right hand over large rolled chords to be taken up by both hands, imitating the sound of a vocalist singing over an instrumental accompaniment.  Liszt indicated that this accompaniment is to sound harp-like and this section to be played "Piu lento con intimo sentimento" (slower, with intimacy and sentimentality).  The section ends in a recitative that resolves to an E major chord that ushers in the following section.

Measures 80 through 97 (4:51 through 5:26): a transformation of the primary theme.  This time, it is stated in the right hand as thicker chords with a heavier accompaniment consisting of repeated triplet chords and broken octaves (4:51 through 5:13).  Liszt marks this section "molto animato" (very animated) and gives the directions to play "trionfante" (triumphantly) and fortissimo.  The melodic line is gradually deconstructed into chromatic broken octaves, providing a transition to

Measures 98 through 119 (5:27 through 6:26): the most technically difficult section of the entire etude.  This section is a transformation of the secondary theme, at full volume and in a higher register than the first full statement in the third section, over a dense repeated chord accompaniment.

Measures 120 through 129 (6:27 through 6:45): a reprise of the fourth section, which leads to

Measures 130 through the end (6:46 through the end): a slow, quiet ending beginning with an interlude consisting of a fragment of the secondary theme over similar rising and falling figurations in the left hand as the ones in the second section and ending with a restatement of the introductory theme over large rolled chords that fades away to three simple D-flat major chords.

These seven sections are different enough in terms of technique required and character that I ended up trying to learn and polish one section at a time.

The first several measures of the introductory theme (top), primary theme (middle), and secondary theme (bottom).  As can be seen, much of the writing in this etude consists of large chords.  Source: imslp.org

As mentioned before, this etude is a study in chords, and unsurprisingly, every one of these sections contains an abundance of them, including large rolled chords spanning two octaves or more as in the introduction (0:00 through 2:39), central arioso (3:37 through 4:50), and ending (6:46 through the end), and unbroken chords.  Like with just about every other piece in the piano repertoire, learning the notes should start with slow practice at half tempo.  But while learning this etude, perfecting the attack of the chords is also extremely important.  All of the notes in each unbroken chord must sound simultaneously, particularly in the softer passages during the introduction and the first appearance of the primary theme (2:40 through 3:25), where any chord where the notes do not sound completely together are especially noticeable.  During each chord, the keys need to be depressed fully with firm fingers as well in order to give the chords fullness and sonority, even in these softer sections.  Also, these chords must be played with the weight of the entire arm rather than just the forearms and wrists; achieving as full or rich of a sound with the latter is very difficult.  Plus, doing the latter will quickly lead to fatigue in the denser chordal passages in measures 80 through 129 (4:51 through 6:45).

After learning the notes and perfecting the attack of these chords, the next step is to determine which notes are the important ones to emphasize and which are there for texture and coloration.  This is crucial for maintaining clarity in the dense passages of this etude.  In my performances of this etude, I sought to emphasize any statement or fragment of the introductory, primary, or secondary themes; most often, these ended up being the topmost notes of a sequence of chords, but on some occasions, such statements occurred in the tenor (0:47 through 1:09) or in the bass (1:30 through 1:40).  I also sought to make sure individual notes within such statements were connected to one another.  I would usually make sure each note within the theme were sustained enough so that their sound would last until the next note, while also being careful not to elicit a harsh tone.  

Meanwhile, for the rolled chords in measures 24 through 29 (1:41 through 2:07), 59 through 76 (3:40 through 4:40), and 143 through 153 (7:23 through 8:18), care must be taken to avoid what my teacher calls a "hit and run", namely when the pianist is so concerned about getting ready to play the next notes that he ends up glossing over the current ones.  Every note in each of these broken chords needs to have a robust sound, and this is  achieved by not only depressing the keys fully, but also using the arms to guide the fingers in each rolled chord rather than simply stretching the fingers from note to note.  In the introduction and ending, the topmost notes of these rolled chords also spell out the introductory theme; the top note in each rolled chord needs to be voiced over all the others, and they, like with the unbroken chords, need to be connected.  For the rolled chords, I also made sure the topmost notes were sustained enough to last until the next note in the theme.  

