Predicting Your Ironman Performance from Your Best Marathon

Predicting Your Ironman Performance from Your Best Marathon

Have you ever thought about how your marathon training could be transferred to an Ironman triathlon? As a veteran marathoner, you have endurance, mental strength, and specific run training—suitable bases for triathlon success. But just how close are you to becoming an Ironman?

The marathon-to-Ironman transition is much more doable than one might think. The goal of this blog is to shed light on the interaction between marathon performance and Ironman outcomes, to give a data-based attempt at predicting your Ironman run times and overall Ironman time through marathon achievements, to guide you to patterns that will hopefully guide your training and race expectations through actionable insights.

TLDR:

  • For men, each second added to their marathon time translates to an approximate 1.204-second increase in their Ironman run time. This strong predictor reveals a direct multiplier relationship between marathon and Ironman run times.
  • For women, each second added to their marathon time results in an approximate 1.159-second increase in their Ironman run time. This strong predictor also indicates a direct multiplier relationship between marathon and Ironman run times.
  • Use the various charts and tables below based on your gender, age group, and marathon performance to see what your Ironman times may be!

Why This Analysis Matters

Understanding the relationship between marathon and Ironman performance is more than just an academic exercise; it has implications for the athlete at almost every level. Here's why you should care:

Maximize Training Efficiency: Understanding how your marathon performances can predict your eventual Ironman time, you can individually tailor training schedules and training areas, leading to the most significant performance gains.

Set Realistic Goals: Whether you're aiming to finish your first Ironman or trying to achieve a personal best, a solid prediction can set realistic goals and help you manage expectations for the day.

Boost Confidence: Now that you know what you can likely pull off according to your marathon performance, you enter the world of triathlons pumped.

Race Strategy: The analysis gives personalized insights in which the user should plan or implement their race to use the strengths and deal with the weaknesses of the study.

From an understanding of these relationships, you can intelligently choose your training and racing strategy. The journey from marathoner to Ironman athlete is epic, full of thrills and spills. If you are confused about the many different and often contradicting pieces of advice on the web about triathlon training, this blog should give you a basis to navigate this with confidence. From these streams of data, we unlock insights that can truly change your training approach and race day performance. Whether you are a newcomer to triathlon taking your first steps into the world of endurance sports, or a veteran marathon runner looking for a new challenge, this article offers a blueprint for your Ironman triathlon aspirations.

A previous analysis, titled 'Personal Best Times in an Olympic Distance Triathlon and a Marathon Predict an Ironman Race Time for Recreational Female Triathletes' by Christoph Alexander Rüst and Beat Knechtle (published in The Chinese Journal of Physiology, June 2012), found that personal best times in an Olympic distance triathlon and marathon were strong predictors of Ironman race times. This study underscored the relevance of marathon performance as a key indicator for Ironman success, aligning well with our current findings.

Stay tuned as we go deeper into the specifics, providing in-depth analysis, predictive models, and actionable advice to help you cross the finish line with confidence and pride.

Data Collection Methodology 

  1. Ironman Dataset Collected using a custom Python Scraper written in selenium, From Ironman Results Websites and Endurance Data
  2. Marathon Datasets Collected using custom python scraper written in selenium, pulling from athlinks.com

If you are interested in obtaining the scraping tool from me, please reach out to me via the contact form on this website. For details on how the dataset was constructed and my Athlete Matching Methodology, please visit the end of the article in the appendix.  

Dataset Descriptive Graphics

I constructed a final dataset of 912 results for detailed analysis, consisting of 711 male and 201 female athletes. The input datasets were from 432,407 Ironman 140.6 Finisher Results and a subset of large marathons totaling 166,844 Marathon Finisher Results. Additional marathon data could be used to make the models even more robust, but given the sample sizes, I felt this was enough data to post this analysis as is.

 

Men (red) have a slightly faster average Ironman Over and Ironman Run Time.

I also conducted a manual investigation of many finishers, including some at the pointy end of the data. I identified the following amazing athletes: Ben Fuqua, Cory Mayfield, Lars Petter Stormo, Niels Esmeijer, and Jeff Dorrill. In addition, I validated the analysis on five people I know personally and conducted a survey on reddit r/IronmanTriathlon. The Reddit results line up perfectly with the models I've created.

