Correct way to do Compression test

Carb open? or removed?

How many "pumps" 2-3 and first?

I do 3 puffs,I never bothered with the carbs and not sure if it makes a difference.Having the engine hot does

It helps if you can warm the motor. Round bores and running piston to wall clearances should help too. Then i pull all the plugs and spin the motor till the gauge stops gaining pressure. It may take as many as ten compression cycles since the gauge and line have to be pressurized . Open the throttle if you don't pull the plugs on all the cylinders.

I have a very UNscientific approach:
- Carb wide open
- Crank the engine till the gauge doesn't go any higher

I hear people say to look for how quickly it builds pressure in the first "n" # of "pumps", but I figure the engine doesn't just run for 2 or 3 pumps after I turn the key. Crude as it may be, the max cranking pressure tells me one thing, and I look for something else w/ a leakdown test.

And I know people who say a leakdown test doesn't tell you sh!t. But if 7 cylinders show 4-6% and one shows 20%, you gotta think "something" is going on w/ that 20% hole.

Oh, and I don't do either on a cold engine, although it doesn't have to be anything more than adequately warmed up for me.

I agree with this

This is also what I do

Starter speed is a variable so you have to make sure you have a fully charged battery.

It doesn't hurt to have a low amp battery charger hooked up and charging also

one thing I know .. you can always tell an engine with a good valve job!! when you crank it.. it " Pops " off instantly.......... I have seen an engine with 1 valve barely tweaked / bent , and it kills the vaccume on start up... also , if your cyls are round cold they should be round when hot. how do you bore an engine at the exact temp. it will run at ?? if your pressures are close to the same across, you have what you have. a leak down test shows how efficient your valves- and - rings are at holding their seal.. that is a sure way to see if quality is in your build...

At the risk of show my ignorance / inexperience, what are we looking for with:

A) Compression testing? What should I theoretically expect to see and how do we correlate real world results to the expected value?

B) Leak down – measure of ring / valve seal efficiency only?

By the book, all plugs out, charger on battery, 4 pulses, throttle open. The main thing on the pulses is just do the same on each cylinder. Plugs all out makes it spin faster, and keeps you from missing head gasket blown between two cylinders. Try a compression test with the throttle open and then with it closed, you should see a few more lbs with it open. Also a running compression test can be valuable as well. One cylinder at a time, with the engine idling you will see fairly low compression, blip the throttle and the compression should come up to 80% of the cranking compression, tests for loose valve guides, broken valve springs, worn cam lobes, carbon build up on the intake valves. Also dry and wet compression test, if compression is low, squirt a little oil and retest, if compression goes up it's the rings, if it does not it's the valves or head gasket. I still prefer a leak test, main reasons is battery charge does not affect results, and I can leak test an engine not in the car. When buying a used engine first thing I do is leak it, saved me money several times. After it's built on the engine stand leak it, just to make sure you did it right, again saved me several times.

Compression test measures the ability of the engine to take in air and compress it.
Leak testing just checks the ability of the combustion chamber to hold air. Leak testing is also good to check for blown head gaskets and cracked heads, pressure the cylinder, take off the radiator cap and look for bubbles.
I also like to use a vacuum gauge when testing. Performance engines have low idle vacuum, but will still show 18 to 20 inches if held above the lope of the cam. Low vacuum means something is wrong, if the needle is bouncing while low its in the valve train, just one example.

When we do a compression check on the dyno we do it when swapping parts. A perfect time to run the compression test is when we're doing a carb swap. That way the engine is warm and there is zero restriction in the intake manifold. Pull the plugs and look at them, pull the carb to make a change, run the compression check, then put the new carb on it for more pulls. When you do it that way it doesn't take much time and it provides a good baseline number.

As stated carbs open, plugs out(easier on starter), battery on charger to ensures it does not get weak. Engine warm is essential, particularly in an aluminum block. Pay attention to the first pump number and then total number. Some engine builders like to use the first pump number as a bench mark as it compares to total. As for a total c cranking compression number lots of variable there for sure. Aluminum blocks will generally tend to have less cranking compression that their equally done iron brethren.

As for leak downs a lot of variables there. Good to fund think like slightly bent valves etc. I think if you ask engine builders they would rather see a cranking compression number over a leak down number as to engine health. If you ran a vacuum pump that is another way to gauge the engines overall "health". If that number begins to fall you will likely see that in a compression test as well.