In measures 98 through 119 (5:27 through 6:26), on top of the chords from the preceding sections, Liszt introduces the second technical focus of this etude: leaps.  A full fortissimo statement of the secondary theme occurs here as right hand chords in the higher registers of the keyboard; in between each note of the theme, the right hand plays a repeated chord accompaniment one octave lower, requiring the pianist to rapidly displace his right hand to cover both the theme and the accompaniment.  Meanwhile, the left hand mirrors these leaps in contrary motion.

The leaps in the central section of the etude.  The notes with the upward stems in the right hand comprise the secondary theme, whereas the notes with the downward stems comprise the repeated chord accompaniment.  Source: imslp.org.

Perfecting these leaps also requires slow practice, but after weeks of repeating the same passage over and over again at half tempo, a pianist learning this piece will be able to almost intuitively sense exactly where he needs to strike.  These leaps require an attack from above the keyboard, where the pianist is to use the full weight of his entire arms to drop down upon the notes, as opposed to preparing the placement of his hands and then using his forearms and wrists to execute the chords; the faster tempo of this section allows no time for the latter, and the former guarantees a fuller sound without inducing the fatigue that the latter does.  Plus, the notes of the theme needs to have a sufficiently robust sound to allow them to sustain over the heavy chords in the accompaniment; thus, care must be taken not to "hit and run" the notes of the theme, especially given the strong temptation to do so with the very busy right hand.  Glossing over the notes of the theme in this manner will produce a much weaker sound and will most likely cause the melodic line to become buried in the thick textures.

Bringing the Piece Up to Performance Standards

After learning the notes and perfecting the attack of the many chords in this piece, the next step is to incorporate some musicality so that this piece sounds more like a concert piece and less like a technical study; again, this etude is meant to be a technical study disguised as a concert piece.  Liszt gave this etude, along with nine others in the set, programmatic titles; he had a very strong vision of the imagery that the music in this etude would conjure.  Pianists performing this etude should aim to invoke "Harmonies du Soir", or evening harmonies with their playing.  When I perform this, I try to invoke a musical depiction of a tempo run along the monuments here in Washington, DC on an evening in November.

An effective performance of this piece requires a firm understanding of the broad structure of this piece, specifically identifying the direction of the music within each individual section and of the entire piece.  Measures 1 through 37 (0:00 through 2:39) is an introduction that sets the context of the entire etude with hints of the themes to come.  I consider the piece to have two climaxes: the true climax during the third and final appearance of the primary theme in measures 120 through 129 (6:27 through 6:45), and a false climax during the first appearance of this theme in measures 38 through 57 (2:40 through 3:36), where the music builds to the high point (3:12) and then abruptly fades away.  The arioso in measures 58 through 79 (3:37 through 4:50) is a calm interlude, and measures 80 through 119 (4:51 through 6:26) is one long buildup to the true climax.  The last section, measures 130 on (6:46 on) closes out the piece.

The overall structure, along with Liszt's dynamics and tempo indications and the textures and harmonies, are useful for determining the character and color of each of these different sections.  For example, the primary theme enters in full three times during this etude, and as mentioned before, I consider the last of these three statements as the climax of the entire etude.  Thus, I make each appearance of this theme more intense than the last.  Whereas I try to achieve an airy feel in the first statement in measures 38 through 57 (2:40 through 3:36) by being a little more liberal with the rhythm, holding down the damper pedal through each phrase, and fully releasing each right hand chord before playing the next one due to the pianissimo marking, the open fifths outlined in the left hand, and the instructions to hold down the soft pedal, I play the second statement in measures 80 through 90 (4:51 through 5:13) more majestically and in strict rhythm, and the final statement in measures 120 through 129 (6:27 through 6:45) more ecstatically and pay less attention to the rhythm.  Meanwhile, the constantly shifting tonality and the unusual harmonic progressions in the introduction (0:00 through 2:39), and Liszt's request to play much of this part piano or pianissimo all indicate that this section is to sound hazy and dreamlike.   

Within each section, we need to consider how to shape the statements of the three main themes so that they have a clear trajectory.  For the introductory and primary themes, I typically shape each phrase within these themes according to their melodic contours, treating the highest note in each phrase as the destination; I incorporate a long crescendo through the notes preceding this high point and a diminuendo through the notes following.  For the secondary theme, I usually give the notes more equal emphasis with a more marcato (strongly articulated) attack.  Equally as important is how phrases within the themes relate to one another, and this is also crucial in giving a clear direction for the entire section.  For example, if a phrase is repeated one whole step higher, such as in measures 38 through 40 (2:40 through 2:49) and measures 42 through 44 (2:51 through 3:00), I give the repeat statement more intensity than the first.

