Update README.md

Addressed Harriet's merge #17 comments

training/README.md

@@ -1,10 +1,12 @@
+[RAP Champions]:https://github.com/NHSDigital/RAP_CoP_dev
+
 # Training your team
 This is just a stub page, but soon we intend to populate this with guidance on how to approach training your team in RAP.
 
 Some initial pointers to get you started:
-- **get your team learning the basics of Python and Git**: They're both used a lot in RAP, so the sooner they can get even a slightly familiarity with these, the easier they will find it to engage with any RAP training
-- **make time for RAP**: people need time to learn and use what they've learned - if they don't get defended time they won't develop these skills, and will potentially fall behind and become demotivated. 
-- **It's ok to start simple:** there is nothing wrong with starting very basic: wrap your existing SQL code in a bit of Python to orchestrate it, get that working, include some very simple tests to prove it works and then add it to Git and publish it. This is a great way to start and provides a solid foundation on which to build and learn more.
-- **copy other people:** Look at your colleagues code, copy chunks of it, play around with it, try and break it - encourage others do the same. Re-use is at the heart of RAP and it's also a great way to teach and learn.
-- **Talk to the RAP champions:** The RAP champions have probably trained a number of people - ask them for advice.
+- **Get your team learning the basics of an open source language and Git**: confidence using version control, and a programming language of choice (at NHS Digital, this is usually Python), will make it easier for your team to engage with RAP training.
+- **Make time for RAP**: people need time to learn and use what they've learned - if they don't get time set aside specfically for RAP they won't develop these skills, and will potentially fall behind and lose motivation. 
+- **Pragmatism is ok!:** You don't leap up to the level of full pyspark pipelines in one go. There is nothing wrong with starting very basic: wraping your existing SQL code in a bit of Python to orchestrate it, get that working, include some very simple tests to prove it works and then add it to Git and publish it. This is a great way to start and provides a solid foundation on which to build and learn more.
+- **Copy other people:** Look at your colleagues code, copy chunks of it, play around with it, try and break it - encourage others do the same. Re-use is at the heart of RAP and it's also a great way to teach and learn.
+- **Talk to the RAP champions:** The RAP champions (both your RAP trained colleagues and the Data Science RAP squad) have probably trained a number of people - ask them for advice. You can find out how to get in touch [here][RAP Champions].
 - **The thin-slice approach is built with training in mind:** [See our pages on the the "thin slice" approach of doing RAP](/rollout-approach/thin-slice-strategy.md) - by doing a simple part of the publication from end to end, you create something people can learn from, and then with help expand. This can provide the structure around which you can train people.