We can try to compute some of the greatest human ST scores through Basic Lift, which is equal to (ST x ST)/5. This is especially tempting because we have access to a lot of olympic weightlifting records .
For this article, we've focused on the Squat , Clean and Jerk and Deadlift techniques. In the first, you are holding the bar over your head and bend and extend your knees. The second is basically about lifting a weight from the floor above your head, using the well known "tilting" movement, while the other is simply pulling a weight off the floor for a few seconds. We can assume that this translates, in GURPS, as a Two Handed Lift (p.B353), for 8 times the Basic Lift for the Squat and the Clean and Jerk, and a Shove and Knock Over (although here we don't knock over the weights!) for the Deadlift, for 12 times the Basic Lift. It's not easy to map real weightlifting techniques to GURPS situations because the game is oversimplified for playability's sake. Therefore, we'll average the three techniques' results, as an attempt to minimize the errors.
Among the best weightlifters we can find Paul Anderson , Tommy Ingalsbe and Don Reinhoudt .
As those results are pretty far from each other, we can conclude that our computations are going to be quite vague. Hopefully, Basic Lift being the square of ST, divided by 5, and rounded, those differences will only imply a lack of precision of 1 point, which doesn't matter much when it comes to upper limits.
Anyway, let's conclude these stats by considering a Basic Lift of 95.15 lbs.
Now, let's take Extra Effort (p.B356) and the Lifting skill (p.B205) into account. For every 10% gain we want, the Will-based Lifting roll is at -1. An olympic champion would have a high Will score in order to break through, giving us some room to take a rather large penalty to the roll. With Will+Lifting at 25, we can apply a -15 and still have an effective level of 10 which is rather safe to beat. This adds 150% to Basic Lift! The rules don't really mention a limit to how much you can improve your abilities through sheer effort, but this sounds an awful lot to me. I'd be tempted to cap this to only 50%. That would give us an original Basic Lift of 63.4 lbs. This means that our olympic champion had ST 18.
As an alternative, consider the default 18-year-old male fit for military service (Kromm's default-attribute baseline). The US Coast Guard (a branch of the US armed forces) allows a bench press test to be used as its strength test for enlisted members. The passing score is 25 repetitions of 80 pounds for men in the 18-25 age range, or an estimated maximum bench press of 125-130 pounds (by the estimation method in this study Comparing this to the bench press record of about 713 pounds (without bench shirt) gives a ratio of 5.6:1, or an effective ST 24. Assuming a 50% weight increase due to Extra Effort (as above) gives an effective ST 19.3, suggesting that a normal distribution with mean 10 and standard deviation 1.5 would effectively capture the range of ST scores.
A calculator for converting from number of standard deviations ("z-score") to probabilities is here For example, by this scale ST 13 (z=2), would be found in 1 of 44 men, whereas ST 19 would be found in 1 of every billion men. Assuming fractional ST is rounded to the nearest integer (i.e., anyone between ST 9.51 and ST 10.5 is considered ST 10), this gives the following distribution:
It is pretty hard to evaluate an olympic weightlifter's ST. Physical feats, in general, are difficult because of the role of Will, fatigue, or even drugs, that are part of the given situation. However, a maximum realistic ST of 18-20 looks rather satisfying. A punch from such a guy inflicts 1d+1 crushing damage, which can knock you down in one blow, without being unrealistically devastating (e.g. breaking your skull).