In addition to changes in the loudness of the notes of each phrase, taking time, letting up the intensity, and playing with rhythm in well-chosen places are also very helpful in shaping the melodic lines.  For example, at the high point of the first entry of the primary theme (3:12 through 3:17), I elongate the right hand chords a little, making this feel like somewhat of an arrival.  I do something similar during statements of the introductory theme in the introduction and ending (1:41 through 2:07 and 7:23 through 8:18).  Some well-placed delays in the music can also enhance some important notes, making them ring a little more.  In measures 98 through 119 (5:27 through 6:26), I do not play strictly in time; after each note in the theme, I take a little time before beginning the following triplet repeated chords.  Doing this helps the theme resonate a little more in this passage.  I also delay a little before unusual chord progressions such as in measure 72 (4:25), drawing attention to the unexpected harmonies.  Dropping the intensity at appropriate locations within the heavy chordal passages can also help make the climaxes even more powerful; I bring down the volume to a mezzo forte in measures 112 and 113 (6:05 through 6:10) right before the transition to the final reemergence of the primary theme in measure 120 (6:27).  Doing this allows an even more noticeable crescendo toward the climax than had I played this entire section at full volume.

Critique of this Performance

The technique felt sound during the performance; most of my issues were minor musicality issues.  I felt the introduction could have been more hazy, and I could have provided even more dynamic contrast between the introduction and the first emergence of the primary theme in measures 1 through 39 (0:00 through 3:25) and the louder chordal passages.  I could have taken even more opportunities to breathe and play with rhythm in the central section, to really give this statement of the secondary theme here some power.

Overall, I was pleased with this performance.  This certainly will not be the last time I perform this etude in public; I enjoy performing this piece, and, like many of Liszt's other etudes, it is usually well-received by the audience.

Race Post-Mortem: Wineglass Marathon

This photo perfectly captures how we were feeling after this marathon.

Sunday, October 6, 2013
Bath, NY to Corning, NY
Difficulty score: 2/10 (0 for weather, 2 for course profile, 0 for altitude)

I finished the Wineglass Marathon in 3:19:31 (7:37/mile), more than 11 minutes slower than my PR in the USA Marathon this past March, even in spite of higher mileage during this training cycle, a significant PR in the Philadelphia Half-Marathon three weeks prior, and the Wineglass course supposedly being easier and faster than the USA course. The weather was the problem here; race day was the last day of an Indian summer that hit the Northeast. At the start of the race, it was 64 degrees Fahrenheit with 99% humidity; these were by far the most unpleasant conditions under which I had run a marathon.

Even though I had wanted to qualify for Boston, which would have required a 3:04:59 or better, or at least get a marathon PR, which would have involved beating 3:08:13, I was still reasonably happy with my performance. As frustrating as it was, the weather was completely beyond my control, and even in those conditions, I still managed a decent time. Besides, the conditions were simply not there for me to even get a PR; 7:10/mile pace, which was around my USA Marathon pace, felt like a noticeable effort even in the early miles. Perhaps if I went out at 7:25 to 7:30/mile pace, I might have been able to avoid the leg cramps and the wall during the last four miles, and that might have taken a few minutes off my marathon time. Regardless, I doubt I could have done much better than a 3:19:31 on that day.

Training

This time around, I tried to increase my weekly mileage. During the previous two marathon training cycles, I ran five days per week except during race weeks and the taper, where the Tuesday runs were tempo runs, Wednesday and Saturday runs were recovery runs ranging from five to six miles, Thursday runs were track workouts consisting of 400 meter, 800 meter, or 1200 meter repeats, and Sunday runs were long runs. Going against conventional wisdom, I took these long Sunday runs significantly faster than what most runners would recommend, running them at around 20 to 25 seconds slower than my target marathon pace; most training programs recommend doing long runs 10 to 20 percent slower than marathon pace, which in my case would have been about 45 to 90 seconds slower. I averaged 38 miles per week, hitting a peak of 45 miles per week. This all had yielded two 3 hour 8 minute marathons in which I was able to finish strong, running the last 10K at a noticeably faster pace than the first 20 miles.

