We were often asked how Vmaxpro builds up a velocity profile, how the 1 rep maximum (1RM) is deter­mined and how exactly the whole thing actu­ally comes about at the end. To be honest, even me as the devel­oper cannot predict or explain the exact paths to the result. Nowa­days, many algo­rithms work in parallel, e.g. to deter­mine the 1RM, the results of which are calcu­lated on the basis of different confi­dence inter­vals. All algo­rithms are also contin­u­ously opti­mized on the foun­da­tion of millions of executed repe­ti­tions with Vmaxpro.

However, one ques­tion can be defin­i­tively answered — the results can be extremely accu­rate if a few boundary condi­tions are met during usage. We are constantly working on making our algo­rithms as stable as possible. But even the most stable calcu­la­tion cannot give good results with bad input values. With this short article I would like to shed some light on the three ques­tions and demys­tify the magic behind Vmaxpro’s algo­rithms.

The velocity profile

The basis for velocity-based strength training is the quasi-linear rela­tion­ship between the concen­tric execu­tion velocity and the load used. In simple terms — the higher the load, the slower it can be moved. If the load is maximal, it is moved at the minimum velocity threshold (MVT). If the load would be increased further, the move­ment would be slower and the athlete would no longer be able to master the concen­tric phase. The MVT is the velocity at maximum load. It is impor­tant that this rela­tion­ship is different for each athlete and each exer­cise. The profile must there­fore be built up for each athlete in each exer­cise. Luckily Vmaxpro does this completely auto­mat­i­cally in the back­ground on the base of all data.

The 1 rep maximum

As already mentioned, the under­lying rela­tion­ship is linear. At least it is in the area of interest we have. With this knowl­edge, the slope of the straight line can be deter­mined by executing submax­imal loads at maximal velocity. It is recom­mended to use a load of more than 55 % 1RM, because the maximal velocity with light loads is diffi­cult to achieve for inex­pe­ri­enced athletes and this is simply not possible in some exer­cises. Although two loads would be math­e­mat­i­cally suffi­cient, several loads lead to a more reli­able result. Vmaxpro uses the best repe­ti­tions of all repe­ti­tions of the last 6 weeks with loads above 60% of the maximum load used. Why 6 weeks? It takes some time to change the velocity profile through training. One speaks here of about 42 days adap­ta­tion time. Within 42 days the rela­tion­ship is quite stable.

Vmaxpro takes a close look at all repe­ti­tions in order to elim­i­nate possible incor­rect measure­ments or record­ings of partial move­ments. For example, in squats, repe­ti­tions with missing depth are not consid­ered in compar­ison to all other repe­ti­tions because they are not repre­sen­ta­tive.

In addi­tion to the slope of the straight line, the MVT for the accu­rate deter­mi­na­tion of the 1RM is now missing. The MVT can be deter­mined by pushing an athlete to his 1RM or by moving a submax­imal load to tech­nical failure. Here, the higher the load, the more reli­able the result. Anything over 80 % usually leads to a very good result. The liter­a­ture recom­mends caution when doing dead­lifts. Besides very high corre­la­tions in all other exer­cises, the approach of the submax­imal failure theorem does not always seem to be correct for dead­lifting.

Inter­nally Vmaxpro auto­mat­i­cally uses the slowest, valid, last repe­ti­tion of a set with a load above 70 % 1RM within the last 6 months. The MVT is adjusted in each training if neces­sary. The MVT is very stable over time and can there­fore be consid­ered over 6 months.

Besides the pure velocity based approach Vmaxpro also calcu­lates the 1RM over submax­imal sets until failure. Depending on the exer­cise and the scope, various algo­rithms are used. With the execu­tion of a single set to failure, a reli­able profile is created.

As an example: 10 repe­ti­tions of pull-ups with a body weight of 75 kg and a maximal execu­tion velocity of 0.7 m/s, in one of the first repe­ti­tions, results in a 1RM of about 95 kg. With the slowest velocity in this set of 0.35 m/s, two loads with two veloc­i­ties and an MVT of 0.35 m/s result.

The accu­racy

In order to clarify how accu­rate the whole thing can actu­ally be, the meaning of the calcu­lated 1RM must first be discussed. The calcu­lated 1RM is always a theo­ret­ical 1RM under optimal condi­tions. Espe­cially with less advanced athletes, with loads close to 1RM, tech­nique or local fail­ures can occur. Such factors cannot be calcu­lated, but they are signif­i­cantly reduced with increased training expe­ri­ence.

When an athlete’s profile is set, the 1RM can be reli­ably predicted in the range of ±2.5%. However, it is partic­u­larly impor­tant that this error is not random but system­atic. This means that the 1RM, for example, is always under­es­ti­mated by 1%. This does not nega­tively influ­ence the use in daily training to deter­mine the training load used.