In the next section I will share some reference tables that you can quickly use to look up how people typically performed given various Marathon performances. 

How to Use the Tables Below

The tables provide a summarized view of Ironman finish times based on open marathon finish times. It's designed to help athletes predict their potential Ironman performance based on their marathon times. Here's how to use it:

  1. Identify Your Gender:

    • The first column indicates the gender. Make sure you refer to the appropriate table based on your gender.
  2. Find Your Marathon Time Bin:

    • Locate the range in the "Open Marathon Bins" column that includes your marathon finish time. For example, if your marathon time is 3 hours and 20 minutes, you would look in the bin labeled "[03:15:00, 03:30:00)".
  3. Review Predicted Ironman Times:

    • Median Ironman Time: This column shows the median Ironman finish time for athletes whose marathon times fall within the specified bin. The median provides a central value, reducing the impact of outliers.
    • Average (Avg) Ironman Time: This column shows the average Ironman finish time for the same group. The average gives a general idea of what to expect, though it might be influenced by extreme values.
    • Standard Deviation (SD) Ironman Time: This column indicates the variability of the Ironman finish times. A lower standard deviation means the times are more consistent, while a higher standard deviation indicates greater variability.

Stand Alone Marathon to Ironman Total Time - Example of Usage

Suppose you are a male athlete who completed a marathon in 3 hours and 20 minutes:

  • You would look at the table for men.
  • Find the bin "[03:15:00, 03:30:00)".
  • The median Ironman finish time for this bin is 11:54:02.0, the average Ironman finish time is 12:07:28, and the standard deviation is 01:21:16.

This means that, based on historical data, a male athlete with a marathon finish time between 3 hours and 15 minutes and 3 hours and 30 minutes typically finishes an Ironman in around 12 hours, with a standard deviation of about 1 hour and 21 minutes.

Table to predict Ironman Overall Time from Open Marathon Time for Men

Gender Open Marathon Bins Median Ironman Time Avg Ironman Time Standard Deviation Ironman Time
Male [02:30:00,02:45:00) 09:40:01.5 09:36:42 00:44:07
Male [02:45:00,03:00:00) 10:30:41.0 10:41:02 01:03:58
Male [03:00:00,03:15:00) 11:27:20.0 11:38:14 01:16:34
Male [03:15:00,03:30:00) 11:54:02.0 12:07:28 01:21:16
Male [03:30:00,03:45:00) 12:32:54.0 12:34:29 01:28:43
Male [03:45:00,04:00:00) 12:47:45.0 12:59:16 01:35:37
Male [04:00:00,04:15:00) 13:39:35.0 13:36:25 01:26:21
Male [04:15:00,04:30:00) 13:50:08.0 13:59:12 01:26:14
Male [04:30:00,04:45:00) 14:22:47.5 14:29:37 01:13:23
Male [04:45:00,05:00:00) 14:54:31.0 14:42:36 01:11:30
Male [05:00:00,05:15:00) 15:45:43.0 15:17:22 01:15:36
Male [05:15:00,05:30:00) 15:17:52.0 15:02:45 01:16:42
Male [05:30:00,05:45:00) 15:13:03.0 15:13:52 01:14:24
Male [05:45:00,06:00:00) 14:57:11.0 15:18:18 01:10:43
Male [06:00:00,06:15:00) 16:03:59.0 15:49:35 01:04:37
Male [06:15:00,06:30:00) 16:26:34.0 16:26:34 00:07:33