For this cycle, I wanted to run six days per week; my plan had me averaging 46 miles per week, with a peak of 54 miles in the eleventh week. My week would look similar to before, except I would add a Friday recovery run ranging from four to six miles, and I would only do track workouts every third Thursday, rotating between them, hill workouts, and tempo runs. I had hoped the overall increase in mileage would be able to chop off those 194 seconds standing between me and a Boston Qualifier (BQ).

At first, all of the workouts were going fine with the exception of the long run. Like in the previous two cycles, I had targeted a long run pace of about 25 seconds slower than my target marathon pace. On my 12-mile and 14-mile long runs during the second and third week of training, I could not hit those paces; I simply thought that was a result of the late June heat and humidity of Washington, DC. It was not until my fifth and sixth weeks where I began to get concerned; these two weeks respectively ended with a 16-mile and 18-mile run, both in which I hit the wall. The 18-mile run was so disastrous that I had to cut it short at 17 miles after hobbling back to my starting point, and had to take the next day off from work because I felt sick. I was puzzled as to why I could not execute my long runs; I was able to finish them just fine last summer, even though the temperatures were hotter last year. After seeking some advice and insight, I decided to slow down my long runs to about 45 to 90 seconds slower than my target pace. The remainder of my long runs that cycle went fine. Looking back, I think it was the combination of the higher mileage plus the long run pace that had pushed me over the edge; when I had been running lower mileage, I might have had enough in me to run my long runs at that faster pace, but now that I was averaging about seven or eight miles per week more, I had to slow down my long runs to avoid overtraining.

The rest of the cycle went fine. My body adjusted to the increased mileage and I was able to complete pretty much all of the workouts on the plan. Also, summer 2013 was noticeably cooler than Summer 2012, which allowed me to hit some pretty good paces during my tempo runs and my target times during my track workouts. Initially I was concerned that I would lose speed from doing so much of my training at a significantly slower pace, but those worries disappeared three weeks prior to the Wineglass Marathon, when I ran my tune-up race, the Philadelphia Rock and Roll Half-Marathon, and finished with a 100-second PR in 1:25:25 (6:31/mile). Given my half-marathon time and my higher mileage, I was confident going into the taper period leading up to the marathon.

The Days Leading Up to the Event

The week before the marathon was stressful.  On the Sunday before the marathon, I caught a cold. I went home in the middle of the day on Monday to rest a little to expedite my recovery. Fortunately, I recovered more rapidly than usual, being able to run the marathon completely symptom-free.

The government shut down that Tuesday, which meant that I was forbidden to work lest I wanted to deal with a hefty fine, disciplinary action, and, according to some coworkers, possible jail time. During the first few days, no one was certain whether or not we would get paid retroactively. I was hoping we would, but others in Washington, DC told me not to expect it. Government workers did receive retroactive pay during the last shutdown in late 1995 and early 1996, but Congress had no Tea Partiers back then. I really did not want to have to dig into my savings. We did end up getting retroactive pay, but the first time I heard of this decision was the day after the marathon.

However, neither getting sick nor getting furloughed stressed me out nearly as much as the weather forecast for race day. As of Tuesday of that week, the forecast predicted a heat wave for the rest of the week, and it was not predicted to subside until after the marathon. The forecast underestimated how hot it would actually be that week, and the prediction for the temperature and humidity for race morning progressively got worse. That Friday, the forecast predicted 64 degrees that morning with 99% humidity. I was prepared to slow down for this race, but I really did not want to do that; I had trained hard for the past 16 weeks, and I wanted, at the very least, a PR as a reward.

Otherwise, the taper, the car ride up, packet pick-up, and the carb-loading dinner the night before all proceeded without a problem.

The Race

64 degrees and 99% humidity did not feel terrible. In fact, as I pulled into the parking lot of Phillips Lighting, I even felt a little underdressed in just a sleeveless T-shirt and shorts. As I stood at the start waiting for the gun to go off, I thought about my execution of this race and decided that I would proceed with my original plan rather than slow my pace for the high humidity.

