I've been using the PMC for some time now to track my training and to try to get the balance right between doing too much and not enough but in a more structured way. Recently I have paid more attention to trying to use the PMC to predict when I thought I might go well and to try to arrange things so that I would have the right balance of freshness and fitness on a particular day without compromising too much my CTL for events that are not high priority events for me.
My first efforts were rather scuppered, in the Cheshire Cat sportive as I had a viral illness and an URTI so any form I may have had was trashed anyway so it was not possible to see whether how the PMC "thought" I would feel actually matched up with my real life experiences.
Anyway, I've continued with this method of trying to manage my training load and have had some very helpful advice as I have gone along. The first piece of advice concerned the setting of the CTL/ATL constants in the PMC which have default values of 42 and 7 respectively. It has been suggested to me that if an older rider such as myself used an increased ATL constant of say 10 the PMC might better match my expected training status on a particular day, this has been suggested by both Andrew Coggan and Steve Palladino, both of whom are in my age range, sorry guys!
Accordingly I cloned my PMC and have set various ATL constants ranging from 7 to 12 and I have been watching the TSB values and gauging how I feel in comparison with the TSB figures generated by the charts. I have generally felt during this trial period that an ATL constant of about 10 "feels" about right.
Based on this I continued my investigations and found on the Internet a presentation given by Andrew Coggan specifically about the PMC and how it operates. In that presentation many figures were presented, I was particularly interested in one that seems to illustrate that in efforts lasting longer than 10 minutes individuals have tended to produce their power personal bests with a TSB score in the +5 to +15 range. I have enclosed a graph from Andrew Coggan's presentation below which illustrates this, I hope he won't mind, didn't think it would be a problem the presentation having been published on the Internet for general consumption.
I decided therefore to plan my training for my next event trying to hit a TSB of +10 to +15 in the hope that this would coincide with feeling Ok on the day and I reasoned that if this were to be the case that this would form a good basis for structuring future preparations. It could be fine tuned if it didn't work out very well by targeting a higher or lower TSB on future event days depending on how things worked out, at least I would have a point of reference to work from.
Following advice from Steve Palladino I entered into my WKO+ calendar a series of "dummy" training sessions during the 2 week period leading up to my next event and put in TSS values for those sessions which are largely known and based on previous data. Doing this I could see what my TSB was going to be on a date in the future. As I did the sessions I deleted the dummy data and replaced it with real workout data and could fine tune things as time went on to try to land gently on the required TSB on the day in question, in this case 01.06.2008, Sunday.
At the moment my PMC is as shown below, the right hand column represents my predicted TSB on 01.06.2008 which should be about right (currently predicting +13.8) and will at least give me something to work from. The event itself should deliver a TSS of approximately 400 which will still leave me after the event with a CTL of 100, hardly down at all from where I was so I'll not have "spent" too much CTL on the event which is not a high priority event for me, more an event to try out this approach!
Ok, that's how I'm trying to approach things at the moment, thought it might be of interest to someone out there, I'll let you know how it works out! It may sound like a load of work but it's not at all really, all you do is upload your workout data and look at the charts with the different ATL constants set.