Two ways of comparing a total across bodyweights, and they do not agree, because they were built from different populations for different purposes. Both are here with the difference explained rather than one picked for you.
Want to know where the total actually sits? Your percentile against real results says more than a coefficient does.
Dots is a polynomial fitted to raw lifters. It replaced Wilks for most purposes because Wilks was fitted in the 1990s to a population that no longer looks like the sport, and it disadvantaged the lightest and heaviest lifters. Dots has one curve per sex and takes no view on equipment.
IPF GL Points is what the IPF actually uses to place lifters across classes at their own competitions, so if you compete under the IPF it is the number that decides a Best Lifter award. It has four curves: raw and equipped, male and female. It is expressed against 100 for a world-class performance, so a GL of 80 means roughly four fifths of the way there.
They disagree, and neither is wrong. A coefficient is a model of a population, so changing the population changes the answer. Use GL if you compete in the IPF, Dots if you want the more widely used cross-federation number, and neither if what you actually wanted to know is whether the total is any good, which is a different question with a better answer.
Where the formulas came from, and how they were checked. Both are published coefficients, implemented here from the specification. None of OpenPowerlifting's code was used and none of it could be: their code is AGPLv3 and this site is not. Their DATA is public domain, which made a real check possible: every score here was computed for 3,706,453 rows of their results and compared against the figure they publish for the same lift. Worst disagreement across all of them, 0.005, which is rounding.
A coefficient tells you how a total compares. It does not tell you which lift is holding it back. Training maxes is where that starts.
Work with DQ