Table to predict Ironman Overall Time from Open Marathon Time for Women

Gender Open Marathon Bins Median Ironman Time Avg Ironman Time Standard Deviation Ironman Time
Female [03:00:00,03:15:00) 10:32:43.5 11:19:21 02:12:16
Female [03:15:00,03:30:00) 11:57:49.0 12:08:15 01:05:08
Female [03:30:00,03:45:00) 12:46:01.5 12:45:35 01:14:10
Female [03:45:00,04:00:00) 12:56:57.5 13:00:32 01:25:23
Female [04:00:00,04:15:00) 13:15:44.0 13:17:12 01:02:44
Female [04:15:00,04:30:00) 13:40:50.0 13:44:21 00:59:45
Female [04:30:00,04:45:00) 14:12:55.0 14:07:52 01:25:42
Female [04:45:00,05:00:00) 15:10:16.0 15:04:45 01:16:58
Female [05:00:00,05:15:00) 14:15:10.0 14:17:35 01:30:35
Female [05:15:00,05:30:00) 15:03:40.0 15:16:57 00:57:03
Female [05:30:00,05:45:00) 15:49:06.0 15:49:06 00:10:44
Female [05:45:00,06:00:00) 16:11:27.0 16:04:49 00:18:30
Female [06:00:00,06:15:00) 16:45:36.0 16:45:36 00:11:01

 

Stand Alone Marathon to Ironman Run Time - Example of Usage

Suppose you are a male athlete who completed a marathon in 3 hours and 20 minutes:

  1. Identify Your Gender:

    • Since you are male, refer to the table for men.
  2. Find Your Marathon Time Bin:

    • Look for the bin that includes your marathon finish time. In this case, the time bin is "[03:15:00, 03:30:00)".
  3. Review Predicted Ironman Run Times:

    • Median Ironman Run Time: This column shows the median Ironman run time for athletes whose marathon times fall within the specified bin. For the bin "[03:15:00, 03:30:00)", the median Ironman run time is 04:14:17.0.
    • Average (Avg) Ironman Run Time: This column shows the average Ironman run time for the same group. For this bin, the average Ironman run time is 04:16:06.
    • Standard Deviation (SD) Ironman Run Time: This column indicates the variability of the Ironman run times. For this bin, the standard deviation is 00:30:41.

This means that, based on historical data, a male athlete with a marathon finish time between 3 hours and 15 minutes and 3 hours and 30 minutes typically finishes the Ironman run portion in around 4 hours and 16 minutes, with a standard deviation of about 30 minutes and 41 seconds.

Table to predict Ironman Run Time from Open Marathon Time for Men

Gender Open Marathon Bins Median Ironman Run Time Avg Ironman Run Time Standard Deviation IM Run Time
Male [02:30:00,02:45:00) 03:09:59.0 03:14:01 00:17:33
Male [02:45:00,03:00:00) 03:40:27.0 03:43:04 00:26:01
Male [03:00:00,03:15:00) 04:02:12.5 04:07:43 00:30:16
Male [03:15:00,03:30:00) 04:14:17.0 04:16:06 00:30:41
Male [03:30:00,03:45:00) 04:27:52.0 04:32:52 00:34:06
Male [03:45:00,04:00:00) 04:38:58.0 04:47:15 00:39:44
Male [04:00:00,04:15:00) 05:04:43.0 05:04:41 00:34:18
Male [04:15:00,04:30:00) 05:10:04.0 05:09:30 00:39:43
Male [04:30:00,04:45:00) 05:18:06.5 05:16:07 00:31:01
Male [04:45:00,05:00:00) 05:33:09.0 05:32:00 00:32:41
Male [05:00:00,05:15:00) 05:54:44.0 05:50:33 00:32:03
Male [05:15:00,05:30:00) 05:49:26.0 05:43:03 00:28:53
Male [05:30:00,05:45:00) 05:53:30.0 05:51:35 00:26:54
Male [05:45:00,06:00:00) 05:32:59.0 05:45:56 00:33:52
Male [06:00:00,06:15:00) 05:47:41.0 05:47:43 00:17:40
Male [06:15:00,06:30:00) 06:23:09.5 06:23:09 00:12:40

