Strange CSS Results And Gaming The System...

At Swim Smooth, one of the things we're most pleased about is how many people are using CSS to improve the training side of their swimming preparation*. We've heard from so many of you about how you've taken 5 or 10 seconds per 100m off your swim times after just a couple of months of CSS training - which is fantastic!

For more about CSS and see the calculator here:

One of the most common questions we receive about the CSS calculation is that if you keep your 400m pace the same but get faster over 200m then your CSS pace actually gets slower! This seems quite counter-intuitive that you are getting faster and yet your CSS pace is getting slower - in fact we get a lot of emails from swimmers (and coaches) telling us the calculator is broken and we need to fix it in a hurry!

The calculator is actually right though and is just telling us something interesting about fitness which is very relevant to how you should train. The calculation looks at the rate of drop-off between your 200m and your 400m times to predict what you would do over a longer distance closer to CSS pace (e.g. 1500m). So someone who is very quick over 200m but dies a little over 400m will have a slower CSS than someone who is much more evenly paced, even if the first swimmers times are quicker combined.

Mind bending isn't it? What the test is assessing is whether your current fitness is more attuned to sprinting or distance swimming. If you have a large drop off in pace from the 200m to the 400m then this suggests you're more sprint-based and you will be significantly slower over 1500m. If you have little drop off between the 200m and 400m then you're well trained for distance swimming and you will be able to swim 1500m at only a slightly slower pace than 400m.

If you think about it, this has to be the case otherwise the best sprinters in the world would also be the best distance swimmers in the world and vice-versa. You can see this explained more visually here:

Of course if you're a triathlete or open water swimmer then you are aiming to be fast over longer distances - developing your diesel engine. If your CSS is getting slower it could be you have been inconsistent with your training or you could be doing too much sprint training (very fast with long recoveries) and not enough CSS type training (not quite as fast but with short recoveries).

Example Athlete: Michael Japp

Michael has improved both his 400m and his 200m times since May, which is great news. However, his CSS has slowed by 1 second per 100m - how so? Michael improved his 400m by 5 seconds but his 200m by 7 seconds. As the 200m is much shorter, this means his speed improved relatively much more over 200m than 400m which is symptomatic of biasing training towards very fast, short intervals with lots of rest and recovery rather than longer intervals at around CSS pace.

In Michael's case being of a very strong athletic background we felt that his anaerobic system has come back to life a little quicker than his aerobic endurance and recommended more CSS training and even some "tough love" Red Mist endurance sets:

Gaming CSS

There's even been some discussion recently about how you can "game the test" in order to get a better CSS result. You might do this by swimming deliberately slowly over the 200m which would give you a better CSS pace but this would be totally pointless as the whole idea of the test is to get a real picture of your current fitness so that you can train accurately going forwards.

Gaming the test will only see you having to target unrealistic CSS target times in training which will ruin the quality of your training - so please don't do it! Both the 200m and 400m time trials have to be maximum effort but they still need to be well-paced to yield your best times. This in itself can be confusing - how can an all-out effort be well paced? See our classic Gradual Crescendo post here:

Swim Smooth!

* Swim Smooth didn't invent the CSS calculation but we are huge fans and advocates of using it to help swimmers train better. To improve the accuracy of this training try doing it in conjunction with a Tempo Trainer Pro to set your training paces.


Anonymous said...

then how do you explain this:
200 2:57
400 6:22
Calculated CSS 1:43
Actual 1500m time 24:28, average pace/100 1:38.
So, pace for 1500m 5 seconds faster than predicted CSS. What's going on here?

Unknown said...

Big fan of CSS and love the red mist sessions and have seen real improvements with it. 1 question altough the calculator has a 25m and 50m pool option do you think CSS should be different if your doing it in 25m or 50m, as i do both and find CSS pace in 50m harder work. Mani

Paul said...

Hi Anonymous

OK, couple of questions:

1) was the 1500m conducted in the same pool or is this an open water wet-suited result perhaps?

2) how far apart are these two tests - could you have improved / regressed depending upon training volume / intensity in the interim?

3) That 400 swim looks a bit slow given that it's only 2.5s/100m slower than your 1500m pace - did you just have an "off" day on the 400? The 200m looks much better matched to your 1500m time.

4) Were you swimming all tests by yourself with zero chance of any drafting affecting the result?

The CSS test gives a slow CSS time because of the large differential between the 400 and 200 (as explained in the article) - by answering the above we might be able to deduce why.



Paul said...

Hi Mani, great to hear from you and that you're making great progress!

Yes, this is a very good point!

Ideally if you train between two pools (25m and 50m) you should really do a CSS test in each as everyone will be different with respect to the efficiency of their turns etc. Of course, this can be a bit of a faff though!

As a (very) rough rule of thumb - I'd suggest 1.5-2.5s/100m slower in the 50m pool than the 25m pool.

Hope this helps.


Kevin Morice said...

Anonymous, a few possible answers.
1. Your 1500m was open water or wasn't an accurate distance? (I am in this box with my open water time for a supposed 1900m open water in rough conditions being under my 1500m pool time).

