I want to do some more snatching to improve my various tests (eg “Snatch 5min” and “Snatch 200”). I find the programming for this a bit tedious though – I tend to end up doing the same thing all the time. Enter a pair of dice to determine what snatch workout to do. But how exactly to do it?
Whilst the problem is not quite difficult, it is worth thinking about a good parametrisation, ie one where all the relevant workout parameter are modified, within reasonable ranges, and in good combinations. And one might also want to think about the shape of the distribution, eg whether one might want to do one tough workout from time to time, and balance this with a number of significant lighter one’s, or whether all workouts should be a bit more “average”.
Before we go into specifics, I quickly want to suggest some ways of converting the throw of one die into a meaningful number. Three suggestions:
- Start with 100, and take off  at every step
- Start with 100, and multiply with [0.66] at every step
- 100 / Number, also 100 / Number + 1
ie one where the increase / decrease is a constant number (#1), a constant percentage (#2) or another formula depending on the Number thrown (#3). Obviously the above options are quite different in that the “average” number obtained is quite different, so we correct for that in the following table
|Reduc 33%||1/x||1/x+1||Reduc 10|
and in the following chart
There is no point overanalysing it, but I like the 1/x and 1/x+1 best: they have a bit of a curvature (1/x more so than 1/x+1) which means that from time to time there will be a very challenging workout, and the remainder of the time it will be rather tame, and they are easy to compute.
For most HIIT style workouts, the following three coefficients provide in my view the best parametrisation
- The overall number of reps
- The number of reps per set
- The Recovery : Load Ratio (“RLR“)
Note that I would not use the number of sets as a direct parameter, simply because in this case some very large numbers of reps can happen (normally when you do very long sets you want to do fewer of them, but that parametrisation does not allow for this).
For the number of reps I would use a slightly less skewed distribution – I want to make sure that I do enough, but that I don’t overdo it. 1/x+1 fits the bill quite nicely – it can be seen from the table that the overall range between the smallest and larges number is 3.5x. An average of 100 snatches per workout sounds slightly to high, so I’d go for 80. Keeping in mind that a maxvalue of 188 corresponds to an average of 100, we simply need to reduce this number by 20%, ie 188 x 0.8 = 150. The formula to convert the Number on the die into a number of reps is hence
Total Reps = 150 x 2 / Number+1 = 150, 100, 75, 60, 50, 42
As for the reps per set, a slightly more skewed distribution would do, so I’d use 1/x. Let’s say I want a maximum of 42 reps per set (21 per side; we cant really do more than that, otherwise a set might be longer than the total). The formula here would be
Reps Per Set = 42 / Number = 42, 21, 14, 10, 8, 7
The number of sets is then the ratio of the Total Reps and the Reps Per Set
Now there are evidently two ways of dealing with the fractional number of sets: one can either shorten the last (or first) set, or one can suck it up and round the number of sets up (rounding down is not an option). A bit of a personal decision – I’d probably suck it up for up to 14 reps per set, and do fractionals for 21 and 42.
Finally the RLR. This is a bit more pedestrian. I’d say, reasonable values go in principle from 1:3 to 3:1. However, 3:1 is rather tough, and 3:1 takes maybe a bit too long. So I’d probably go for three choices only
1:2, 1:1, 2:1
ie combining throws of 1&2, 3&4, 5&6