Table to predict Ironman Run Time from Open Marathon Time for Women

Gender Open Marathon Bins Median Ironman Run Time Avg Ironman Time Standard Deviation IM Run Time
Female [03:00:00,03:15:00) 03:34:50.5 03:57:04 01:00:35
Female [03:15:00,03:30:00) 04:00:37.0 04:04:24 00:18:13
Female [03:30:00,03:45:00) 04:30:33.5 04:31:17 00:31:26
Female [03:45:00,04:00:00) 04:37:01.5 04:41:47 00:36:03
Female [04:00:00,04:15:00) 04:49:08.0 04:48:13 00:23:19
Female [04:15:00,04:30:00) 04:57:31.0 05:05:48 00:24:51
Female [04:30:00,04:45:00) 05:01:35.0 05:09:35 00:33:19
Female [04:45:00,05:00:00) 05:23:36.0 05:27:10 00:35:07
Female [05:00:00,05:15:00) 05:01:22.0 05:18:47 00:43:56
Female [05:15:00,05:30:00) 05:57:13.5 05:52:22 00:26:38
Female [05:30:00,05:45:00) 05:49:37.0 05:49:37 00:11:10
Female [05:45:00,06:00:00) 05:57:34.0 05:57:19 00:14:05
Female [06:00:00,06:15:00) 06:24:24.0 06:24:24 00:09:55

In the next section I present graphical representations which is similar to the above tables and can also be used to quickly figure how what your most likely performance will be. 

How to Read the Charts Below

The first chart presents a graphical analysis comparing Ironman run times for men and women based on their open marathon times. The chart uses a smoothing model to highlight trends and nuances in the data, providing a visual representation of the relationship between marathon times and Ironman run times. Here’s how to interpret the chart:

Key Elements of the Chart

  1. Axes:

    • X-Axis (Horizontal): Represents the open marathon finish times.
    • Y-Axis (Vertical): Represents the Ironman run finish times.
  2. Data Points:

    • Each black dot represents an individual athlete’s marathon and corresponding Ironman run times.
  3. Smoothing Line (Blue Line):

    • The blue line represents the smoothed average relationship between marathon times and Ironman run times for both men (right) and women (left). It shows the general trend in the data.
  4. Confidence Intervals (Gray Shaded Area):

    • The shaded area around the blue line represents the confidence interval. This interval provides a range within which we can expect the true average relationship to fall, with a certain level of confidence (usually 95%).
  5. Dashed Lines:

    • The red and blue dashed lines represent confidence thresholds of 90%. This means there is a 90% chance your result will finish within these range, based on the data analyzed.

Using the Chart for Prediction

To predict your Ironman run time based on your open marathon time, follow these steps:

  1. Identify Your Marathon Time:

    • Suppose your marathon time is 3 hours and 20 minutes (03:20:00). Locate this point on the x-axis.
  2. Find the Corresponding Point on the Smoothing Line:

    • Move vertically from the 03:20:00 mark on the x-axis to intersect with the blue smoothing line.
  3. Read the Predicted Ironman Run Time:

    • From the point where your marathon time intersects the blue line, move horizontally to the y-axis to find the predicted Ironman run time. For example, for a 03:20:00 marathon, the predicted Ironman run time for men might be around 4 hours and 14 minutes, and for women, it might be slightly different as indicated by the left chart.
  4. Consider the Confidence Interval:

    • The gray shaded area around the blue line provides a range of likely outcomes. If your marathon time is 03:20:00, the vertical distance covered by the gray area at this point gives you the range of potential Ironman run times. This means that while the average predicted time might be around 4 hours and 14 minutes, the actual time could reasonably fall within the upper and lower bounds of the shaded area.

Example

If you are a male athlete with a marathon time of 3 hours and 20 minutes:

  • Locate 03:20:00 on the x-axis of the right chart (for men).
  • Move up to the point where it intersects the blue smoothing line.
  • From this intersection, move horizontally to the y-axis to find the predicted Ironman run time, which might be around 04:14:00.
  • The gray shaded area indicates the confidence interval. If it spans from approximately 03:50:00 to 04:40:00, this means your Ironman run time is likely to fall within this range, considering the variability and confidence in the model.

Smoothing Models to Predict IM Run by Gender 

 

Here is the same chart but for Ironman Total Time on the Y-axis instead. 

Smoothing Models to Predict IM Overall by Gender 

 

Graphical Analysis Using a Linear Model to Highlight Prediction Differences by Continent

Below is a similar analysis that breaks down the differences by both gender and continent. I focused on Europe and North America due to sample size considerations. For more details on the differences between European and North American athletes, see my other blog on Estimating Your Ironman Finish Time.

