# Modelling Human Performance **URL:** \ **Category:** Uncategorized\ **Created:** [5 September 2022 07:42 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091 "2022-09-05T07:42:38Z")\ **Posts on this page:** 20\ **Page:** 1
**Author:** ![David\_Scheers](https://forum.intervals.icu/user_avatar/forum.intervals.icu/david_scheers/32/38707_2.png) [@David\_Scheers](https://forum.intervals.icu/u/David_Scheers)\ **Post date:** [5 September 2022 07:42 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/1 "2022-09-05T07:42:38Z")
[https://www.biorxiv.org/content/10.1101/2022.08.31.506028v1?s=08](https://www.biorxiv.org/content/10.1101/2022.08.31.506028v1?s=08) According to this article, the power-duration curve follows the power law, instead of hyperbolic, as the CP- model states. I’ve always found the CP model quite unrealistic for longer durations. It assumes I can hold 95% of my FTP almost indefinitely, which sadly just isn’t true 🙂 Maybe worth adding the option to plot de power curve according to the power-law ? ---
**Author:** ![david](https://forum.intervals.icu/user_avatar/forum.intervals.icu/david/32/30189_2.png) [@david](https://forum.intervals.icu/u/david)\ **Post date:** [5 September 2022 18:59 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/2 "2022-09-05T18:59:12Z")
That’s very interesting. Tx. I saw the paper come up on Twitter but hadn’t had a chance to look at it yet. ---
**Author:** ![Coach\_Ron](https://forum.intervals.icu/user_avatar/forum.intervals.icu/coach_ron/32/36042_2.png) [@Coach\_Ron](https://forum.intervals.icu/u/Coach_Ron)\ **Post date:** [5 September 2022 19:30 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/3 "2022-09-05T19:30:13Z")
Ah. Good ol’ Riegel strikes back. Basically this is what it is … time, t = ax^b , where a is a constant, x = distance, b is the fatigue factor. Some of my writings on this : > **[New Fatigue Factors for Running : Women's Road and Track Racing](http://www.georgeron.com/2017/09/the-fatigue-factor-for-running-womens.html)** > > A blog about endurance sports, climate and complex systems. My thoughts : 1. power law implies constant pacing is the way to race to your best performance. We know that idea, when put to test in events like the marathon, also falls apart. There is no such thing as “constant pace”… in the real world, we vary pace according to perceived exertion. 2. the authors’ model looks like a work in progress, according to paper. ---
**Author:** ![david](https://forum.intervals.icu/user_avatar/forum.intervals.icu/david/32/30189_2.png) [@david](https://forum.intervals.icu/u/david)\ **Post date:** [5 September 2022 19:52 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/4 "2022-09-05T19:52:04Z")
🙂 I should probably wait a bit to see what their updated model is. But it might be worth implementing Riegel for runners soonish … ---
**Author:** ![Coach\_Ron](https://forum.intervals.icu/user_avatar/forum.intervals.icu/coach_ron/32/36042_2.png) [@Coach\_Ron](https://forum.intervals.icu/u/Coach_Ron)\ **Post date:** [5 September 2022 20:02 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/5 "2022-09-05T20:02:55Z")
I do raise a question. If a model is supposed to help understand a real world phenomena, such as the drastic change in physiology at transition points along the intensity spectrum , and if the critical power model in being hyperbolic was able to describe that, how is it still called an artifact ? ---