The best word I could use to describe this course is desolate. The first three miles were through the streets of the town of Bath, but then for the next six miles, the course ran down New York State Route 415, a rural state highway running parallel to the Southern Tier Expressway. After a brief trip through Campbell, NY near around Mile 9, the course became sparse again, going down on some roads where the only things around were wilderness and the occasional house. The next sign of civilization did not come until around Mile 20, when we approached Corning, NY. The last 10K went through various neighborhoods of Corning, with the end at Centerway Square. Other than a few clusters of spectators in Bath, Campbell, and Corning, very few people watched or cheered for the runners. The size of the race was also limited to a few thousand runners, so after around Mile 5, we were also quite spread out. In spite of the sparsity of this course, the first 20 miles of it were at least quite aesthetically pleasing, particularly with the Appalachian Mountains in the background and the leaves changing color. I also noticed that even though the course had a substantial net downhill, there were quite a few substantial uphills that were surprisingly challenging.

The first two miles went according to plan. I eased into my target 7:05 to 7:10/mile pace, running the first mile in 7:21 and the second in 7:12. However, in subsequent miles, I had trouble staying in the 7:05 to 7:10/mile range; I was working, but I did not want to push any harder out of fear that it would cause me to hit the wall later. I figured I could try to make up the time later on in the race; for the time being, I would just try to stick to this pace. I also noticed I was passing quite a few people.

Meanwhile, I fueled using the same plan that worked for me during my previous two marathons, taking three Shot Blocks about every three to four miles starting at Mile 4 and getting Gatorade at every water stop. At the water stops, I even grabbed an extra cup of water and poured it over my head and on my forearms in an attempt to keep myself cool.

I eventually began to notice the effects of the high humidity. Even as early as Mile 11, I noticed people already walking, having clearly hit the wall. At Mile 13, a girl collapsed right in front of me and had to be carried to the nearest medical tent. At Mile 14, I began to feel tired. Some guy that I passed around Mile 2 that had been running behind me since passed me back around this point, and I was unable to keep up with him. At Mile 15, my pace slowed from 7:05 to 7:10/mile to 7:15/mile, and my pace kept getting slower with each subsequent mile. At around Mile 18, I realized I did not have it in me to pick up the pace and run the last 10K with a finishing kick similar to the ones I had for the last two marathons. Any hopes of running a PR that day quickly vanished.

At around Mile 23, my legs started cramping badly and I hit the wall. I finished the last four miles trying to run, but cramping up and needing to walk it off after about a quarter of a mile of running, and then repeating the entire process. At around Mile 24, the 3:15:00 pace group leader, who was running by himself, passed me. My paces for Miles 24, 25, and 26 were slower than 9:00/mile; at that point, I considered the possibility that I might not even beat the time from my first marathon, 3:23:36.   

I eventually finished, with a bicycle escort riding with me to the finish after she found me cramping up near the 25-mile mark. But when I arrived, I suddenly felt very lightheaded and had trouble standing. Some volunteers wheeled me to the medical tent, where I stayed for a few minutes, and after sitting for a little while, I began to feel better.

I waited for Thorne to finish and shortly afterwards, we boarded the bus back to the start.

Lessons Learned

Two important lessons came from this experience.  The first is that I will never win against Mother Nature. If race day turns out to be hot and/or humid, then I will need to accept that I will not be able to perform my best and there is nothing I could do about that except adjust my pace accordingly. This goes particularly for the marathon; I have actually been able to run a half-marathon PR in similar or worse conditions, as I did in Frederick last year, but the marathon magnifies the effect of every adversity at least tenfold.

I was not the only one whose performance suffered in this weather. On the bus ride back to Bath, Thorne and I talked to some of the other marathoners about their race. Many of them, just like I had, increased their mileage noticeably while training for this marathon and ran a substantial PR in their tune-up half-marathon. Yet they did not perform nearly as well as they did in previous marathons either. When I returned to the hotel room, I looked up my overall place. I had finished in 107th place out of 1687 finishers; in 2012, when it was 42 degrees and much less humid at the start, I would have finished around 144th place out of 1586 had I also run a 3:19:31 then.

On a related note, high humidity coupled with temperatures in the high 50s or low 60s is a particularly insidious combination. Those temperatures will not feel particularly warm at the start, while 99% humidity seems much less noticeable than temperatures in the 80s or above. This may trick runners into thinking they can maintain their normal pace in those conditions, until the effects of the high humidity suddenly take their toll at Mile 15 of the marathon.

The second lesson I learned from training for and running the Wineglass Marathon was that just because the marathon itself did not turn out as well as I hoped does not mean the entire training cycle was a waste. I did improve my fitness from increasing my weekly mileage and hitting faster tempo paces during this training cycle. That certainly showed during my half-marathon in Philadelphia three weeks before Wineglass in the form of a 100-second PR.