 

 

Using a Linear Model Summary for Predicting Ironman Run Time

Male: Marathon Time vs Ironman Run Time

Dependent Variable: How long it takes to finish the Ironman run

  • Time_Seconds:

    • Coefficient: 0.820
    • Standard Error: 0.030
    • Interpretation: For each second increase in marathon time, the Ironman run time increases by approximately 0.820 seconds. This is a very strong predictor, meaning marathon time is closely related to Ironman run time.
  • Constant:

    • Value: 5,482.691 seconds (approximately 1 hour and 31 minutes)
    • Standard Error: 416.735
    • Interpretation: This is the starting point for the Ironman run time when marathon time is zero, though in real life, marathon time can never be zero. It’s just a necessary part of the math model.
  • Observations: 711

    • Interpretation: The number of athletes' data points used in the model.
  • R² (R-squared): 0.518

    • Interpretation: About 51.8% of the changes in Ironman run times can be explained by marathon times.

 

Female: Marathon Time vs Ironman Run Time

Dependent Variable: How long it takes to finish the Ironman run

  • Time_Seconds:

    • Coefficient: 0.796
    • Standard Error: 0.057
    • Interpretation: For each second increase in marathon time, the Ironman run time increases by approximately 0.796 seconds. This is a strong predictor, meaning marathon time is closely related to Ironman run time.
  • Constant:

    • Value: 5,563.677 seconds (approximately 1 hour and 32 minutes)
    • Standard Error: 864.694
    • Interpretation: This is the starting point for the Ironman run time when marathon time is zero, though in real life, marathon time can never be zero. It’s just a necessary part of the math model.
  • Observations: 201

  • R² (R-squared): 0.493

    • Interpretation: About 49.3% of the changes in Ironman run times can be explained by marathon times.

 

 

 

Both models show that marathon time is a strong predictor of Ironman run time for both men and women. The R² values indicate that the models explain a substantial portion of the variance in Ironman run times, with slightly higher explanatory power for men (51.8%) compared to women (49.3%). The constant terms, while not practically meaningful, are necessary for the linear model calculations. The number of observations indicates the sample size used to build each model, ensuring robust results. 

 

Explanation of Linear Model Summary for Predicting Ironman Run Time (No Intercept)

Male: Marathon Time vs Ironman Run Time (No Intercept)

Dependent Variable: How long it takes to finish the Ironman run

  • Time_Seconds:

    • Coefficient: 1.204
    • Standard Error: 0.006
    • Interpretation: For each second increase in marathon time, the Ironman run time increases by approximately 1.204 seconds. This is a very strong predictor, indicating a direct multiplier relationship between marathon time and Ironman run time.
  • Observations: 711

Linear Models to Predict Ironman Run (No intercept)

Female: Marathon Time vs Ironman Run Time (No Intercept)

Dependent Variable: How long it takes to finish the Ironman run

  • Time_Seconds:

    • Coefficient: 1.159
    • Standard Error: 0.009
    • Interpretation: For each second increase in marathon time, the Ironman run time increases by approximately 1.159 seconds. This is a strong predictor, indicating a direct multiplier relationship between marathon time and Ironman run time.
  • Observations: 201

 

Comparison and Contrast: Intercept vs. No Intercept Models

Intercept Model

  • Pros:

    • Provides a baseline Ironman run time when marathon time is zero, although this is not practically meaningful.
    • Can account for additional factors that might contribute to Ironman run time that are not directly proportional to marathon time.
  • Cons:

    • The constant term can be difficult to interpret in a real-world context.
    • Slightly more complex interpretation since it includes an intercept.

No Intercept Model

  • Pros:

    • Simpler interpretation: the coefficient is a direct multiplier, meaning it directly relates the marathon time to the Ironman run time.
  • Cons:

    • Assumes that the relationship between marathon time and Ironman run time is purely proportional and passes through the origin, which may not capture all nuances.
    • May oversimplify the relationship by ignoring any baseline factors not captured by marathon time.