**Author:** ![gbleach](https://forum.intervals.icu/user_avatar/forum.intervals.icu/gbleach/32/34156_2.png) [@gbleach](https://forum.intervals.icu/u/gbleach)\ **Post date:** [6 September 2022 10:37 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/6 "2022-09-06T10:37:09Z")
> [@Coach\_Ron](#): > > - power law implies constant pacing is the way to race to your best performance. We know that idea, when put to test in events like the marathon, also falls apart. There is no such thing as “constant pace”… in the real world, we vary pace according to perceived exertion. I’d be interested in reading more research on pacing if you have links to any. Also, I don’t know which pacing is optimal according to the CP model. ---
**Author:** ![Nikolai\_Stenfors](https://forum.intervals.icu/user_avatar/forum.intervals.icu/nikolai_stenfors/32/37912_2.png) [@Nikolai\_Stenfors](https://forum.intervals.icu/u/Nikolai_Stenfors)\ **Post date:** [13 August 2023 14:32 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/7 "2023-08-13T14:32:35Z")
The power-law model paper now published, perhaps something that could be implemented in Intervals.icu?: [Modelling human endurance: power laws vs critical power | SpringerLink](https://link.springer.com/article/10.1007/s00421-023-05274-5). Twitter/X-synopis: [Axel Finke on Twitter: “Critical power vs power laws in endurance sports: A brief 🧵 1/10 https://t.co/lqJ8WfgdRe” / X](https://twitter.com/AxelFinke/status/1690022339637903360) ---
**Author:** ![Flo\_Gen](https://forum.intervals.icu/user_avatar/forum.intervals.icu/flo_gen/32/29226_2.png) [@Flo\_Gen](https://forum.intervals.icu/u/Flo_Gen)\ **Post date:** [7 August 2024 14:12 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/8 "2024-08-07T14:12:58Z")
Any news about a possible implementation of a new feature capturing the power laws? ---
**Author:** ![GarageLab](https://forum.intervals.icu/letter_avatar_proxy/v4/letter/g/fbc32d/32.png) [@GarageLab](https://forum.intervals.icu/u/GarageLab)\ **Post date:** [7 August 2024 23:53 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/9 "2024-08-07T23:53:29Z")
Out of curiosity, how will you use the power law/fitting parameters for training and race planning? Of the models I have, the power law fit is pretty bad. But it could be useful to compare fitting parameters over time, especially the “fatigue” factor. ![image](https://forum.intervals.icu/uploads/default/original/3X/9/6/96d135f73c283c1953afdd6835cd5f1f498cb05b.png) ---
**Author:** ![iinrlv00](https://forum.intervals.icu/user_avatar/forum.intervals.icu/iinrlv00/32/47059_2.png) [@iinrlv00](https://forum.intervals.icu/u/iinrlv00)\ **Post date:** [13 December 2024 08:49 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/10 "2024-12-13T08:49:52Z")
Hi do you have any news for the implementation of the power-law fit for cycling? These days people is forced to use excels outside any platform, and will be a real difference maker for intervals.icu. ---
**Author:** ![Luisma\_Gallego\_Soy\_P](https://forum.intervals.icu/user_avatar/forum.intervals.icu/luisma_gallego_soy_p/32/26303_2.png) [@Luisma\_Gallego\_Soy\_P](https://forum.intervals.icu/u/Luisma_Gallego_Soy_P)\ **Post date:** [16 December 2024 11:31 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/11 "2024-12-16T11:31:06Z")
I created this, I don’t know if it will be useful to you, let me know if you need any modifications. ![image](https://forum.intervals.icu/uploads/default/original/3X/3/1/317f83b191b4727e3b323224ba7a653a554e309d.png) ![image](https://forum.intervals.icu/uploads/default/original/3X/e/b/eb09f2c32c8435abd9042117dec3eae6b8af9197.png) ---
