You can listen to the audio by clicking above, or download it here
Part 1 of understanding the ranking systems of USAT and 3 other major triathlon federations, and how you can maximize your USAT ranking through effective race scheduling and strategy. You will be weighed, you will measured, but will you be found wanting? Let’s find out today, on Tri Talk.
Welcome to this special edition of Tri Talk your podcast source for triathlon tips, training, news and more. Welcome to the new listeners who have recently sent me e-mails from Denmark and Japan. Your stories of how you have managed to stay in shape over the northern hemisphere’s winter are inspiring. Outstanding dedication. My goal at Tri Talk is to help you swim, bike, and run faster, and in this case, to help you get ranked higher, to meet your personal triathlon goals. Whether you are an elite or amateur triathlete, we cover sprint distance to Ironman distance. I’m your host, David Warden, and this is Tri Talk Episode 66.
Thanks for joining me in this special episode of Tri Talk. I want to make a few things clear about this episode before you invest your time in listening. First, although my goal is always to help you swim bike and run faster, this episode will not necessarily make you swim bike or run faster. It will help you appear as though you are swimming, biking and running faster by influencing your ranking within your federation. Second, this episode will really only benefit members of USAT and to a much lesser extent members of British Triathlon, Australia Triathlon, with a brief nod to Triathlon Canada. Or, if your triathlon federation uses a ranking system similar to one of these federations, you may still get some benefit out of it. I primarily bring up the other federations’ ranking systems so that we can compare them to USAT. Also, this information is mainly for individuals who race for glory and ranking, and not for achievement or affiliation. So all you achievement and affiliation types can skip to the Josh Grobin tracks you’ve got on your iPod. Although I will be covering the ranking systems used by Triathlon Canada, British Triathlon, and Australian Triathlon, they will each take about 1 minute, and the USAT system will take the remainder of the podcast. This is not because I harbor over-nationalistic sentiments for my home country, it really is that the ranking systems for the other English-speaking federations are that simple, and the US ranking system is freaking complex. It makes the US college football BCS ranking system look like 2+2.I hope you’ll forgive the US focus, but with 70% of Tri Talk listeners in the US, this is relevant information for a vast majority of Tri Talk listeners, and I hope you’ll allow this brief focus on my home triathlon federation. My purpose for this episode is to help you maximize your 2009 USAT ranking through knowledge of how the system works, race selection, and a race strategy which compliments the ranking system.
Episode 66 will be released in 2 parts. This episode will focus on understanding the ranking systems. Episode 66a, released in another week, will give you very detailed strategies for race selection and strategy in order to maximize your USAT ranking. For those of you outside the USAT system, Episode 67, released in March, will go back to an international focus to help all of us swim, bike, and run faster.
Before we get onto the good stuff, I had some significant response to episode 65. As I suspected, Professor Jenks corrected me on a clear error. Regarding my theory on why my Zipp wheels had the same roll down time as my Mavics, my thought was that because the Mavic’s were heavier they would obviously carry me faster down the hill. However, Professor Jenks kindly reminded me that an object’s mass does not influence how fast it falls, but rather the density. Mass does matter when trying to use force to overcome gravity, but it does not matter when falling. Therefore, the weight of the wheels does not really explain the Zipp vs. Mavic times during a rolldown test. Unless I could prove that the Mavic wheels were somehow more dense. This is possible, since the wheels are very similar in size and shape, and one reason that the Mavics are heavier even with very similar dimensions could be that portions of that wheel are more dense. Kudos to my friend Paul from Southern California as well who pointed this same thing out to me. However, Paul, as one of my former clients, you have broken the sacred rule of never questioning your master. You are now dead to me.
Regarding my numbers around the potential benefit of the legal draft, Tri Talk listener Jay feels that the benefit from passing cyclists can only be realized if you can keep your momentum during the pass. Clusters of cyclists, or side-by-side riders that cause you to brake or stop pedaling prior to the pass could easily neutralize your legal draft benefit. He goes on to say that even when being passed, if the lead cyclist passes you very slowly, you might have to brake in order to get back into a legal draft zone within 15 seconds.
