Best bike split accuracy


I completed Ironman Wales last weekend. It was my first ironman and I loved it and learnt a lot.

Now looking at my ride details I notice big differences in the best bike split projections and my actual race times.

I rode an average of about 175 and nominal 198 for a 6.23 bike split.

Best bike split projects much faster times for that given power even if I adjust the CDa quite a bit.

Just wonder if others have found similar results, perhaps the nature of the course with a lot of elevation makes these predictions very complicated?

Here are the Garmin results and two different best bike split data predictions :

Just another data point, I used best bike split for a Gran Fondo last weekend and it predicted 5:00:30 at 173 average watts. In the end I did it in 04:54:38 at 174 watts. So my numbers seem much more accurate.

BUT - best bike split said it would be 184w NP and 260 TSS, while it was actually 229 NP and 396 TSS !

I assume it’s to do with me going harder than planned on climbs and coasting on descents when I could have been pedalling.

In your case, maybe try playing with the “climb position”, “max descent” speed and “power limits” sections to see if that brings the result more in line? Or do you know if the predicted wind conditions varied much from what you actually had on the day?

Thanks for the reply, I’ll have a play with the variables as you suggest, it just seems quite a long way off I I felt the power was fairly consistent.

The weather was also very good with minimal wind.

I would think best bike split will only be able to give an accurate estimation if you follow the power recommendations. You rode with a variability index of ~1.14 against a planned 1.04.

Also in the case of a fondo, it’s hard to take into account if you’re drafting throughout with the CdA. If you had stuck to the recommended power plan it would have been 184W and 260TSS but you rode with a variability index of 1.24 which as you said is to do with going harder on climbs and coasting down along with potentially drafting and taking turns.

I used BBS and the garmin head unit field for a half distance tri earlier this year and managed 2:32:47 (213W/225NP) split against a planned 2:30:40 (212W/217NP). I burnt some matches passing traffic and paid for it which you can see in the NP difference but overall a decent prediction.


In my experience riding in Wales, BBS power recommendations don’t match the varying terrain well enough. Often it will tell you to go easy just as you hit a short, sharp climb over a small lump. I think this might be down to the accuracy of terrain model not being sufficient.

1 Like

Another thing to consider is that if the Welsh roads are anything like Scotland’s, an average Crr of .25 is very ambitious, even with the best tyres on. This won’t account for the whole variance, but will probably be a part


Three more anecdotal data points:

IM Hamburg last year predicted 4:51 on 168NP, did 4:54 on 163
Local half predicted 2:28 on 182, did 2:31 on 178
Mont Blanc Man (1800m climbing) predicted 2:51 on 210NP, did 2:58 on 203

So for me it is working fairly well. It took some fiddling with cda, but for my current TT and road bike it is close enough. I’m using the free version and don’t actually pace by BBS, but I’ll use it to calibrate my race plan and expectations.


I’ve used BBS split for a few races now and I’ve found it extremely accurate IF you follow the power recommendations.
The BBS estimate is based on that. You can ignore the power recommendations, end up with a similar NP and your time will be nowhere close. Most likely though your time will be slower than the estimate in this case because you’re not being efficient with your pacing.

Had it within 2 minutes on an Olympic distance, within 2 minutes of an extremely flat 70.3 (basically hold constant power for 2 hours) and it was even pretty close for the mountainous Nice course. This it was within 5 minutes or so.

I find it’s real benefit to be the power plans, not so much the estimation of time. Especially in a Triathlon where I don’t want to go over a target bike NP. For Nice I split the power plan into 15km chunks to align with the hills and descents, which really helped. Certainly helped pace myself since the the course was so variable. 230W NP first 15km, 300w NP for the main climb, 190w NP for one of the downhill portions, etc.


What tools and techniques have you found to be successful in receiving timely notifications while on course of the recommended power intervals?

For Nice, it was pen and duct tape on my handlebars for my targets for each 15km section. I also just kind of memorize the general power target for flats, small hills, large hills, etc, instead of knowing the exact power target for every little hill along the course. I’m not looking at my watch every 5 seconds so I don’t really feel that need to have it that detailed. Basically just check every once and a while to see if I’m roughly close to the target.

When I did Calgary 70.3 it was so flat that there was basically 1 power target, so I just aimed for that.

My watch is set up to autolap every 15km. On the main screen I show 10 second power, lap NP, speed and lap HR.

1 Like

Thanks for the detailed reply… I’m sure there is a lot of truth in that… I did try to stick roughly to the plan in that I set maximal wattage for climbing and minimum for descending… I guess perhaps I should have been more specific but Wales is so variable that it was difficult.

Having adjusted the crr and weather best bike split is now within about 10 minutes and .5mph so not so bad… I’m sure the rest is due to my variability and spikes in power.

Thanks again :+1:

You can export your BBS race plan as a Garmin Power Courses, this gives you on course power instructions. There’s a BBS app that you can add to most newer garmin head units, that picks up the course and has the data fields all set up. If this fails there is also a low tech ‘cheat sheet’ for each plan, that gives power targets for different terrain.

Enjoy Jonny :biking_man::dash:

Hi! Does someone has used BBS on an Edge 130 and/or Edge 130 Plus (the little one).
I have a 70.3 on Sunday and this was the only bike computer that fits my setup.
Many thanks!!!