**Author:** ![iinrlv00](https://forum.intervals.icu/user_avatar/forum.intervals.icu/iinrlv00/32/47059_2.png) [@iinrlv00](https://forum.intervals.icu/u/iinrlv00)\ **Post date:** [17 December 2024 08:59 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/12 "2024-12-17T08:59:09Z")
Thanks Luisma! It’s a custom chart? I can’t find it in the custom chart list. May I’m doing somthing wrong? ![power-law](https://forum.intervals.icu/uploads/default/original/3X/7/3/7309d34e5c2cb3b2592a7685a96f54b21b8a5f8f.png) ![custom_charts](https://forum.intervals.icu/uploads/default/original/3X/b/5/b5d613dfb47c14c6b3c012c0cc92d6a92391e33b.png) ---
**Author:** ![Gerald](https://forum.intervals.icu/user_avatar/forum.intervals.icu/gerald/32/28423_2.png) [@Gerald](https://forum.intervals.icu/u/Gerald)\ **Post date:** [17 December 2024 11:26 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/13 "2024-12-17T11:26:53Z")
It’s an activity-level chart: Open the activity, go to power tab Scroll to bottom Activity Charts Type in Power-Law ![IMG_7339](https://forum.intervals.icu/uploads/default/original/3X/d/0/d0a20d1f8594a114175d786c10a4b78741089aed.png) ---
**Author:** ![GarageLab](https://forum.intervals.icu/letter_avatar_proxy/v4/letter/g/fbc32d/32.png) [@GarageLab](https://forum.intervals.icu/u/GarageLab)\ **Post date:** [27 March 2025 23:39 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/14 "2025-03-27T23:39:23Z")
Just updating an older thread… I’m sharing a spreadsheet that has several models included to fit your mean maximal power curve and develop modeled power duration curves. Several of the models are typical models that you’ll find everywhere, such as the two and three parameter models. I also have some of the more modern models like the Omni-Domain Model, Extended Critical Power Model (found in Golden Cheetah) and a modified Perronet and Thibaut model (presumably similar to the WKO model). A power law model is also included. Hopefully these model implementations can help you estimate a value for critical power to use for training. This spreadsheet has not been quality checked by anybody and you should presume that there could be errors somewhere in the spreadsheet. This spreadsheet has been set up for Microsoft Excel, however you probably can get it to work in LibreOffice pretty easily. To use the spreadsheet, you’ll need to paste your mean maximal power curve into the summary sheet and then go sheet-by-sheet and use the built-in solver to come to a solution for the model fitting parameters. The sheet is set up to take the data from Intervals.icu CSV for your mean maximal curve of interest and paste directly into the sheet. Or you can use a mean max curve with up to 4,208 rows of data. You can automate this task, but I find that going model by model and fit by fit gives you a better idea of whether your model fits are reasonable or not. The sheet > **[MMP Modeling