2. You are a super racer who thrives on competition and adrenaline and can put out 5% more on race day? (I know more than a few of those).

3. You aren't working hard enough on your 400m test? If you can do ~1:28/100m for 200m but only put in ~1:37/100m for a 400m that seems like a huge drop even if you are a pure sprinter. And yet you able to hold 1:38 pace for a whole 1500m, effectively doing just under 4x400m reps with 4secs recovery each?

Paul said...

Kevin, great point on #2!

Unknown said...

Great article, thanks. I spent 2013 gaming the test, took 10 seconds of my CSS and was left wondering why I couldn't keep up when I got promoted to a faster lane.

After that, I've spent 2014 ignoring the test results and applying what you describe in your (excellent) 'taking control' article for H2Open which I've found to be very helpful (

...I now have renewed faith in the 400/200 test and will apply the results to my training.


huwf said...

In the past I've not had enough recovery between the two and been very surprised with the CSS score from having such a slow 200m!

Think I got it better this week though with 5:54 and 2:53 - both 2s outside my PBs. Or maybe the hard work of the last couple of months is paying off and I can recover from a TT better than before. So how much is the minimum recovery should you put between the two to get an accurate idea of the times?

David Kaufman said...

That's a great article. Gaming the system has been on my mind since the first time I took the css test (a few months ago) and played around with the numbers. Great explanation!

Paul said...

Great to hear you're on a roll Nick - are you swimming the Channel yourself?

Huwf - ideally take a good 8-10 mins active recovery between the 400 and 200.

David - naughty, naughty! ;-) Glad to see you're thinking it through though!


John said...

After a couple month layoff would you have swimmers test for CSS in the first week back to get a current snapshot, or have them get some meters in for a few weeks prior to testing?

Paul said...

Good question John. I would approach it one of two ways:

Either 1) do as you say, get six sessions under your belt then try or 2) add about 1s/100m for every week you've had off to your previous CSS as a rough rule of thumb.

Hope that helps.


Anonymous said...

Very relevant post for me, I've been focusing on CSS training for the past few months and found my CSS increase each time I did a time trial. Along with this, my sets (red mist, Mega locks, fresh and fruity) have been too easy, so instead of time trials, I've been dropping one second off my CSS each week until the sets have become challenging, but achievable (my last time trial came out at CSS 1:41/100m, but I've found 1:38 to be about right at the moment)
Is this approach ok?, my background has been in open water distance and my sprints are pretty slow (1:32/100m).

Anonymous said...

Are there any correlations between the Stroke Rate Ramp test and the CSS test? It seems to me that one swimming speed is determined from the stroke rate test, while possibly another speed from the CSS test. Would you use a SR determined by the ramp test for the CSS test?

Paul said...

Hi Anonymous # 1 - by "CSS increase" I assume you mean you have been getting faster each time?

If so, yes, 1s per week is a good thing to aim for and/or you could try 1s per fortnight whereby you have two attempts at the same key sessions at the same speed before notching it up again on week 3 etc.

Anonymous # 2 - it's certainly a good idea to check your self-selected SR during the 2nd 100m of the 400m CSS test to see how it matches up against your SR ramp test results. Let us know of you findings!


Unknown said...

It took me a while to work out how the CSS worked, but I just trusted it anyway! For me it correlates very well. Recently did 6:50 for 400m and 3:15 for 200m; CSS = 1:48. Just did a 36:24 for 2000m. A little bit slower (1:49/100), but probably due to extra 500m compared to a 1500m.
One thing I was wondering Paul is what is the optimal time between intervals when training at CSS pace. I've found that going from 10 sec rest to 15 second rest dramatically improved my times for 15x100m intervals. It almost felt too easy, so I went back to 10secs. Also, I was wondering what you thought of the strategy of, say when you lose a bit of fitness after a layoff, coming back at the same CSS as before, lengthening the rest period (to say 15 secs) and then progressively decreasing the rest period back to 10 secs over a few weeks? I've found this works for me, because if I increase the CSS time instead, I tend to lose my rhythm as well and tend to struggle a bit.

Paul said...

Hi Stuart

Really it sounds like you've already found your own ideal way of doing things and as they're working and would simply say - yes to the 10s and yes to the 15s post return.

Well done! Keeping it simple is key.


Anonymous said...

Hi Paul

How often do you recommend doing the CSS testing? And what's the best way to improve CSS times? Is red mist the answer? :-)


Paul said...

Hi Sarah

About every 6-8 weeks is a good time frame.

Lots of info on how to develop your CSS by typing "CSS" into the search bar at the top left of

Hope this helps.


Subscribe to Feel For The Water
And receive the amazing Mr Smooth animation as your optional free gift.
Find out more: here

* required
I consent to receiving tips to improve my swimming and occasional information about our products and services from Swim Smooth. You can unsubscribe at any time. See our Privacy Policy
Powered by Blogger.


Blog Archive

Recent Posts