Next Steps

My running group has a marathon party the day of the Marine Corps Marathon, which this year will take place on Sunday, October 27. All runners who have run at least one marathon or half-marathon during the previous year are invited to wear some or all of their race medals. I will certainly wear my Wineglass Marathon medal. For those of you who have not seen it, it is a handcrafted medal made from the same type of glass used in wine bottles. It is definitely unique, and I plan on showing it off.

After I recover from this marathon, I have planned several short races to take advantage of the fitness I have gained from this training cycle. These include the AIDS Walk 5K at the end of October, which would hopefully be enough time for me to recover enough to the point where I could race a 5K, the Veteran's Day 10K two weeks later, and the Jingle All the Way 8K in December. Perhaps I could get a PR or two in these shorter distances.

And regardless of the unpleasant race day conditions, I still have not been turned off from marathons just yet. Quite the opposite, actually; I have already planned a marathon for the early spring. Hopefully, I could increase my mileage and speed up my training paces a little, and hopefully this time, the weather on race day would not nullify the effects of my more intense training over this past year. I will probably run the Shamrock Sportsfest Marathon in Virginia Beach, which takes place the day after the USA Marathon. I have already run the USA Marathon twice, and I'm in the mood for a change.

Race Difficulty

The difficulty of a race is related to how likely you are to run one of your stronger finishing times during it.  Major factors that determine whether a race is a PR-friendly one or one in which you will struggle to even finish include:

• Typical weather conditions on race day.  Races in which the temperature at the start is typically cool (high forties or lower for marathons, high fifties or lower for other races) are conducive to strong finishing time.
• The profile of the course.  The ideal profile for a strong race performance is flat, or slightly downhill, and straight.
• Terrain of the course.  Courses with substantial portions on technical trails will be more difficult.
• Whether the race takes place at altitude.  Courses at altitude will really increase the difficulty, particularly for sea-level creatures like myself.

The race difficulty score that I give to races is a numerical score that takes into account the above factors and summarizes them into a numerical score ranging from 1 (PR-friendly) to 9 (simply finishing the race is extremely difficult).  Interpretations of this difficult score is as follows:

• 1: PR-friendly.  These races feature courses are flat or slightly downhill, and feature few turns, take place at sea level, and are on asphalt.  Temperatures on race day are usually ideal.  Example: Cherry Blossom 10-Miler.
• 3: conducive to faster finish times.  These races have mostly fast courses with a few minor difficulties (e.g. mostly flat courses with rolling hills or hairpin turns).  Temperatures on race day are often ideal, although unfavorable weather may occur on occasion.  Example: Richmond Marathon.
• 5: faster finish times possible, although more difficult.  These races feature some very noticeable difficulties, but a faster finish time is still feasible with careful preparation and intelligent race execution.  Such races may, for example, feature some large uphill portions or take place in a city or during a time of year where warmer weather is typical.  Example: Rock and Roll DC Half-Marathon.
• 7: finishing times likely to be significantly slower.  These races feature some very prominent difficulties that often prevent most runners from achieving some of their faster finish times for that particular distance.  Such races may feature some extremely difficult elevation profiles (e.g. long or steep uphill portions), unfavorable weather, difficult terrain (e.g. large segments on gravel), high altitude, or combinations of the above.  Example: Capitol Hill Classic 10K.
• 9: simply finishing the race is extremely difficult.  These races extreme difficulties including severe elevation profiles, typical unfavorable weather on race day, difficult terrain, or high altitude.  Example: Kauai Marathon.

Even-numbered difficulty scores are used if I feel like a particular race is somewhere in between two of the above levels.  For example, a race that is generally flat and fast but may have a few terrain-related or weather-related difficulties such as long segments on loose gravel or strong headwinds may be given a score of 2; it is not completely PR-friendly, but definitely less difficult than a race with a score of 3.

One caveat: this difficult score assumes that you are adequately trained for a race of that distance.  Any race would be an extraordinary challenge to complete if you arrive unprepared.

For fun, I also include an index of all of the races I have run to date.

Index of Race Difficulty Scores

Last updated May 13, 2017.  Numbers in parentheses following the difficulty score indicate the year in which I last ran the race (that year's race is thus the basis for the score).  Races that are currently defunct (e.g. the Earth Day 5K in Silver Spring, MD) or have changed substantially since I have run them (e.g. Rockville Twilighter 8K in Rockville, MD) are not included in this list.