Let’s get onto the good stuff! Starting with the ranking system of British Triathlon. Unlike USAT, the British Triathlon Ranking Series is only based on 5 events. Being somewhat smaller than the US geographically, this makes sense as you can rank everyone based on the exact same races head-2-head, and the travel burden is not as heavy if you were to do that in a physically larger country. Plus, you only have to race at 3 of the 5 events. This is true for both the triathlon and separate duathlon series.
To determine points, after each of those 5 British Triathlon (BT) events, all non-federation members are stripped out of the results. I’m not sure I agree with that, but you are essentially only using other BT members to determine points. The BT athlete who has a time closest to the 40th percentile is identified as the pacesetter time, and that individual gets 100 points. Meaning, 39% of the field was faster than them, and 59% of the field was slower.
All other BT members get points based on that pacesetter time using the formula 100 x the pacesetter time divided by their actual time. For example, if 90 minutes was the pacesetter time that came in closest to the 40th percentile, an individual with a time of 60 would get 150 points, which is 100×90/60. An individual who finished in 120 minutes would get 75 points, or 100×90/120. At the end of the year
In can summarize BT’s ranking system in 3 sentences: In each of the 5 series races, determine the pacesetter time by taking the time of the 40th percentile among BT members only. Points are then individually allocated for that race at 100 x pacesetter time / actual time. Final ranking is determined by adding up the total of the athlete’s best 3 races in the series of 5. 3 sentences. Piece of cake.
Australia is even easier. Like BT, Australia’s Triathlon Series only offers a select number of races, in past years it has been 6 events. While BT and USAT are based on time. Australia is based on placement in your age group. With up to 25 points allocated to the first 25 in each age group at those 6 specific events, it is very straightforward. First place in the age group gets 25 points, 10th place gets 15 points, 25th place gets 1 point. Also like BT, only current members of AT are awarded the points. If a non-AT member finishes 1st in an age group, the next AT member to finish in that age group would get the 25 points.
In can also summarize AT’s ranking system in 3 sentences: In each of the 6 series races, 1-25 points are awarded inversely to the athletes’ age group finishing spot in each age group. Non-AT athletes do not receive points, and those points roll down to the next AT finisher. Final rankings are a sum of all event points.
It is interesting to note 2 more things about the ATS. First, in 2008 those points were not really used to determine a final ranking, but were used to determine a separate ITU Selection points and who would represent Australia at the World Championships in Vancouver. Second, I don’t see any literature that AT will be continuing their ATS point system in 2009, which would be a shame. Perhaps one of our fine Australian listeners can let me know the status of the ATS in 2009. For now, I’ll assume that they will.
Canada has the easiest system of all, and I can summarize theirs in one sentence, nay 4 words. They don’t have one. They used to have one, but discontinued it in 2007. Far be it for me to criticize my great neighbor to the north, but you only have a few thousand Triathlon Canada members, right? If you sent me your data, I could run your rankings system myself. For a country that pumps out an overwhelmingly disproportionate number of world-class triathletes for their population, this is a glaring inconsistency in the way Triathlon Canada is run.
OK, now for USAT. Before I go into detail on USAT’s ranking system, I’m actually going to give you the simplest explanation possible first.
After any USAT sanctioned event, identify which participants received an overall ranking in the previous season and define them as pacesetters. Convert all participants’ times into real numbers representing minutes to the fourth decimal place. Multiply each pacesetter time by their previous season’s pacesetter ranking and divide that by 100. Drop the top 20% and bottom 20% of pacesetter times. Average the remaining pacesetter times to determine the event par time. Divide the par time by each participants’ finish time and multiply that by 100. An additional 10% will be added to the female participants’ scores. Year end points are determined by averaging the inverse value of each of an athlete’s top 3 events, inverting that resulting average again and multiplying that number by 100.
I am not kidding when I say that is the shortest and simplest way to describe the USAT ranking system. If you could understand it from that brief description, you can skip the next 10 minutes. For the rest of us mortals, let’s dig in. I promise that you will impress your tri friends when you show them you actually understand the system. Here we go.
In any given USAT sanctioned event, not just specific events like BT or AT, and this applies to all triathlon and duathlon distances, the ranking first identifies which participants have a ranking from the previous year. These participants are going to be the pacesetters for that event. It is important to note that only participants who have a ranking from the previous year will be used to determine the future par time, although all participants will be able to have a point value associated with their time at the end of the event. Easy so far. Next, convert all participants’ times into real numbers representing minutes. For example, a Sprint-distance time of 1:04:37 would be converted to 64.6167 minutes.
