Page 1 of 2

Time/distance predictor...

Posted: Wed Oct 24, 2007 9:30 pm
by Thomas W-P
Discussion in another thread has made me start this.

If you have times for some pieces, how do predict your time for another. I would like to add a utility for this...

There seem to be several ways.

1) MaxDev says split decreases by 5 seconds every time you double the distance.

2) John Glynn has an as yet un-described method...

3) I (as usual) have an over-complicated spreadsheet ("OpenOffice under ubuntu, natch.")

My way relies on the fact that if you plot power against distance with log scales you get a straight line. If you have two maximum effort pieces you can determine the gradient and intercept, and from that predict any other power for any other distance and from that get the time...

My way is very hard to code in javascript so I am open to offers for simpler ways.

I realise that most Free Spirits will not be enamoured with this thread!

Posted: Wed Oct 24, 2007 10:16 pm
by NWakeford
On the C2 program guide section 5.23 table 5.14 there is a pace predictor.

Posted: Wed Oct 24, 2007 10:27 pm
by kirbyt
Hey Thomas. I think predicting is very complicated. It's hard to know what condition you were in for each PB for instance. Also, it sure seems like some of us are more natural sprinters than others. When I compare my times with Petr's for example, mine seem to be better the longer the distance gets but his sprint times just leave me shaking my head. For everything over about 30 minutes I seem to only need to add about 3 seconds when doubling the distance instead of the fabled 5. For everything under 2k--actually I've never done maximal effort under 2k :oops: except maybe this latest ctc. And then we even have the added complication of height and weight that may skew things one way or another. But good luck with the spreadsheet anyway. :wink:

Kirby

Posted: Thu Oct 25, 2007 8:41 am
by Spamuel
Good luck with this Thomas. I think a relyable way of predictiong times may well be the 'Holy Grail' of indoor rowing.

There always seems to be several threads on the C2 site trying to do it. It's the second most poular topic I think behind ones best summed up as 'I think I'm great and want to tell others'

Posted: Thu Oct 25, 2007 9:03 pm
by Thomas W-P
OK, here is my go at finding the Holy Grail...

http://www.freespiritsrowing.com/fs/time_predictor.html

N.B. The two you enter must be truly maximal!

I am making this sticky in the hope that people will try it and give me feedback...

Posted: Thu Oct 25, 2007 9:49 pm
by Stan
I entered my 1k and mile pbs and it predicts 7:16 for 2k. Quite remarkably it predicts 19:19.6 for 5k and 23:28 for 6k, both of which are spot on for me. After that it departs from my actual times, with 40:30 for 10k and 1hour 29 for the HM.
If I put in my 5k and 10k times, then the predicted HM is accurate to within a few seconds.
If I put in my 2k and HM times then it predicts I should be a few seconds faster for 5k, 6k and 10k.
Clearly the closer the 2 maximal times are to the desired predicted distance, the more accurate it is. If you put in a short and long distance it you seem to get a good guide.
Interesting utility Thomas and I am sure some of us will have some fun with it.

Posted: Thu Oct 25, 2007 10:55 pm
by johnglynn
Mine is a spreadsheet, I posted in another thread about a year and a half ago .

http://www.freespiritsrowing.com/forum/ ... .php?t=267

It's designed mainly for long distances but works quite well on short distances also, for duckrunners 2 times I needed to work a bit backwards but it worked all the same .

Mine uses 5K pace as it's base . And normally needs at least 1 longer then 5K normal distance . I started developing a version that could use any of the times as a base, but it was becoming very complex and something else took my interest :roll:

Took a look at your Javascript there Thomas, good standard of code, do they teach programming in physics these days .

Posted: Fri Oct 26, 2007 1:46 am
by NWakeford
Thomas that's wicked - you know what would be really cool, if the Free spirit personal best comparator, could include predicted times for distances not yet covered?

Posted: Fri Oct 26, 2007 9:40 am
by Thomas W-P
Thanks for your comments....

John: thank you for your comments re: code... Remember my past as a management consultant in IT though. :D nothing to do with teaching. (I had a complete DOH moment this morning thoguh and the code does not look as neat).

Nick: Interesting suggestion - I will have a look...

Stan: nice to see it gets is broadly right...

Everyone: I have updated the predictor this morning to adjust for the always skewed 500m times. You can play with the adjustment factor but 2.0 seconds seems about right.

Posted: Fri Oct 26, 2007 10:56 am
by paulgould
Excellent tool Thomas - quite accurate when predicting 2k from 1k and 5k(2.5 secs out) - nowhere near when using 500m and 1k maximal efforts as predictors(probably due to my 500m being way quicker than my 1k - 12secs/500m).
Will have a bit of a play and find out what efforts I SHOULD be producing.

Paul G

Posted: Fri Oct 26, 2007 11:11 am
by Thomas W-P
It will never be 100% accurate but it interesting none-the-less I think.

I twigged that I can predict timed pieces too with a spot of iteration so have added that now.

My work here is done for the moment.

Posted: Fri Oct 26, 2007 12:58 pm
by alistairkbs
Thomas

Fantastic tool. Great effort.