5K Races

Baltimore Running Festival 5K (Baltimore, MD): some hills.  Temperatures are usually conducive to a fast race, although slightly warmer temperatures have occurred in recent years.  4/9 (2007).

Crystal City Friday Night 5K Series (Crystal City, VA): a series of 5K races, with one taking place each Friday evening in April.  All races take place on the same a course, a generally flat one with some small rolling hills, particularly in the last mile.  Since this takes place in April in the evening, weather can sometimes be too warm for a faster 5K time.  3/9 (2016).

Run Around the Square 5K (Pittsburgh, PA): this course features some long continuous uphills in the second mile and extended segments on cobblestone and dirt.  This race takes place in the last weekend of August, which is often one of the hottest and most humid times of the year in Pittsburgh.  7/9 (2009).

San Francisco Pride Run 5K (San Francisco, CA): the course is a loop through Golden Gate Park.  The course features some rolling hills, but nothing too terrible.  The race takes place in late June, where mornings are generally cooler; the shade from the trees throughout also help.  3/9 (2008).

Shadyside 5K (Pittsburgh, PA): a flat, fast course with a few very minor fluctuations in course profile.  Temperatures are usually in the fifties at the start.  1/9 (2009).

SOME Turkey Trot 5K (Washington, DC): a flat, fast course along Freedom Plaza.  Temperatures are generally ideal for a race, although headwinds may make an appearance as they did in 2013.  1/9 (2013).

Walk to End HIV 5K (formerly the AIDS Walk 5K; Washington, DC): a flat, fast course along Freedom Plaza.  Temperatures are generally ideal for a race.  1/9 (2013).

10K Races

Capitol Hill Classic 10K (Washington, DC): this course through the streets of Capitol Hill feature some long continuous uphills, particularly toward the end.  Since this race takes place in May, weather is generally very warm and humid.  7/9 (2013).

The Great Race 10K (Pittsburgh, PA): this very large 10K race has a net downhill course profile with some moderately uphill segments in Mile 4.  The only factor that may prevent this from being a PR-friendly course is the weather; late September is still often warm and humid in Pittsburgh.  3/9 (2009).

Pike's Peek 10K (Rockville, MD): a very net downhill course, with some uphill segments in the third and fourth miles.  Weather on race day has not usually been an issue.  1/9 (2014).

St. Patrick's Day 10K (Washington, DC): this race coincides with much of the 5K course and takes place about an hour afterwards.  The course is generally flat with a few rolling hills.  Temperatures are usually in the forties or fifties on race day.  2/9 (2016).

Veteran's Day 10K (Washington, DC): a flat, fast course along Hains Point.  Weather on race day has been superb for the past seven years, with temperatures in the forties or fifties.  1/9 (2015).

15K Races

Jingle All the Way 15K (Washington, DC): a new race that begins at the Washington Monument, goes along Rock Creek Parkway, the Tidal Basin, and Hains Point, and ends at the location of the start.  This course overlaps with much of the St. Patrick's Day 10K and the Cherry Blossom 10-Miler, so it is generally flat with a few rolling hills.  The early December race date means that temperatures are usually in the forties on the morning of the race or possibly cooler.  2/9 (2016).

10-Milers

Army 10-Miler (Arlington, VA): generally flat and fast, with somewhat difficult segments along 14th Street Bridge and along Virginia Avenue (Mile 3).  Temperatures are usually in the fifties at the start.  2/9 (2016).

Cherry Blossom 10-Miler (Washington, DC): very flat course profile, with the exception of the hill on 15th Street at the very end.  Temperatures have not been an issue on race day in recent years; the only weather-related issue to appear since 2010 were the powerful winds in 2016.  The energy level also makes this a great course for a 10-miler PR.  1/9 (2017).

Half-Marathons


Baltimore Running Festival Half-Marathon (Baltimore, MD): first seven miles is essentially one long continuous uphill, but the last 6.1 miles are net downhill with some rolling hills thrown in.  Provided that runners are smart about conserving energy in the first half, a very fast second half is doable.  This race starts one hour and 45 minutes after the 5K.  5/9 (2011).

