The way I understand cutoff voltage is this. If you discharge your batteries down to zero with a load on them it's possible to reverse the weaker cells. What this means is if you have a six cell battery and five are great offering long runtime but one runs out sooner, then if you discharge at a high amp rate that one "weaker" cell can loose all power and actually reverse it's polarity. (positive becomes neg. and vice versa)
To stop this from occuring it's generally accepted to discharge your cells down to .90 volts per cell. Take the .90 times the # of cells and that should be your cutoff voltage. From there you can tray your cells with one of the many battery trays out there that work with a much lower discharge rate and also don't allow the battery to go below zero volts so thereby protecting them from reversal.
Keep in mind I've only been in the hobby for "almost" a year and I'm sure others will clear up any inconsistency's in my explanation.
Sounds good. The only thing I would add, quick explanation of matching. Cells are rated by discharging them at a constant rate (20 amps, 30 amps, etc.) and timing the discharge from beginning until end. The cutoff voltage is how you define the 'end'... that point where you quit timing. Like Alpha said, .90 volts per cell is the generally accepted standard for determining the end of the discharge. Cells are then 'matched' into packs by putting cells with similar discharge times together. That's the simple explanation. There are other factors that are used to match packs such as average voltage, internal resistance, etc., but discharge time seems to be the main factor.
The industry feels batteries have a point where useable voltage stops some feel it's .90 others use .85 volts per cell. That would be 5.40 volts for 6cell packs, 3.60 volts for 4cells and 6.30 volts for those using 7cell packs at .90 volts per cell.
So if you wanted to test the runtime, attach a load, use a DVM and time the length for battery voltage to reach a .90 volts level stop the timing when it reaches that point. Take the voltage at peak charge and that would be close to the average voltage of each cell. Divide the peak voltage by the number of cells in the pack and that gives you a ball park of the cells average voltage. You can also try recording the voltage levels at different increments 15, 30 or 1 minutes segments to determine which pack has a little higher discharge curve.
Electrical values are used to determine the voltage a cell produces, along with its resistance.
The battery numbers are all important Runtime=tells you how long battery should last at a give load, Voltage -how much the cell delivers higher voltage used more by oval racers, Internal Resistance - here the lower the internal resistance the best cell can produce voltage, Charge rate - not to important but may effect the test numbers, Cutoff (runout) - 5000 is best those using 360 can cause runtime to look low and voltage look high, Test Load (20,25,30,35)- matchers feel the higher testing weeds out weak cells that may get by a lower loads - you can still get packs to perform good regardless of what you use in the end for discharging. Some of the 20amp packs I purchased performed fine once I began using a 30 or 32amp load. In 19turn oval 400/1.179/.25 lower IR are what I look for, mod oval 420+/1.179 1.18 if available stock offroad / TC 400/1.159 to 1.16 (1.17 voltage tend to produce too much low end punch on a loose/loamy dirt track).