I'm a data junkie - interesting none the less (Page 2)
-
No new posts
Moderators: k9car363, alicefoeller | Reply |
|
2009-10-21 6:34 AM in reply to: #2469487 |
Champion 7595 Columbia, South Carolina | Subject: RE: I'm a data junkie - interesting none the less Interesting analysis. It is important, of course, (I'm not saying that this point has eluded you) that there are correlations -- indeed, causal connections -- among the three numbers. (It would be interesting to see what those correlations look like, and to see how those correlations are themselves correlated with overall results.) I think someone mentioned Macca's race report, which illustrates one sort of causal connection that could exist, among many. So we can't jump immediately from these correlations to some conclusion like 'stronger runners do better than stronger swimmers'. One reason, among many, is that the events happen in a certain order, and performance in one is obviously affected by what happened earlier. Consider two strong swimmers. One goes all out on the swim and is not able to recover for a while on the bike. Another takes it easy on the swim and comes out fresh. Etc., etc. (Many of these points are more relevant the longer the race is.) I think that information about the run is especially 'tainted' in this way. (I am 100% sure that the run is my strength (insofar as I have one), but my results, especially in longer races, mostly don't reflect this fact because I have yet to get the pacing exactly right for the other parts of the race.) |
|
2009-10-21 7:01 AM in reply to: #2470000 |
Master 2946 Centennial, CO | Subject: RE: I'm a data junkie - interesting none the less JorgeM - 2009-10-20 6:31 PM Could the data just be representative of how in general triathletes are terrible swimmers given the little emphasis they place on swim training? My guess is that if you would do the same analysis for pro racing (i.e. ITU racing) a different picture would emerge. I'd tend to disagree here. I would think we put in a tremendous amount of time/emphasis on swim training, but I also think that the impact/improvement of swim vs bike/run is less significant. For example. A ave/poor swimmer swimming 2:20/100 may spend significant time working on their swim only to reach 1:45/100 (40+ second improvement) This improvement in a 750 yrd swim is less significant than a ave/poor cyclist @ 16.5 mph improving to 20 mph on a 15 mile bike ride. (Feel that good swimmer is 1:20/100 and good cyclist 22 mph is just an opinion.) Improvement in the swim as Daremo stated will keep you in the race, but will not win it for you. |
2009-10-21 7:40 AM in reply to: #2470452 |
Coach 10487 Boston, MA | Subject: RE: I'm a data junkie - interesting none the less velocomp - 2009-10-21 7:01 AM JorgeM - 2009-10-20 6:31 PM Could the data just be representative of how in general triathletes are terrible swimmers given the little emphasis they place on swim training? My guess is that if you would do the same analysis for pro racing (i.e. ITU racing) a different picture would emerge. I'd tend to disagree here. I would think we put in a tremendous amount of time/emphasis on swim training, but I also think that the impact/improvement of swim vs bike/run is less significant. For example. A ave/poor swimmer swimming 2:20/100 may spend significant time working on their swim only to reach 1:45/100 (40+ second improvement) This improvement in a 750 yrd swim is less significant than a ave/poor cyclist @ 16.5 mph improving to 20 mph on a 15 mile bike ride. (Feel that good swimmer is 1:20/100 and good cyclist 22 mph is just an opinion.) Improvement in the swim as Daremo stated will keep you in the race, but will not win it for you. who is "we"? certainly my experience with the majority of triathletes is that their swim training is more often than not minimal. About the bolded portion, you also assume swim training will only yield improvements in swim speed; that is not the case. You might gain some speed during the swim portion but coming out of the water fresh will yield plenty of speed for the bike and run portion. Anyway, my point stands, IMO the data is more representative of how little emphasis most triathletes place in swim training hence they avg slow times. Anyway, how much time do you devote to swim training just out of curiosity? |
2009-10-21 7:49 AM in reply to: #2469487 |