Unfortunately it still predicts a sub-20 5K that I have yet to do but thats my mental problem

It worked very well plugging in 1K and 60 mins and certainly showed where I should be able to improve

Thanks again

Alistair

Posted: Fri Oct 26, 2007 2:51 pm
by kirbyt
Thomas, that's very interesting. I noticed that my predicted pace for the marathon would be 1:57.1. This seems reasonable to me and about what I had in mind. However, when I input the 2k time that I'd like to be doing by February, my marathon time slows down a fair bit. Is this because I'm changing my profile toward the sprinters?

Kirby

PS: I used my 10k as the other input.

Posted: Fri Oct 26, 2007 2:54 pm
by johnglynn
NWakeford wrote:Thomas that's wicked - you know what would be really cool, if the Free spirit personal best comparator, could include predicted times for distances not yet covered?
Mine does this to an extent Nick, I wrote it for long distances (as they were giving me the biggest problems at the time) , but it is pretty accurate all the way down .

Long Distance Pace Estimator.xls (CLICK ME)

Fill in your Long distance times (into yellow bits) and put the new distance into the orange (current 8848) box, the green areas will have your estimates . It needs a 5K and one (preferably more) other time input .

Posted: Fri Oct 26, 2007 4:07 pm
by 300lb
Hi Thomas,

I guess it was my question about predicitng 2k time from 1k and 5k PB's that spurred you on??

Well I think congratulations are in order, very clever stuff. I got myself mentally ready and went for my first 2k proper this afternoon. Firstly, what a horrible horrible experience, secondly from my 1k and 5k times you predict 7:31.7 the result 7:32.3. correct to within 6 tenths =D>

Not sure how you done but thanks all the same.

Only 32 secs to find!! :shock: Although I'm quite pleased given I only started this lark on the 9th October and sometimes the stairs got me out of puff!

Thanks again

Paul.

Posted: Fri Oct 26, 2007 6:32 pm
by Thomas W-P
300lb wrote:I guess it was my question about predicting 2k time from 1k and 5k PB's that spurred you on??

Well I think congratulations are in order, very clever stuff. I got myself mentally ready and went for my first 2k proper this afternoon. Firstly, what a horrible horrible experience, secondly from my 1k and 5k times you predict 7:31.7 the result 7:32.3. correct to within 6 tenths =D>

Not sure how you done but thanks all the same.

Only 32 secs to find!! :shock: Although I'm quite pleased given I only started this lark on the 9th October and sometimes the stairs got me out of puff!
It was your question - and it has been an interesting challenge, luckily it coincided with half term...

And to be honest I am amazed :shock: (and pleased :D) that it was so close.

As for 32 seconds - I made similar enormous improvements after getting off the couch. My initial target (I started in June 2006) was the U12 world record (7:14) and it took a little while to get there. The other day I reeled three of those off in a row with 5 minutes rest in-between which shows how much you can progress.

Posted: Fri Oct 26, 2007 7:15 pm
by Thomas W-P
johnglynn wrote:Mine is a spreadsheet, I posted in another thread about a year and a half ago .
John, I still cannot get it to work - I plug in 5000m and 60 minutes but nothing magical happens. Do I need to add more figures?



Everyone: I have added the pace predictor to the utilities...

Posted: Sat Oct 27, 2007 4:20 pm
by Oliva
Thomas,

Nice. Interesting tool. =D>
I tried it. And my next race on 2k will be 6:40,5 ... 8) I used 1k and 5k inputs. We will see...

Posted: Sun Oct 28, 2007 1:16 am
by kirbyt
Petr said:
And my next race on 2k will be 6:40,5 ...
Go Petr Go! I think 2K is going to be the next CTC. We're both going to go sub 6:40. :twisted:

Kirby

Posted: Sun Oct 28, 2007 9:12 am
by Thomas W-P
Go for it boys. This means of course that if Petr does it I have to too now that I am coming back to fitness.

BTW - I posted on the c2 forum about my new tool and it has had some interesting (and brain bending) responses...

http://www.concept2.co.uk/forum/viewtopic.php?t=15583

Posted: Sun Oct 28, 2007 12:58 pm
by Stan
Stan wrote:I entered my 1k and mile pbs and it predicts 7:16 for 2k.
Looks like it was pretty accurate yesterday Thomas. Only a second or so out.

Posted: Thu Nov 01, 2007 10:01 pm
by Gregor Andrews
Thomas W-P wrote:I think someone said once that it is all about the pacing ;)

Using my predictor suggests you should be aiming for 7:30 at the moment and a 1:52 split. Though of course if your 1,000m and 5,000m times are very weak...
Interesting predictor Thomas. I managed a 7:18.3 2k last month so I should manage better than that now.
I haven't attempted the 500m or 1000m more than once this season - even more incentive to now to get the best out of the predictor.

(TWP quote taken from CTC thread but I'll post in here as its more relevant)

Posted: Thu Nov 01, 2007 11:39 pm
by beegee
I put in my 1k and 5k - it predicted 7:41 :cry:
I put in my 1k and 10k - t predicted 7:40 :cry: :cry:
I don't want to do this
Brian

Posted: Fri Nov 02, 2007 8:03 am
by Thomas W-P
I think there is too much mystique about the 2k. Every piece hurts in its own unique way... Go for it!

Posted: Fri Nov 02, 2007 11:27 pm
by Gregor Andrews
Based on my 1k and 6k SBs:
"Your predicted time for 2000m is 7:11.6 @ 1:47.9"
Will give it a go later on this week. :)