V2.3.xlsm](https://onedrive.live.com/personal/7aaa504c5ba71805/_layouts/15/Doc.aspx?sourcedoc=%7B77c35971-e5c9-4d4c-ad66-daf8be813223%7D&action=default&redeem=aHR0cHM6Ly8xZHJ2Lm1zL3gvYy83YWFhNTA0YzViYTcxODA1L0VYRlp3M2ZKNVV4TnJXYmEtTDZCTWlNQjJxcG5saW4zZTdjUE1vMmpxSkZ0amc_ZT05amRKOVo&slrid=e61a1ca2-2078-1000-c4d5-e14fb2e0c956&originalPath=aHR0cHM6Ly9vbmVkcml2ZS5saXZlLmNvbS86eDovZy9wZXJzb25hbC83YWFhNTA0YzViYTcxODA1L0lRQnhXY04zeWVWTVRhMW0ydmktZ1RJakFkcXFaNVlwOTN1M0R6S05vNmlSYlk0P3J0aW1lPWYxdldySjNHM2tnJnJlZGVlbT1hSFIwY0hNNkx5OHhaSEoyTG0xekwzZ3ZZeTgzWVdGaE5UQTBZelZpWVRjeE9EQTFMMFZZUmxwM00yWktOVlY0VG5KWFltRXRURFpDVFdsTlFqSnhjRzVzYVc0elpUZGpVRTF2TW1weFNrWjBhbWNfWlQwNWFtUktPVm8&CID=af6df9a0-2251-4088-ac43-0e77b5c12bbf&_SRM=0:G:82)** Sheets: - Model Summary - 2 parameter, 3 parameter – hyperbolic - 2 parameter – linear work - Ward-Smith - Omni-Domain, Omni-3 parameter, Omni-Exponential - Extended CP - Modified Perronet & Thibault - Power Law Summary sheet ![image](https://forum.intervals.icu/uploads/default/original/3X/b/d/bd545166acd7295e297863b9aec24e069dc90ac3.png) OMPD sheet ![image](https://forum.intervals.icu/uploads/default/original/3X/b/a/bafb4dbba361d0eac2ecc25675b0445acd394a53.png) More sheets and models coming… ---
**Author:** ![GarageLab](https://forum.intervals.icu/letter_avatar_proxy/v4/letter/g/fbc32d/32.png) [@GarageLab](https://forum.intervals.icu/u/GarageLab)\ **Post date:** [28 March 2025 20:47 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/15 "2025-03-28T20:47:18Z")
Here’s another spreadsheet. Mader’s basic model Its a very crude implementation the model from the 1986 paper. The sheet may have errors and I don’t have a bio/physio background, so consider this a simple numerical calculator, of sorts. The main references are in the sheet. Hauser, 2014 and Mader, 1986. I coded the sheet in **LibreOffice Calc** – it’s free and useful. There are some modifications you can make to have this in Excel (charting, VBA coding). _Input parameters_ Not sure where _vol. rel._ comes from - if you look at Incsyd reports, Vol. rel. seems to be based on a calc using body characteristics (height, weight, body fat, etc.) Seems _K”s_ is adjusted based on VLaMax magnitude, high = 1.728, low = 1.331 Where to get VO2Max? – Watch estimate, ASCM formula, 5m Cycling power correlations, Software (WKO5, GC), other correlations (ramp test, Cooper test, etc.) Where to get CP? – intervals, WKO5, GC, maybe TR (what model?). Use FTP also. Need good proxy for mlss. How to estimate VLaMax? – Back-calculate. Use spreadsheet to iterate by varying VLaMax to find CP/FTP (mlss) Fat and CHO correlations – I have no idea where they come from. The sheet: > **[MLSSc UpdatedV3.ods](https://onedrive.live.com/personal/7aaa504c5ba71805/_layouts/15/Doc.aspx?sourcedoc=%7B541697d7-63c2-4306-8161-9ac0de04e1ca%7D&action=default&redeem=aHR0cHM6Ly8xZHJ2Lm1zL3gvYy83YWFhNTA0YzViYTcxODA1L0VkZVhGbFRDWXdaRGdXR2F3TjRFNGNvQnNTWHVBSDh3SzhHa00zd0Q1TmZhd3c_ZT16dmpObkw&slrid=e61a1ca2-d03f-1000-c28f-8ca327fece32&originalPath=aHR0cHM6Ly9vbmVkcml2ZS5saXZlLmNvbS86eDovZy9wZXJzb25hbC83YWFhNTA0YzViYTcxODA1L0lRRFhseFpVd21NR1E0Rmhtc0RlQk9IS0FiRWw3Z0JfTUN2QnBETjhBLVRYMnNNP3J0aW1lPUlfcm5ySjNHM2tnJnJlZGVlbT1hSFIwY0hNNkx5OHhaSEoyTG0xekwzZ3ZZeTgzWVdGaE5UQTBZelZpWVRjeE9EQTFMMFZrWlZoR2JGUkRXWGRhUkdkWFIyRjNUalJGTkdOdlFuTlRXSFZCU0RoM1N6aEhhMDB6ZDBRMVRtWmhkM2NfWlQxNmRtcE9ia3c&CID=1d4103c6-6f72-424a-8084-db47db6441d0&_SRM=0:G:523)** A _very_ quick video showing the basic operation: > **[Microsoft