Champion 9600 Fountain Hills, AZ | Subject: RE: I'm a data junkie - interesting none the less Very interesting, although I think conclusions are difficult to draw. I went back and did a quick survey of my own training. On a weekly basis based on time, swimming represents about 15-20% of my total weekly training time. For example last week I swam 3 hours out of 17.5 hours of S/B/R. It is my weakest sport in terms of OA placing, although I tend to be at the top of my AG in the swim and I win my AG at most races. |
2009-10-21 7:54 AM in reply to: #2470000 |
Champion 9407 Montague Gold Mines, Nova Scotia | Subject: RE: I'm a data junkie - interesting none the less JorgeM - 2009-10-20 9:31 PM My guess is that if you would do the same analysis for pro racing (i.e. ITU racing) a different picture would emerge. Here you go Jorge; from Beijing, male and female athletes. Shane (finish.png) Attachments ---------------- finish.png (12KB - 5 downloads) |
2009-10-21 10:31 AM in reply to: #2470540 |
Extreme Veteran 541 Colorado | Subject: RE: I'm a data junkie - interesting none the less gsmacleod - 2009-10-21 6:54 AM Here you go Jorge; from Beijing, male and female athletes. Shane Wow! From this the run was decisive while swim and bike were effectively irrelevant. That seems odd. I've done lots of these for distances up to HIM and have never seen such a scatter for the bike position vs overall. Edited by KirkD 2009-10-21 10:35 AM |
|
2009-10-21 10:57 AM in reply to: #2469487 |
Elite 5316 Alturas, California | Subject: RE: I'm a data junkie - interesting none the less Alright there is just a lot of mud in the swim variance. I wonder why that is. Even considering that it might not be a linear relationship it still looks bad. Only a multiple regression coeffiction would drag out what little contribution it has. However, I will still work on swimming, because I want to become a better swimmer. But ya the take away is work on the run and bike more than the swim for sure in total hours. Run foucs this year, bike focus next year. |
2009-10-21 11:00 AM in reply to: #2469487 |
Master 1790 Tyler, TX | Subject: RE: I'm a data junkie - interesting none the less In non-drafting olympic distance events (where the swim is proportionately longer) do the data suggest that swimming is more important than in other distances? Brian |
2009-10-21 11:05 AM in reply to: #2471076 |
Extreme Veteran 541 Colorado | Subject: RE: I'm a data junkie - interesting none the less Baowolf - 2009-10-21 9:57 AM However, I will still work on swimming, because I want to become a better swimmer. But ya the take away is work on the run and bike more than the swim for sure in total hours. And I'll still be going to my Master's swim class twice per week. As I mentioned before, I wasn't trying to suggest that everybody slack off on swimming. My original intent was strategy. I still stand by my original strategic conclusions that during an event, I can save some energy (and stress) on the swim without jeopardizing my overall placing. Everyone who took this as reason to back off swim training - double laps for you!! |
2009-10-21 11:09 AM in reply to: #2471001 |
Champion 9407 Montague Gold Mines, Nova Scotia | Subject: RE: I'm a data junkie - interesting none the less KirkD - 2009-10-21 12:31 PM Wow! From this the run was decisive while swim and bike were effectively irrelevant. That seems odd. I've done lots of these for distances up to HIM and have never seen such a scatter for the bike position vs overall. However, here are the differentials (male/female): Swim - :58/1:14 Bike - 1:31/4:50 Run - 7:17/8:16 Shane |
2009-10-21 11:11 AM in reply to: #2471118 |
Extreme Veteran 541 Colorado | Subject: RE: I'm a data junkie - interesting none the less gsmacleod - 2009-10-21 10:09 AM KirkD - 2009-10-21 12:31 PM Wow! From this the run was decisive while swim and bike were effectively irrelevant. That seems odd. I've done lots of these for distances up to HIM and have never seen such a scatter for the bike position vs overall. However, here are the differentials (male/female): Swim - :58/1:14 Bike - 1:31/4:50 Run - 7:17/8:16 Shane I'm not sure I understand what the differentials are. Could you elaborate, please? |
|
2009-10-21 11:12 AM in reply to: #2469563 |
Pro 4054 yep, | Subject: RE: I'm a data junkie - interesting none the less Daremo - 2009-10-20 4:29 PM I've faked my way through sprints, olympics and halfs with swimming. And while you cannot win the race on the swim, you can certainly lose it! I've missed out on overall and podium spots because I swim for sh-t. But I'm a strong cyclist and runner. Triathlon is one sport. Swimming is just a part of it that determines how well you do overall.