Here is where it gets interesting. Take each pacesetter’s numerical time in minutes and multiply that by their previous year’s ranking, dividing that result by 100. For example, our same athlete who finished a Sprint in 64.6167 minutes has a 2007 USAT ranking of 89.9764. A quick reminder that your USAT ranking is a number between 1 and about 110, with higher being a faster athlete. So we take their finish time of 64.6167 and multiply it by their 2007 ranking of 89.9764 which and divide by 100, which equals 58.1398. Why is that number important? That number is called the par time.
Taking an athlete’s finish time and multiplying it by their previous ranking gives us an idea of how fast someone with a ranking of 100 could finish that event. That result of 58.1398 means that based on that one athlete’s race result and previous ranking, we could predict an athlete with a ranking of 100 to finish in 58.1398 minutes, which then becomes par for that event. This was the part where I stared at computer monitor for 20 minutes with a blank look on my face. If you can get this concept in the next minute, you are way ahead of me.
Let me give a better example of this. From here on our I’m going to strip out the decimals and leave the numbers as integers to make this very simple. Let’s look at two athletes, once fast and one slower. We’ll use our same athlete who was ranked 90 in 2007 and finished in 65 minutes. Another athlete was ranked 73 and finished in 81 minutes. Taking the faster athlete’s information, we take their finish time of 65 times their ranking of 90 divide by 100 and we get 59. The second athlete is ranked 73 times a finish time of 81 divided by 100 also equals 59. Because a fast athlete has a low numerical value for a finish time, but a high numerical value for a ranking, the result is almost the same as an athlete with a high numerical value finish time and a low USAT ranking. In this case a finish time of 65 times a ranking 90 is the same as a finish time of 81 times a ranking of 73. Think of the previous years ranking in this formula as the handicap. The lower the ranking value, the greater the handicap. And that is why we can reasonably predict how fast an athlete ranked at 100 could finish the race based on other athletes times and rankings. That athlete ranked at 100 essentially has a 0 handicap. That prediction becomes the par time for that event.
OK, we are about 2/3 of the way through. Stay with me.
That process is repeated for all ranked athletes in the event. If there were 300 athletes in a race, and 250 of them have a ranking from the previous year, we take 250 athlete’s finish times, multiply it by their ranking, and divide that by 100.
Next, we drop the top and bottom 20% of those par numbers. So, if we had 300 participants, with 250 ranked athletes, we would strip the top and bottom of those remaining 250 athletes after the par calculation and are left with 150 values. It is important to note that we do not strip out the top and bottom 20% of all times. We strip out the top and bottom 20% of the results after the time multiplied by ranking result. This means we are not dropping the best and worst overall times, we are dropping the best and worst performances which significantly over-performed or under-performed based on their previous ranking. Athletes with high rankings but very poor performances or athletes with low ranking but unusually high performances are stripped out of the calculation of par time.
So, in our example, what started out as 300 athletes, is whittled down to only 150 inputs based on ranked athletes only and dropping the top and bottom 20% of par times. The remaining average of those 150 par times becomes the official par time for that event. That whole first process was just to determine the par time for a race.
Now that we have a par time for an event, the rest is a lot easier. In our example of an athlete with a 65-minute finish, let’s say that the par time for this event was still 59 minutes after we averaged all 150 par times. We divide the par time by the finish time and multiply that by 100. 59 divided by 65 times 100 is 90.7692. We finally have a score for that athlete for that race! We repeat that process for all 300 participants. Although we stripped out non-ranked and top and bottom 20% to calculate the par time, once we have that par time everyone in the race who finished gets a score.
In order to understand this ranking system, it is very important to note that we divided the par time by the finish time and NOT the finish time by the par time. The results would be totally different. 59 minutes divided by 65 minutes gives us our USAT score of 91, but dividing the finish time by the par time would have resulted in a score of 110. Why don’t we use that higher number instead as the point total? Remember, USAT could have done this, and instituted a scoring system where the lower the number the better the ranking. But, we are talking about a goal of a higher number being a better ranking. As their times go down, their ranking goes up and the USAT ranking you get is an inverse relationship you’re your overall time. This inverse relationship means that to get a true average of an athlete’s top 3 events for a year, we have to invert their points one more time.
