Weight Training Loading Schemes and Their Applications
1. Ramping up (starting light and going heavier each set, and doing the same number of repetitions per set). The first few sets are essentially warm up sets. The last 2 sets are where the work is done. This method is best if you are doing singles to triples, basically working up to a heavy attempt. Probably great for skill based lifts too (Olympic lifts) because the lighter weights provide some skill practice before a heavy attempt.
2. Sets across (Using the same weight for each set and doing the same number of repetitions per set): This is a bit more intensity than above because all of the sets are reasonably loaded. If done correctly, you are stopping 1-2 reps shy of failure on each set until the last set, which should be really tough. Also, the idea is to load heavy and rest a long time between sets (>3 minutes). This loading scheme is great for strength because it gives you many quality reps with a reasonably heavy load. 5 sets of 5 repetitions is the “gold standard” for this type of loading scheme but anywhere from sets of 3 to sets of 12 have been used. In fact, making the loading “wavy” (varying the reps) seems to be more effective in the long run. Sets across also helps to prevent overtraining and allows you to progress by “auto regulating.” You increase weight when you are able to perform all assigned reps (this is auto regulating).
3. Descending (Start heavy, go lighter each set, but go to failure on some or all of the sets): This is the most intense of the three if done properly. The idea is to go to failure, or nearly to failure on each set. This results in ALL of the sets being done with (relatively) heavy loading. The fatigue effect is most pronounced with this method. This is probably best for hypertrophy because hypertrophy requires some degree of fatigue stimulus. The idea also is to rest less between sets (30 seconds to 1 minute).
So, my take: ascending for heavy singles-triples and skill lifts, sets across for pure strength work, and descending for hypertrophy.
However, there are no hard and fast boundaries here, and there is some overlap between the methods.