This is the biggest part. As you can see in many races the swim dictates very little on your overall placement. |
2009-10-21 11:26 AM in reply to: #2471126 |
Champion 9407 Montague Gold Mines, Nova Scotia | Subject: RE: I'm a data junkie - interesting none the less KirkD - 2009-10-21 1:11 PM I'm not sure I understand what the differentials are. Could you elaborate, please? Sorry, the differentials are between the fastest split and the slowest split. So, IMO, what this shows is that an athlete that turns themself inside out to stay with the pack will, more often than not, have a long day on the run. Shane |
2009-10-21 12:00 PM in reply to: #2471184 |
Champion 10018 , Minnesota | Subject: RE: I'm a data junkie - interesting none the less Thanks for the info. This is interesting to me, although I am struggling to read the charts (so embarrassed!). Anyway, I get the drift. I experienced this myself at an early season sprint this year. I STRUGGLED with the swim (i.e old fashioned panic attack, swam head out of water the whole way), yet it was only 400 yards or so and therefore only ended up about 1-2 minutes behind the main group. Fortunately I am a better than MOP cyclist and had a good run that day and ended up placing better than usual. And more importantly better than my competitive friends who usually beat me due to the swim I realized that as a slow swimmer, shorter races would always be better for me as the ratio of bike and swim to run was much greater. Although I will say that even at an Oly distance, my poor swim very much negatively effects me. I can come out of the water 15 minutes after the leaders. Big difference and too much to make up for me, even with the bike of my life... This is all fine with me because I don't care enough to do the swim training required to improve much more. Just thought I would share the anecdote... and I realize now that when you get outside of the fast group it seems that the swim can have a bigger impact. |
2009-10-21 12:18 PM in reply to: #2471320 |
Master 2460 | Subject: RE: I'm a data junkie - interesting none the less Actually, BikerGrrl - you'd do much better in a full IM where the swim is a very small percentage of the overall race time. In sprints/olys, the swim is a significant portion of the race. |
2009-10-21 1:35 PM in reply to: #2471001 |
Master 2380 Beijing | Subject: RE: I'm a data junkie - interesting none the less KirkD - 2009-10-20 11:31 AM gsmacleod - 2009-10-21 6:54 AM Here you go Jorge; from Beijing, male and female athletes. Wow! From this the run was decisive while swim and bike were effectively irrelevant. That seems odd. I've done lots of these for distances up to HIM and have never seen such a scatter for the bike position vs overall.Shane Perhaps the run was only decisive due to its position as the last discipline? In order to truly test this, it is imperative that we gather the participants of IM China and have them race several more times, to cover all permutations. (Swim/bike/Run, S/R/B, B/S/R, B/R/S, R/S/B and R/B/S) |
|
2009-10-21 2:09 PM in reply to: #2471635 |
Extreme Veteran 541 Colorado | Subject: RE: I'm a data junkie - interesting none the less moondawg14 - 2009-10-21 12:35 PM KirkD - 2009-10-20 11:31 AM gsmacleod - 2009-10-21 6:54 AM Here you go Jorge; from Beijing, male and female athletes. Wow! From this the run was decisive while swim and bike were effectively irrelevant. That seems odd. I've done lots of these for distances up to HIM and have never seen such a scatter for the bike position vs overall.Shane Perhaps the run was only decisive due to its position as the last discipline? In order to truly test this, it is imperative that we gather the participants of IM China and have them race several more times, to cover all permutations. (Swim/bike/Run, S/R/B, B/S/R, B/R/S, R/S/B and R/B/S) Excellent idea. I'm certainly available to analyze the data once you've gathered it. Thanks! 8^) |
2009-10-21 2:39 PM in reply to: #2471635 |