OneDrive](https://photos.onedrive.com/share/7AAA504C5BA71805!52414?cid=7AAA504C5BA71805&resId=7AAA504C5BA71805%2152414&ithint=video&e=srbfnA&migratedtospo=true&redeem=aHR0cHM6Ly8xZHJ2Lm1zL3YvYy83YWFhNTA0YzViYTcxODA1L0VRVVlwMXRNVUtvZ2dIcS16QUFBQUFBQlQ2bldBM1J2Z0dlVmNWUk5obWs2dmc_ZT1zcmJmbkE)** ![image](https://forum.intervals.icu/uploads/default/original/3X/7/c/7cb557f7d3f4d94ef23700bf03755c7dcfed6760.jpeg) A better explanation of the model can be found at the link I always share: > **[Anaerobic Threshold via VO2max and vLamax v2](https://www.kaggle.com/code/rasaff57/anaerobic-threshold-via-vo2max-and-vlamax-v2/notebook)** > > Explore and run AI code with Kaggle Notebooks | Using data from No attached data sources ---
**Author:** ![Andi\_Sp](https://forum.intervals.icu/user_avatar/forum.intervals.icu/andi_sp/32/25896_2.png) [@Andi\_Sp](https://forum.intervals.icu/u/Andi_Sp)\ **Post date:** [29 March 2025 14:19 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/17 "2025-03-29T14:19:42Z")
Thanks for sharing. But it seems I can’t get it work. I have not these calculation buttons and I miss some graphs. Any idea? ![image](https://forum.intervals.icu/uploads/default/original/3X/7/9/79990c75907e9a0494ed7f156deed2e7d7d33dbe.png) ---
**Author:** ![ilbonvi](https://forum.intervals.icu/user_avatar/forum.intervals.icu/ilbonvi/32/78007_2.png) [@ilbonvi](https://forum.intervals.icu/u/ilbonvi)\ **Post date:** [29 March 2025 16:09 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/18 "2025-03-29T16:09:33Z")
yep, this doen’t work. Just like the CP one ---
**Author:** ![GarageLab](https://forum.intervals.icu/letter_avatar_proxy/v4/letter/g/fbc32d/32.png) [@GarageLab](https://forum.intervals.icu/u/GarageLab)\ **Post date:** [29 March 2025 16:28 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/19 "2025-03-29T16:28:10Z")
Let me find a better way to share the sheets. I believe OneDrive is the problem. ---
**Author:** ![GarageLab](https://forum.intervals.icu/letter_avatar_proxy/v4/letter/g/fbc32d/32.png) [@GarageLab](https://forum.intervals.icu/u/GarageLab)\ **Post date:** [1 April 2025 18:37 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/20 "2025-04-01T18:37:52Z")
I changed the location of the sheets to a Dropbox folder and checked the downloads on a separate computer. The files should work now. Just a quick note about the Mader sheet…make sure you modify the macro security settings for the file (if you trust it). Also, you’ll need a [64-bit JRE](https://www.java.com/en/download/manual.jsp) for goalseek. For some reason I can’t edit my posts, so here’s the updated link: [https://www.dropbox.com/scl/fo/fjlrvq52lbz8xav2mlzz1/ABboAKaYX7gB9QEY9ifosJM?rlkey=hzqddu37dow05cyl023ij0ahn&st=pxuf9jxg&dl=0](https://www.dropbox.com/scl/fo/fjlrvq52lbz8xav2mlzz1/ABboAKaYX7gB9QEY9ifosJM?rlkey=hzqddu37dow05cyl023ij0ahn&st=pxuf9jxg&dl=0) ---
**Author:** ![GarageLab](https://forum.intervals.icu/letter_avatar_proxy/v4/letter/g/fbc32d/32.png) [@GarageLab](https://forum.intervals.icu/u/GarageLab)\ **Post date:** [1 April 2025 18:38 UTC](https://forum.intervals.icu/t/modelling-human-performance/19091/21 "2025-04-01T18:38:37Z")
The CP spreadsheet has been updated a bit. I just added buttons on each sheet to automate the solver function. See the new link just above for the file. [Next page](https://forum.intervals.icu/t/modelling-human-performance/19091.md?page=2)