Both models show that marathon time is a strong predictor of Ironman run time for both men and women. The intercept model provides a slightly more complex but potentially richer understanding by including a baseline Ironman run time. The no intercept model simplifies the relationship, making it easier to interpret and apply, especially for practical use. The choice between the two depends on the context and the level of detail needed in the prediction.

 

IM Run by Age Group

Using the Charts Below

To predict your Ironman run time based on your marathon time:

  1. Identify Your Gender and Age Group:

    • Locate the appropriate panel based on your gender and age group.
  2. Find Your Marathon Time:

    • Locate your marathon finish time on the x-axis of the relevant panel.
  3. Determine the Predicted Ironman Run Time:

    • Move vertically from your marathon time to intersect with the blue linear model line.
    • From this intersection, move horizontally to the y-axis to find the predicted Ironman run time.
    • The gray shaded area gives a range of likely outcomes, providing an estimate of variability.

Example

If you are a male athlete, 30 years old, with a marathon time of 3 hours and 20 minutes:

  • Locate the panel for "AgeGroup 30" in the bottom row.
  • Find 03:20:00 on the x-axis.
  • Move up to intersect the blue line, then move horizontally to the y-axis to read the predicted Ironman run time.

 

IM Overall Time by Age Group

Here is the same chart but for Ironman Total Time on the Y-axis instead. 

 

 

Summary

This blogspot explores the exciting link between marathon performance and Ironman triathlon results, which will be very informative for an athlete where the change from marathon to triathlon is intended. Using this vast real-world data and very meticulous and sophisticated analysis, this post suggests a detailed framework able to predict Ironman run times and overall Ironman times based on marathon performances.

Key findings showed that for every added second in their marathon, men had about a 1.204-second increase in their Ironman run time while women had about 1.159. To be a valid strong predictor, a direct multiplier relationship between the marathon and Ironman run times would place marathon performance as the most critical predictor of Ironman success.

Appendix 

Step 1: Preparing Ironman Data

The first step is to prepare the Ironman data:

  • Selecting Relevant Columns: Only the essential columns (Name, Run, Total_Time, Year, AgeGroup, Continent) are kept.
  • Cleaning Data: The Gender-AgeGroup column is converted to a numeric Age column, and all names are converted to lowercase to ensure consistency.
  • Filtering Duplicates: For each athlete (identified by Name and Year), only the best run time (minimum Run) is kept to avoid duplicates.
  • Removing Duplicates: The dataset is further cleaned to remove any remaining duplicate records.

The resulting dataset is a cleaned version of the original Ironman data, with one record per athlete per year.

Step 2: Preparing Running Data

In the next part, I process the run-only marathon dataset that contains additional running data from a collection of marathons:

  • Selecting and Filtering: Relevant columns are selected, and rows are filtered based on specific keywords in the filename, such as the race location.
  • Data Conversion and Cleaning:
    • Time Conversion: The run times are converted to numeric seconds.
    • Name Consistency: Names are converted to lowercase.
    • Age Calculation: The Age column is extracted and cleaned, and the athlete’s birth year is calculated.
    • Categorizing Age Groups: The data is categorized into high and low age groups based on how Ironman and marathons calculate Age Group. For example, a 34-year-old doing a race on March 30 would be considered 35-39 by Ironman if their birthday is July 2. Capturing the edges of the age groups in this way allows for better comparison.

Step 3: Combining Data

  • Creating Lagged Data: A version of the running data with the year lagged/lead by one year is created to allow only comparisons when an Ironman and marathon have been finished within ±1 year of one another. This helps ensure fitness carries over and improves the matching ability (athletes who are likely still fit and active).
  • Joining Data: The running data and Ironman data are joined on multiple criteria (Name, Age Group, Year, Continent) to find matching records.
  • Filtering and Cleaning:
    • Best Performance: For each athlete, the best performance (minimum time/run ratio) is kept to identify the peak of what’s possible for each athlete (in case of multiple comparisons).
    • Gender Identification: Gender is determined from the AgeGroup column.
    • Applying Filters: Additional filters are applied to remove outliers and ensure the data is within reasonable ranges for analysis.