Extreme Veteran 591 New Port Richey, FL | Subject: RE: I'm a data junkie - interesting none the less moondawg14 - 2009-10-21 2:35 PM KirkD - 2009-10-20 11:31 AM gsmacleod - 2009-10-21 6:54 AM Here you go Jorge; from Beijing, male and female athletes. Wow! From this the run was decisive while swim and bike were effectively irrelevant. That seems odd.Shane Perhaps the run was only decisive due to its position as the last discipline? Not so odd when you consider this data is referring to a draft-legal race. |
2009-10-21 4:32 PM in reply to: #2469487 |
Veteran 183 Bellingham, WA | Subject: RE: I'm a data junkie - interesting none the less I'm also a data junkie, and ran some numbers on the results that I had on my computer (female Boise 70.3 finishers, 333 total). Although, rather than looking at raw finishing place, I used a metric I called "percent behind leader." I figure that while there may not be a significant difference between the 1st and 2nd place finisher based on rank, there could be a HUGE difference in the time gap. Essentially, my % behind leader was a way to "normalize" the swim results. Also probably could have used pure difference between the times, but I had some Excel formatting issues. I would be interested in seeing the ITU race results filtered using this method, since I suspect that with the packs that form, there is a stronger correlation with % behind leader than rank.I put a linear as well as a 3rd order polynomial fit on the curve just to see what it looks like. Interesting that the 3rd order is a pretty good fit. Kind of suggests that the FOP and BOP has a stronger correlation than MOP.The shading is by run time. Red being fastest, blue being slowest. Interesting that the fastest swimmers also tended to be the faster runners (also held true for bike times). Note the woman that led the swim (26 min) finished MOP. (Boise Results.JPG) Attachments ---------------- Boise Results.JPG (83KB - 5 downloads) |
2009-10-21 4:38 PM in reply to: #2472102 |
Extreme Veteran 541 Colorado | Subject: RE: I'm a data junkie - interesting none the less Interesting metric, FeS. Do you happen to have graphs of the same for bike and run? |
2009-10-21 5:10 PM in reply to: #2472118 |
Veteran 183 Bellingham, WA | Subject: RE: I'm a data junkie - interesting none the less The swim does show more scatter than either the bike or the run. All three do show a correlation though. ETA: shading on these is by age because I forgot to change the marker Edited by FeS 2009-10-21 5:12 PM (Boise Bike Results.JPG) (Boise Run Results.JPG) Attachments ---------------- Boise Bike Results.JPG (69KB - 13 downloads) Boise Run Results.JPG (71KB - 11 downloads) |
|
2009-10-21 5:28 PM in reply to: #2469487 |
Extreme Veteran 541 Colorado | Subject: RE: I'm a data junkie - interesting none the less The bike graph is very tight. I think it is interesting that they all follow a 3rd order curve shape. I'm not sure what kind of conclusion to draw there. I suppose the easiest explanation is that the FOP has a tight correlation for obvious reasons, the MOP having enough scatter to flatten the curve while not having dramatic changes in time differences, and BOP representing the really slow folks that cause an upward shift on the Y-axis while maintaining the relative position overall. Essentially the delta in times becomes larger as you move through the overall positions at the BO-MOP and BOP causing the upward inflection. |
2009-10-21 7:34 PM in reply to: #2469487 |
Veteran 274 Everett, WA | Subject: RE: I'm a data junkie - interesting none the less Very interesting. You've given me hope. I always assumed that the bike split would be the biggest determining factor in final placement. It appears the run has nearly as significant an impact, so my better than average run speed might be more beneficial than I previously thought. |
2009-10-21 8:00 PM in reply to: #2469487 |
Veteran 307 Montréal | Subject: RE: I'm a data junkie - interesting none the less Very interesting Kirk! Another thing that would be interesting, from a duathlete's point of view at least, would be to see if we observe the same phenomenon with the first run of duathlons than with the swim of a tri. I have the impression that the results would tell me to improve my run... Lol |
2009-10-21 10:28 PM in reply to: #2472420 |
|