Let me give you another example of what I mean. If one athlete finished in 60 minutes, and another finished in 72 minutes, that second athlete was 20% slower than the first, right? 12 minutes is 20% of 60 minutes. We could give the fast athlete a ranking of 100 and the slower one a ranking of 120. But, since we want a slower time to be a lower number, we invert 1.2 with a result of 0.83, or 1 divided by 1.2. This will become important later on.
Once that athlete does at least 3 USAT sanctioned triathlons or 2 USAT sanctioned duathlons, they are eligible for a year-end ranking. Only their top 3 triathlon event points count toward their year-end ranking, and top 2 duathlon events. Could it now be as simple as taking the average of those 3 triathlon events to get a year end score? Of course not!
For example, lets say that our athlete’s top 3 triathlon event points were 91, 90, and 94. Again, USAT will round to 4 decimals, but I am going to keep it simple. We are not going to average 91, 90, and 94. We are going to invert those 3 numbers and multiply by 100 to get 1.10, 1.11, and 1.06. We average those 3 numbers and get 1.09. We then invert 1.09 to get the final year-end ranking of 92. Frankly, the process of either averaging the 3 point totals, or taking the top 3 and inverting, averaging, and inverting again is very close, but they are different numbers and USAT is right to do the calculation that way.
You may find that it will be easier to understand this if you read it rather than listen to it, and I’ll have the transcript of this podcast available on PowerTri.com.
Is anyone still there?
Sweet, now we know exactly how the ranking system works. If you had enough time, you could calculate your ranking from each event long before USAT publishes them. That would require you looking up the previous year’s ranking for each and every athlete in the race you did, so you would have to be very motivated.
I’d like to share some thoughts on the USAT system. Frankly, I went into this research not understanding the system at all, and I was determined to hate it. This started out as a planned expose on the faults of the USAT system, and by the time I was finished, I was a raving fan, and I’ll tell you why.
First, The USAT system rewards for speed, unlike a ranking system that ranks based on age-group placement. You could be 15 minutes behind 10th place in your age group, but you get the same points as someone who came in 1 second behind 10th place. I don’t like that.
Second, The USAT ranking system allows for a fairly accurate comparison. Unlike BT and AT, where the total points at the end of the year mean nothing in terms of the ability to predict one athlete’s performance to another, with USAT, I can confidently say that athlete A is 6% faster than athlete B.
Third, I perceive USAT’s system as having a smaller margin of error. For example, in a small race with 50 people, the par time for BT is based on only one individual, while the USAT system would include up to 30 athletes to determine par time.
Fourth, it is resistance, not immune, but resistant, to race topography. Because it is based on an average of the field, external factors that influence all athletes equally will not effect your points. If there is a killer wind and you end up going 3 minutes slower than you normally race, the par time of the entire event will go up as a result.
Finally, if we didn’t use this system, I’d be curious to know what system we would use instead that would be better? Of course the current system has flaws, but can you recommend to me a system that works better? Actually, I would be interested in hearing a method that you think would be better. Surely, we have some statisticians out there who have some ideas.
The one thing I would change is that I would break down the rankings into distance categories. The ranking system seriously favors short course racing. You can race 12 sprints a year, and taper for 8 of them and have a tremendous chance at getting 3 outstanding performances. Athletes who race at the IM level will do maybe 2 full and 2 70.3 in a year, which means they can only afford one bad race. This is why I would encourage long distance racers to throw in a few short events, or at least schedule some short course for after your last long event.
I should disclose that I am the Regional Vice Chair for USAT for the Rocky Mountain Region, and that could be interpreted as tainting any endorsement I give to USAT. As Vice Chair for my region, I don’t work for nor am I compensated for my USAT work. It is an elected position representing the members of my region. My endorsement of the USAT ranking system is based on it’s merits, and not on my affiliation with USAT.
That’s it! You made it through the USAT ranking system!
Episode 66a, released in another week, will give you very detailed strategies for race selection and strategy in order to maximize your USAT ranking. This episode will be much shorter than this episode, but don’t miss it.