Wednesday, January 11, 2012

Thoughts on Scoring the Decathlon for Masters...


Haven't written in quite a while, I thought maybe I would start up again by putting my thoughts on decathlon scores for masters down on paper. It's easy for anyone to see by looking at the records and reasonably deep all-time lists that scores are lower for masters, even after age adjustments, than for the open-aged. The average world best using the age groups 40-75, which are the 8 age groups that really count for masters, is 8069. Now, 8069 is a decent score for a 25 year old, but it's nearly 1000 points (10.6%) under the open world record. When I look at the marks necessary for me in the 45 age group to score 9026, it seems inhumanly possible. For anyone to even have those 10 performances as PB's after 45 would be extremely difficult, let alone doing them all at once. Furthermore, I compared the total score for the world best mark by young decathletes in each decathlon event(decs over 8k) for the open guys and get 10319; for the same 8 masters age groups as above I get an average of 9200. Actually I was surprised it's even over 9k. The difference there is 10.8%. Furtherfurthermore, I compared a decathlon score using last year's best overall marks from each age group in each decathlon event between masters and the open bests, and the difference here is 11892 to an average of 10760, a difference of 9.5%.
My assertion is that the current age graded tables might be fine for single events(or each decathlon event taken singly), and certainly fine for fairly determining the order of finish for any particular competition, but it's not a sufficient or fair means for comparing a decathlon score by a master to an open score.
The reasons, in my opinion, are(in some/no particular order): 1) it's harder to train for 10 events as a master than when young. The older body is slower to recover from workouts, which affects every athlete past 40, but I would submit the effect is exaggerated for decathletes. 2) The functional career span for the open decathlete is around 14 years(20-34), but once you're over the hill, your body is constantly deteriorating. In other words, you're at your peak or near it for several years, but you're only 45 or 67 for one year. Although I haven’t calculated the average number of decathlons the young guys do at/near their peak, but certainly at least 20, and some do many more than that. Roman Sebrle did 39 decathlons over 8000 points between the ages of 22-32 We masters can only do at most 4-5 during the first two years in any particular age group. 3) Larger/heavier athletes tend to have a tougher time with the running and jumping events as they age, but young athletes overcome this with their energy and elasticity, and in fact use it to their advantage in the throwing events.

The first issue at hand is, why age grade at all? The answer is that in the higher age groups, some athletes weren't able to get any points for their performance in certain events using the open tables, so they would just start the event, but then not finish. This satisfies the requirement that each event be attempted, but ruins the spirit of the competition. Age grading allows the older athlete who wants to do a decathlon, and runs 22 seconds in the hurdles, to get some real points for the effort. Handicapping for age does make sense and lends some elegance and authority to the whole affair, and the people involved with doing the statistics and math to come up the current tables have been very diligent and come up with numbers that are probably meaningful. The alternative to age grading each event would be to just handicap the total based on the open tables, with some other kind of adjustment for an event where the athlete can’t score any points.
Before I submit the results of my own age-grading calculations , I'll admit one can also say: the best athletes in the world in their best condition do athletics in their 20's, and then they tend to stop. Very few top elite tracksters go on to work out and compete as hard as masters. This should explain some discrepancy between the open world record, and the over-40 marks. If everybody doing the decathlon as a master had scored over 8000 points as a youngster, then yes, the averages for masters scores would be higher. But I'm here to tell you Roman Sebrle will not be able to score an age-graded 9026 as a 45 year old, no matter how hard he trains, but I hope to God he tries because it would be fun to watch. One can see the result of the current age-graded tables on Sebrle’s 8109 from last summer at age 36: 8440, an adjustment of 331 points or 4%. Sebrle scored over 8440 20 times during his prime, so the question is: is the discrepancy due to a lesser training regimen, is Roman not aging well, or are the age adjustments just too low? I suggest the latter. Although the age-graded tables don’t start until 35, I think this is already well past the prime of the average decathlete. The last time someone over 26 broke the world record was Bill Toomey back in the 60’s, and frankly he had some extenuating circumstances that allowed that to happen, namely that he started the dec at later age and had health problems(in other words he might have had his best score earlier if he had started training for the decathlon in high school), Sebrle is one of only 4 guys to score over 8000 after 35, and I’ve found just 5 gentlemen that scored over 8K in their prime who have done decathlons after 40, so there’s not a lot of data to draw from on that front. There are, however, many hundreds of people that have done decathlons as masters, including many very good athletes, so in total there’s probably enough data statistically speaking to make accurate adjustments.

So the method I came up with is to establish a performance for each age group from open to 75 that represents the best that a decathlete should expect to perform in any particular event for that age group. This is for a world class decathlete in top shape and relatively balanced between all 10 events. No one athlete should expect to achieve all 10 of these marks in one decathlon. To create the baseline in the open class, I calculated the average of the best mark in the world in each event for each of the past 20 years by a decathlete who scored over 8000 points. That point total comes out to 9509. As an aside, I estimated what I thought the marks should be and came up with 9475. I’m just a decathlon nerd I guess.
Those marks, along with the rest of my calculations are on the linked spreadsheet. Then, for each age group starting with 35, I tried to come up with a comparable performance and make them fit on a graph in a relatively smooth curve. I started with the best mark that a decathlete in that age group had ever done in that event, and then adjusted from there to make it fit the curve better with consideration as to whether the mark was an outlier by comparing it to the second and third best marks in the age group event. Finally, I compared the graphed curve I had at that point to the curve of the world records for the event by age group.
I like the results of this method; while not perfect event by event for every person, overall, it's more satisfying to me intuitively. The average using my factors for the 8 age groups identified above comes to 8425. Still not 9026, but closer. I think this difference now can be attributed to the factors mentioned above. In the two youngest masters age groups, 35 and 40, the scores are now higher than the PBs for those guys as open athletes, but I don't have a problem with this. Kip Janvrin's ability at 35 and 40 in relation to his average performance during his peak career really is fantastic, and I believe his age adjusted scores should be the 8820 and 8816 that I calculated.
The factors are not meant to exactly "predict" what an athlete should have been capable in each event, but for the average of 10 events, there should be some correlation, or else what's the point to coming up with age factors at all?
I absolutely welcome any and all comments relating to my thinking on this, just keep the language clean!