Colonial Williamsburg Half-Marathon (Williamsburg, VA): very hilly course, particularly in the second half, with much of the race taking place on trails.  The race takes place at 1:00 pm, which has resulted in some warm temperatures in the past few years, especially for February.  6/9 (2014).

Frederick Half-Marathon (Frederick, MD): this course features some rolling hills throughout and a very large one at Mile 12 right before the race track at the end.  Weather can be warm and humid as this takes place in May.  4/9 (2012).

Navy-Air Force Half-Marathon (Washington, DC): a generally flat course with a few rolling hills and a number of turns.  Weather can be a bit of an issue on race day; September in DC can be warm and humid.  3/9 (2015).

Rehoboth Beach Seashore Half-Marathon (Rehoboth Beach, DE): the course is very flat, although Miles 6 through 11 occur largely on crushed gravel paths.  Temperatures are conducive toward a fast time, although the humidity can be high.  3/9 (2014).

Rock and Roll DC Half-Marathon (Washington, DC): some rolling hills, particularly in Mile 7 through Adams Morgan and Mile 12, and one very large, steep hill leading out of Rock Creek Park.  Temperatures are usually conducive for a fast time.  5/9 (2015).

Rock and Roll Philadelphia Half-Marathon (Philadelphia, PA): this course features a few long uphill segments, particularly the one on Fairmount Avenue by the Eastern State Penitentiary.  Weather has been an issue a few times over the past few years; the 2014 and 2016 races took place on unseasonably warm and humid mornings.  4/9 (2016).

Shamrock Half-Marathon (Virginia Beach, VA): very flat course, although strong headwinds are regularly a problem, particularly along the stretch in Miles 7 through 8 through Fort Story.  Temperatures are usually in the forties at the start and typically do not get much higher than the fifties.  2/9 (2016).

Marathons

Baltimore Running Festival Marathon (Baltimore, MD): first three miles is essentially one long continuous uphill, and so are Miles 15 through 20.  Other segments are not particularly easy, such as the miles on the cobblestone right before the UnderArmour Headquarters.  But like the half-marathon, if runners execute the first 20 miles intelligently, a fast final 10K is doable.  The last ten miles of this race coincide with the last ten miles of the half-marathon; the marathon starts one hour and 45 minutes before the half-marathon, so marathoners will be running the last miles of their race with the half-marathoners.  5/9 (2012).

Boston Marathon (Boston, MA): the sum of many little things make this a tricky race, including the net downhill profile, the appearance of the hills between Miles 16 and 21, and the downhills in the last five miles that can thrash your quads even further.  The 0.7 mile walk from the Athlete's Village to the starting corrals can also add a bit of wear to your legs.  If executed correctly and if the weather cooperates, it is possible to run one of your faster times in this race, but the weather in April in Massachusetts can be unpredictable.  If warm and sunny conditions occur on race day, the lack of shade on this course often leads to very difficult conditions.  6/9 (2017).

Chicago Marathon (Chicago, IL): extremely flat course, except for the very small uphill at Mile 26 (Mt. Roosevelt).  Although the 2007 race is forever etched in marathoners' minds for its brutal heat and humidity, temperatures on race morning are often in the forties and fifties.  1/9 (2014).

Hartford Marathon (Hartford, CT): a somewhat fast course with some rolling hills throughout, but many turns and a hairpin turn at Mile 17 and two very cruelly placed long uphill segments in the last mile.  Weather typically is not an issue on race day.  3/9 (2015).

Richmond Marathon (Richmond, VA): generally flat with rolling hills, particularly in the first half, lengthy uphills, and a difficult segment across Lee Bridge (Mile 16).  Temperatures are usually in the forties at the start and typically do not get much higher than the fifties.  3/9 (2016).

Rock and Roll DC Marathon (Washington, DC): this race starts with the half-marathon and the first 20K coincides with the first 20K of the half-marathon course.  The second half is quite difficult, featuring a very hilly last 10K through the streets of Anacostia.  6/9 (2013).

Shamrock Marathon (Virginia Beach, VA): the second half of this marathon coincides with the half-marathon course, but the first half is just as flat but also shielded from the strong headwinds that often make an appearance on race day.  This race starts 90 minutes after the start of the half-marathon.  2/9 (2014).

Wineglass Marathon (Corning, NY): the course is net downhill with a few rolling hills.  Temperatures are generally conducive for a fast time, although higher temperatures and humidity have occurred on race day within the past few years.  2/9 (2013).